There are a number of qualified brushless motor manufacturers with products on the market today.
In this article we are focusing on big and powerful brushless motors that have at least 9 kW of power. These are motors for heavy lift drones, eVTOL, electric airplanes, and more.
Skip to the end to see the performance summary table.
In this article we will provide an in-depth comparison of these motor brands:
• magniX |
• H3X
|
• EMRAX
|
|
• Safran
|
|
• T-Motor
|
• Turnigy
|
• Thin Gap
|
• KDE Direct |
• Fly Dragon | |
• Plettenberg | |
• Geiger Engineering | • FreeRCHobby |
Need to test large motors and propellers up to 150 kgf? Check out the Flight Stand 150 test stand.
magniX has become a household name for those following the development of the commercial electric aircraft industry. Through several fruitful partnerships, magniX motors have powered some of the most promising electric aircraft ever tested, including Harbour Air’s electric Beaver seaplane as well as the largest fully electric commercial aircraft to date, the Cessna eCaravan. Noting that in 2019 over 45% of commercial flights were less than 500 NM, magniX hopes to electrify these routes for improved environmental sustainability. Currently magniX lists two electric propulsion units on their website, the specs for the magni650 EPU are below.
Specifications:
Figure 1: Electric Beaver seaplane by magniX and Harbour Air (Photo from Harbour Air)
A new and ambitious player on the scene, H3X is a US-based company with aims to power large commercial aircraft like the Boeing 737. Their unique design combines the electric motor, inverter, and gearbox into a single unit. They currently have one model ready to order, the HPDM-30, with two larger models available for pre-order. Below are the performance characteristics for the HPDM-30:
Specifications:
Further reading: Automated Tests for Your BLDC Motor
The idea for EMRAX was born in 2005 when founder Roman Susnik flew Slovenia’s first flight in an electric aircraft. The flight ended in an emergency landing when the aircraft’s brushless motor failed, inspiring Susnik to develop a suitable motor himself. By 2008 a functional prototype was created and tested, and EMRAX had several requests for collaborations. Now in 2020, EMRAX engines are already used in a variety of electric gliders and aircraft. The motors possess a high power density and air-cooling, making them ideal for aviation applications. The performance figures for their largest model, the EMRAX 348, are summarized below.
Specifications:
Based in Czechia, MGM COMPRO and Rotex collaborate to bring advanced motor-ESC systems to the electric aviation market. They are involved in many innovative projects including eVTOL vehicles, tilt-rotor aircraft, and personal flying machines. Already on the market are the VELO self-launching gliders developed in partnership with GP Gliders and the ΦNIX all-electric plane by the Pure Flight consortium. Here are the specs for their largest off-the-shelf motor, the REB 90:
Specifications:
Figure 2: VELO glider by MGM COMPRO and GP Gliders (Photo from GP Gliders UK)
NeuMotors manufactures both inrunner and outrunner brushless motors. They have been active since 2002. Their largest motors are their 180xx series, whose specs are below.
Specifications:
With a strong electrical engineering portfolio, it is no surprise that Safran has developed their own brushless motors for use in the aviation industry. Called ENGINeUS, the line of smart motors boast some unique features, such as motor controller integration and a sealed design for use in harsh environments. The ENGINeUS is already used in Bye Aerospace's eFlyer, and Safran has announced that they now have hybrid motors available for commuter-sized aircraft. The performance characteristics for the largest fully electric ENGINeUS motor are outlined below.
Specifications:
Further reading: Comparing Drone Power System Designs
T-Motor remains a drone-focused company, but their high-powered brushless motors are suitable for much more than quadcopters. Their U15L, XL and XXL models in particular are highly relevant for the manned aviation industry, producing up to 98 kg of thrust per motor. While no manned applications are advertised to date, theirYouTube channel highlights some of the larger UAVs and eVTOL vehicles that their motors can power. Below are the performance specs of their U15XXL motor:
Specifications:
Having been around for over 20 years, Turnigy takes pride in keeping prices low for consumers through their ‘slick’ supply chain. They offer nine main models of motors, each suited to a unique function. While they are focused on the drone and RC industries, their motors offer sufficient power for certain manned flights, such as manned multicopters and paramotors. They also offer their own pre-made mini paramotor. Below are the specs for Turnigy's largest motor available on Hobby King's website:
Specifications:
Designed with both civilian and military uses in mind, Thin Gap motors are known for their toughness and high torque: weight ratio. They also prides themselves on their proprietary ironless stator and no cogging design, which allows for high power density, smoothness and efficiency. Looking to break into the VTOL industry, Thin Gap partnered with Aurora Flight Sciences to develop a powerful fan system for lifting heavy unmanned vehicles. The research contributed to the design of the Aurora LightningStrike XV-24A, an electric Osprey-like project that has been discontinued. Below are the performance specs for their LSI 267-58 motor.
Specifications:
Figure 4: Thin Gap motor kit (image from Thin Gap website)
KDE Direct approaches manufacturing with the mission to power future-focused applications such as flights, satellites and space exploration. Their brushless motors are designed for longevity and zero-vibration operation, ideal for smooth flights and missions. While they are largely used in drones and UAVs, they also offer sufficient power to lift heavy lift drones. The performance range for KDE Direct’s brushless motor fleet is outlined below.
Specifications:
Further reading: Why You Should Measure Your Brushless Motor's Torque
Pipistrel is a leader in the electrical propulsion industry, with several different models of fully-electric aircraft available for purchase. Their Velis Electro aircraft was the first ever electric-powered aeroplane to receive a type certificate from the EASA, paving the way for the development of the industry. They also offer electric self-launching gliders and hybrid general aviation style aircraft. Powering these aircraft is their E-811 liquid-cooled engine, capable of 49.2 kW of continuous power. Here’s a full summary of its basic characteristics.
Specifications:
Figure 5: The fully electric Velis Electro Aircraft (Photo from Pipistrel)
The brushless motors created by MAD Components were "born for" heavy lift and manned drones, to put it in their words. Each rotor equipped with a MAD Components motor is designed to offer a maximum thrust of up to 114 kgf per rotor. Designed for both commercial and recreational uses, you will see MAD motors in aircraft ranging from emergency rescue drones to paragliders. Below are the specs for their top performing motor, the MAD TORQ M50C35 PRO EEE 34 KV.
Specifications:
ePropelled is determined to shake up the brushless motor industry for eVTOL and aviation by developing solutions that meet the power-to-weight ratio needs of the trade. Their mandate specifically mentions the delivery of “propulsion systems for unmanned aerial vehicles (UAVs) and electric vertical takeoff and landing aircraft (eVTOL).” Their patented technology “produces a dramatically more energy-efficient method of electric propulsion that increases flight range and life expectancy of the battery pack”, allowing the manufacturer to save money on battery costs. Below are the specs for their largest motor, the PM1200L.
Specifications:
Fly Dragon is a Chinese company operating in Chengdu. They offer three motors for heavy lift drones, the largest of which is the G110. All of their heavy lift motors come as a kit that includes a propeller and ESC. The specs for their G100 motor operating at 500 V and 80% throttle are below.
Specifications:
Figure 6: G110 propulsion kit from Fly Dragon (photo from Fly Dragon website)
Plettenberg’s guiding principle is to achieve maximum efficiency at the lowest possible weight. They have several off the shelf electric motors available for purchase and they also do custom projects on request. According to their website, their standard solutions “offer efficiency levels of up to 95 % and power-to-weight ratios of up to 26 kW/kg.” Their Nova 50 motor is the most powerful in their fleet, full specs below.
Further reading: How Brushless Motors Work and How to Test Them
Beyond Motors manufactures a range of axial flux motors for various applications including aviation, marine, automotive and racing. They use a 'patent pending' water cooling system that enables high efficiency and continuous power. Their range of motors starts at a continuous power of 75 kW and continues all the way up to 230 kW continuous with a 430 kW peak in their AXM4 motor. And for those who need it, their motors are designed to be stacked up to 3 in a row for triple power within the same diameter.
Specifications:
Geiger Engineering manufactures electric motors, propellers, and accessories for aviation applications. What makes their motors unique is their Duplex option, which contains two identical power systems with homogenous redundancy to protect the aircraft in the event of one system failing. Below are the specs for the largest member of their High Power Direct (HPD) family of motors.
Specifications:
The MP240150 is Free RC Hobby’s largest off the shelf motor. It offers customizable parameters for sensors, cooling, and Kv. From their website, their slogan is “we can do what you need"!
Website: https://www.freerchobby.cc/
The goal of this report is to demonstrate a simple process for extending your multicopter flight time. Many concepts presented here also apply to fixed-wings vehicles. The concepts applied here can also be used to increase your drone's lift capacity.
Note: We will be using the Series 1585 thrust stand to gather motor and propeller data for this article.
The optimization process is a loop, so we need to start by making some assumptions. First, let’s assume that the drone already flies, so you have an existing design from which you know the weight and the battery size. Additionally, we will optimize a multi-rotor that is mostly hovering (our drone is neither a racing drone nor a competition drone).
Further reading: Drone Design Calculations and Assumptions
Figure 1: The Otus Quadcopter
To better illustrate the process, we will use the Otus Quadcopter as an example, though this method is applicable to any flying UAV.
The first unoptimized version of our quadcopter has:
With this current design, our flight time is ≈ 4 minutes.
The first step is to understand how a drone can fly and take-off. In order for a drone to become airborne, it must overcome its weight and drag. A drone’s weight is the product of its mass times gravity and drag is the force resisting the drone's motion through the air, which is dependent on reference area, air density and flow velocity.
The rotation of the propellers generates thrust and allows the drone to rise and maintain flight. At hover, we can assume that the combined thrust of the propellers is equal to the drone’s total weight.
From this assumption and with the weight of the drone, we can deduce the thrust required by each propeller in order to maintain hover. Here, our drone weighs 777 g, so we need a total thrust of 7.6 N to hover or 1.9 N per propeller.
Figure 2: Drone weight and hover thrust requirements
To keep a good control authority, the maximum thrust achievable by the propeller should be about twice the hovering thrust. Keep in mind this is just a recommendation. Racing quads will have a much higher maximum thrust to weight ratio.
We are looking for the most efficient propeller producing 1.9 N of thrust that has a maximum size of 6 inches, and can achieve 3.8 N of peak thrust.
We can vary parameters such as pitch, size, profile, material, and brand to find our most efficient propulsion system.
Efficiency is the ratio of the output divided by the input. Here, the propellers convert mechanical energy into thrust.
First, make an initial propeller selection based on manufacturer data. Unfortunately, propeller testing is not standardized and you cannot compare the data provided by different manufacturers. Manufacturers often provide the bare minimum technical specifications, so it is up to the buyer to have some knowledge about expected performance before purchasing.
We have compiled a database of drone motor, propeller and ESC data that can help you evaluate propellers before purchasing.
Further reading: How to Use Our Database of Drone Motors, Propellers and ESCs
You can also use our RCbenchmark Series 1585 thrust stand to test all your propellers with the same motor and record thrust, torque, voltage, current, motor rotation speed, and vibration. Optionally, you can use the database upload feature of the app that will walk you through the test process with a test script.
We want to measure thrust, torque and rotation speed. Propeller data is independent from motor data when you rely on torque and speed. The thrust of a specific propeller depends only on the propeller speed and the incoming air speed, not on the motor powering the propeller.
Regardless of the motor you choose, the thrust generated will be the same at a given rotation speed. This property is useful to check that your tests were performed correctly. The data points on a thrust vs. rotation speed graph for a single propeller tested with multiple motors should all be very close to the same line as in the image below.
When you have determined the desired torque and speed for the most efficient propeller at hover, you can perform a search for the motor that is the most efficient at this torque and speed. But first we need to find our propeller.
Let’s focus on only 3 propellers to keep this simple. We will test them with the Series 1585 thrust stand.
The test can be done manually or with a script. We performed the test with a script and compared the results to those in our drone propulsion database. Here is a comparison of the propeller mechanical efficiency (N/W) as a function of thrust (N).
As shown in the graph, at 1.9 N, the most efficient propeller is the Gemfan 6030 at 0.077 N/W. We can rule out the two other propellers as they have a lower efficiency.
At a thrust of 1.9 N, the Gemfan 6030 operates at 1300 rad/s, where it also generates 0.0184 N.m of torque (figure 6).
Now that we have found a propeller, our next step is looking for the most efficient motor at the operating point of 0.0184 N.m and 1300 rad/s. We will limit our search to 2 different motors for the purpose of this tutorial, but in reality, there are many more candidates to choose from.
The graph in figure 8 shows the mechanical efficiency of the tested motors when they are equipped with a Gemfan 6030 propeller. At 1.9 N of thrust, the efficiency of the Multistar is 68% while the efficiency of the EMAX is 60%. Thus, we conclude that for this specific propeller at hover, the most efficient motor is the Multistar Elite 2306.
The graph above allowed us to observe the efficiency difference for the torque-speed lines of two motors. We did not determine that the Multistar is a better motor in general, only that it performs better for this specific propeller. The Emax motor may be more efficient than the Multistar with another propeller.
Another thing we must check is that this motor is also capable of generating the peak propeller thrust for sufficient control authority. Earlier, we said we are looking for the most efficient propeller at 1.9 N of thrust that can also achieve 3.8 N of peak thrust.
We confirm graphically that this motor is capable of generating the 3.8 N peak thrust we need. At 3.8 N, the motor is capable of generating 0.035 N.m of torque at 1783 rad/s.
Once the motor and propellers are chosen, we can select a suitable ESC. For now, we just pick an ESC capable of delivering the motor’s peak current of 7 Amps with a safety factor. We will choose the Afro Race Spec Mini ESC which supports 20 Amps. There are some optimizations that can be done on the ESC, but that is outside the scope of this article.
Finally, we can use the information we have gathered to determine our flight time:
The capacity of the battery (Ebattery in Wh) can be expressed as the Flight Time (FT) in hours, multiplied by the generated power (Power in Watt).
The battery capacity (Ebattery) is equal to the weight of the battery (Wbattery in grams) multiplied by the energy density (sigmabattery in Wh/g).
The total power (Power in Watt) is equal to the weight of the drone (Wdrone (g) = Wframe (g) + Wbattery (g)) divided by the propellers efficiency (propefficiency in g/W).
The propeller efficiency is a function of the total weight of the drone divided by the number of propellers on your drone.
So by combining the equations, we obtain the flight time [eq1]:
An increase in the weight of the battery increases the division term in the equation above and reduces the propeller efficiency. In this way, the battery can influence the flight time in both directions. Larger batteries have a higher capacity, which can increase flight time, but they also have a higher weight, which increases the thrust required for operation, thus decreasing the flight time. Figure 11 shows an example of how to find the 'sweet spot' when balancing battery capacity and weight.
Those formulas are implemented in the spreadsheet here:
Use our Google sheet to easily calculate drone flight time.
There are some assumptions that you must add when using the Google sheet, such as the weight and capacity of your battery. Also, write the total weight of your quad (WITHOUT the battery) and the number of propellers. Now you have everything you need to calculate the flight time!
Further reading: Drone Building and Optimization (eBook)
You can observe the effect of varying the battery capacity on the flight time. As you can see, increasing the battery to a 33.3 mAh increases the flight time by a few minutes, but reduces the control authority as the weight to thrust ratio decreases. Going even bigger shows very little benefits in terms of flight time, but increases noise and reduces control authority.
In this article we covered how to choose a propeller, a motor, an ESC and a battery for our drone, and we looked at how to compare efficiencies, analyze data and calculate flight time.
All these modifications can change the total weight of your drone, especially if you choose another battery. You may need to restart the analysis if you significantly change the weight of the drone.
As we mentioned at the beginning of the article, the drone building process is a loop that repeats itself as we learn about our components and performance through testing. We make assumptions, choose parts based on those assumptions, test the system, swap the parts and repeat the process.
Regardless of the tool used to capture the data, we strongly recommend that you measure torque during your tests. This will allow you to analyze motor and propeller data independently and measure efficiency. We have designed multiple tools to make the data collection and analysis easier and more accurate.
The automatic testing capability of the Series 1585 combined with our database of drone propulsion data should allow you to select the best motor, propeller and ESC in a few hours of testing. The tests can increase flight time, lift capacity, and reduce the heat produced by the system, which increases the life of your components.
If you want to dig deeper in the subject of motor theory to fully optimize your motor, watch the video “How to Maximize Your Drone's Flight Time” on our YouTube channel. In this video we take you through the steps we took to build the Otus quadcopter and how we increased its flight time from 4 to 8 minutes and even up to 13 minutes!
For further reading on this optimization process, check out Part 2 of this article - The Drone Design Loop Calculations and Assumptions
]]>By Lauren Nagel
Brushless motors are seemingly everywhere - appliances, vehicles, tools, and more. But what is a brushless motor and what makes it different from a brushed motor?
Brushless motor means the same thing as BLDC motor. The full form of BLDC motor is brushless direct current motor.
In this article we will answer all the basic questions relating to this versatile type of motor.
Table of Contents
A brushless DC (BLDC) motor is a type of electric motor that relies on repulsive and attractive forces between permanent magnets and electromagnets to drive its rotation. BLDC stands for 'brushless direct current', which roughly describes how the motor works.
The motor’s rotation is managed by a controller that delivers timed bursts of current to the electromagnets in the motor, which in turn controls its speed.
In figure 1, you can see the permanent magnets on the rotor in red and blue and the electromagnetic coils on the stator in copper.
Figure 1: Brushless DC motor diagram
Brushless motors are known for having a high efficiency and long service life compared to alternatives. There are several common ways of describing this type of motor, including brushless motors, brushless DC motors, and BLDC motors.
Brushless DC motors were invented in 1962 by T.G. Wilson and P.H. Trickey, presented in their AIEE paper titled, “D-c machine with solid state commutation”.
Through the paper they outline the problems they wished to solve and the solutions they came up with. In their section on fundamental considerations, they outline the design challenges they must overcome to build a functional brushless motor:
“Since the required switching of armature conductors is to be done with static elements, it is desirable not to have the switching elements themselves rotate. This may be accomplished by placing the armature winding on the stationary member and causing the field poles to rotate. Use of a permanent magnet for the field element eliminates the need for supplying power of the rotor either through brushes or transformer action [...] the axis of the magnetic field produced by the armature current must also be made to rotate and to do so at exactly the same speed as the rotor. To this end, information concerning the angular position of the rotor poles must be continuously transmitted to the stator so that exactly the right switching of armature conductors can take place”.
In their concluding paragraph, they outline some of the ways they foresee their invention will provide an advantage over brushed motors. The improvements they suggested have turned out to be incredibly accurate:
“It should practically eliminate radio interference in the normal sense as well as wear and maintenance now encountered on brushes and commutators. The machine should have extremely high reliability as well as long life.”
The main counterpart to brushless motors, the brushed motor, was invented over 100 years earlier in 1856. While other advances had been made since then, this major technological overhaul was well overdue.
The brushless motor relies on two key parts in order to function: 1) The rotor holding permanent magnets, 2) The stator holding copper coils that become electromagnets when a current is sent through them.
Brushless motors may be inrunners, where the stator is located on the outside and the rotor rotates within, or they can be outrunners, where the rotor rotates outside the stator.
For a more detailed explanation of the brushless motor mechanism, check out our article on How Brushless Motors Work.
Figure 2: An inrunner motor (left) and an outrunner motor (right)
When a current is delivered to a coil of the stator, it becomes an electromagnet and develops a North and South pole. When the polarity of the electromagnet matches that of the permanent magnet it faces, their like poles repel and the rotor spins.
If the current maintained this configuration, the rotor would spin briefly then stop once opposite electromagnets and permanent magnets lined up. For that reason, the current is delivered as a three-phase signal in a way that constantly changes the polarity of the electromagnets so the rotor keeps spinning.
The motor spins at a speed equal to the frequency of the three-phase signal, so if you want the motor to go faster, you increase the frequency of the signal. With a remote-controlled vehicle, the speed is increased by increasing the throttle, which tells the controller to increase the switching frequency.
In order to know how and when to energize the coils, the motor uses Hall Effect sensors to determine the relative position of the rotor and the stator. That way, the electromagnets in the stator are activated in the correct order at the right time, and the motor keeps moving.
The brushless motor mechanism is most easily demonstrated with a diagram / GIF, which you can see below.
Figure 3: Brushless motor GIF
Brushless motors may be inrunners or outrunners. The one in the diagram above is an outrunner because its rotor containing the permanent magnets is located outside of the stator. Each permanent magnet is also called a ‘pole’ and the pole count of the motor can affect its performance.
The permanent magnets are repelled by the electromagnets as they line up, pushing the rotor around in a circle while the stator remains stationary, in this case counter-clockwise (figure 3).
The motor controller on the right, also called an electronic speed controller (ESC), is connected to the power source or battery on one side and the motor on the other. It can be connected directly to the throttle input device or remotely such as through a radio signal. The ESC takes the frequency of the throttle signal from the controller and tells the motor how fast to spin by adjusting its switching frequency.
Further reading: How to Control Electric Motor Speed
Brushed motors are another kind of electric motor that relies on the concept of magnetism to drive its rotation.
The main differences between brushed motors and brushless motors are their mechanism, efficiency, and service life. The differences are summarized in this table, then expanded on below.
Brushed Motor |
Brushless Motor |
|
Commutation |
Mechanical |
Electrical |
Efficiency* |
Lower (75-80% typ.) |
Higher (85-92%+ typ.) |
Heat |
Higher (20-25% of input power) |
Lower (8-15% of input power) |
Initial cost |
Lower |
Higher |
Maintenance requirements |
Higher |
Lower |
Control of motor speed |
Simple |
More complex |
Lifespan |
Lower |
Higher |
Electrical noise |
Higher |
Lower |
Ignition safety |
Good |
Very good |
Low speed control |
Good |
Good |
Maximum speed |
~20,000 RPM |
>100,000 RPM |
Torque to weight ratio |
Lower |
Higher |
* Any motor running outside of its efficient range of torque and speed will have a low efficiency (<50%) and high heat production.
Commutation: In a brushed motor, the current is transmitted from the commutator to the motor windings through physical contact with the brushes. In a brushless motor, the current is controlled electrically through the switching on and off of the stator coils via semiconductor switches.
Efficiency: Brushless motors provide a higher amount of torque per watt of power drawn than brushed motors, thus making them more efficient.
Initial cost: Brushless motors are generally more expensive than brushed motors of the same size due to the increased complexity of their controllers.
Maintenance requirements: Brushed motors tend to wear down quickly where the commutator meets the brushes, and components thus require more frequent replacement. Brushless motors don’t have any components that experience wear to the same extent, so maintenance / replacement is required less frequently.
Complexity of motor speed control: In a brushed motor, the speed of the motor is simply controlled by the voltage applied to it. In a brushless motor, the motor controller uses a 3-phase signal to control the rotation speed, thus the signal delivery is more complex.
Lifespan: Brushless motors have a longer lifespan compared to brushed motors because they do not experience the same wear and erosion between components. That said, brushed motors can sometimes be rebuilt to extend their service life.
Electrical Noise: In addition to the acoustic noise produced by motors, they can also produce electrical noise that can interfere with other components of the system. Brushed motors tend to produce more electrical noise than brushless motors, which can cause electromagnetic interference with local circuits. Without proper isolation, this interference can cause circuit malfunction and reduced performance.
Ignition safety: Brushless motors are much less likely to produce sparks that could ignite flammable material.
Low speed control: Brushed motors have a simple control system that operates well at low speeds. When combined with angular encoders and control electronics, brushless motors can also rotate at very low speeds. This is ideal for applications such as hoverboards, or high power servo motors. ODrives offers an open source controller for such applications.
Maximum speed: Due to the friction inherent to the brushed motor design, they are not suitable for use at high RPM. The brushes in the motor tend to heat up with increased speed, which leads to increased wear and a high temperature situation. Brushless motors, on the other hand, do not encounter this problem due to lack of brushes, so can be run at very high RPM.
Torque to weight ratio: While brushed motors can produce significant amounts of torque, especially at low speeds, brushless motors that produce a comparable amount of torque are significantly smaller, thus giving them an advantage in their torque : weight ratio.
The previous section discusses the important differences between brushed and brushless motors, but the question remains, are brushless motors inherently better?
As with many engineering questions, the decision comes down to how you plan to use the motor. We know that brushed motors aren’t ideal for high RPM applications, so brushless motors would prevail in those areas. The lower efficiency and high maintenance cost of brushed motors also makes them less desirable for certain uses.
That said, brushed motors have a lower up-front cost, simple speed control, and produce smooth, high-torque motion at low speeds. They also have a linear torque-speed relationship, which makes controlling them easier. Even though they wear out over time, those other characteristics make them the ideal choice for certain devices such as low-speed industrial equipment, blenders, windshield wipers, and mobile medical equipment.
Figure 4: Brushed motors in windshield wipers
On the other hand, brushless motors tend to shine in areas where high RPM is required, such as saws, fans, and propellers. They also have an advantage in their torque : weight ratio, which is important for use in certain vehicles such as drones and electric aircraft as well as in areas where a small but powerful motor is required, such as microrobotics.
BLDC motors are also favourable for applications where high usage is expected and high efficiency is required, such as cordless tools, RC vehicles, heating and cooling systems, and industrial engineering. One example is in Light Detection and Ranging (LiDAR) systems, which are used in robotic navigation and localization systems, such as self-driving cars and surveillance drones.
Figure 5: The RPLiDAR System used for robotic navigation applications
Let’s dive a bit deeper into some of the applications for brushless motors.
Industrial engineering: Brushless motors are useful for many industrial engineering processes due to their high torque and longevity. You may see them used in CNC machines, linear motors and servomotors. They are also used as actuators to control movements in industrial robots for tasks such as painting, product assembly, and even welding, such as in Tata Steel's UK facility (figure 6).
At home: Due to their low weight and high efficiency, brushless motors are a top choice for many home appliances. They are found in virtually every room of the house in machines such as vacuum cleaners, coffee machines, hair dryers, hard drives, printers, and more. Since many of these devices are produced in very high quantities, it is important for them to be of high quality and reliability to avoid dealing with customer returns and maintenance.
Cordless power tools: Since brushless motors can run on rechargeable batteries, they are ideal for cordless applications requiring minimal weight and high RPM. Cordless drills became available about a decade ago but now represent up to 50% of models available. This is one area where brushless motors must still compete with brushed motors, so price and frequency of use are key factors when deciding which model to go with.
RC Cars: Brushless motors can spin at very high RPM and deliver bursts of speed rapidly, so they are great for use in RC vehicles. The characteristics of the RC motor, such as motor Kv, size, and power rating, will vary based on the model. RC cars may also have nitro motors, which provide a longer run time, but come with the complications of buying gas and refueling the vehicle. Brushless motors tend to have an acceleration advantage over nitro motors, so they are generally better for racing.
Figure 7: Off-road Monster Truck RC Car powered by a brushless motor
The combined characteristics of brushless motors make them ideal for use in drones. They are lightweight, highly efficient, have a wide speed range and high torque capabilities, all qualities that are beneficial for UAVs.
Brushless motors for hobby drones are relatively inexpensive, which allows drone builders to test out different models to determine which ones are most efficient. Maintenance is rarely required due to their high reliability, and it is often less expensive to replace the motor than repair it.
For larger commercial drones and manned drones, larger brushless motors are being developed all the time to meet the demands for greater power, torque and thrust. These motors are contributing to the electric revolution seen in the transportation industry, which includes electrifying existing plane models and conceiving new eVTOL aircraft.
There are many types of drones ranging in size, style, power source, and function. Brushless motors are suitable for many applications, though gas-powered and hybrid motors are also used.
For more information on how to choose a brushless motor for your drone, check out our free eBook on Drone Building and Optimization.
Brushless motors can be found in almost every area of our lives and they are only gaining in popularity. From the home to industrial settings, these motors impress us with their compactness, efficiency and reliability.
To understand these fascinating machines requires a bit of background on electromagnetism and electrical engineering, concepts we’ve covered here.
If you have any remaining questions, don’t hesitate to leave us a comment below, or check out the other articles in our knowledge centre.
]]>By Lauren Nagel
There are hundreds of drone batteries on the market and it can be difficult to decide which one is right for your build. One of the main challenges is decoding the many performance indicators listed on the battery, such as capacity, discharge rating, and cell configuration.
Choosing the right battery for your drone will not only help improve performance, but also increase it’s lifetime. Just like your motor and propeller, your battery will eventually fail, but this can be postponed by choosing the correct one and operating it safely.
In this article we will cover:
Figure 1: LiPo battery pack used in Otus quadcopter drone
The most common batteries used in drones are lithium polymer (LiPo) batteries. LiPo batteries are composed of a lithium-based cathode and anode separated by a polymer electrolyte.
LiPo batteries differ from other lithium-ion (Li-ion) batteries in that they have a solid polymer electrolyte component rather than a liquid electrolyte. Common polymer electrolytes may be dry, porous or a gel, and include poly(methyl methacrylate) (PMMA), poly(acrylonitrile) (PAN), poly(vinylidene fluoride) (PVdF), and poly(ethylene oxide) (PEO).
The science behind LiPo batteries is the same as in other Li-ion batteries: chemical energy is converted to electrical energy when electrons travel from the battery’s anode to its cathode, creating an electrical current. The cathode contains a lithium metal oxide (such as lithium-cobalt oxide (LiCoO2)), which provides lithium ions, whereas the anode contains a lithium carbon (such as graphite).
The anode and cathode are separated by an electrolyte that interacts with the anode to generate electrons, which creates a charge gradient in the cell. As the anode becomes negatively charged, the electrons travel along a conducting wire to the cathode. The whole system thus undergoes an electrochemical redox reaction (reduction/oxidation): the anode loses electrons and becomes oxidized while the cathode gains electrons and is reduced.
Figure 2: LiPo battery redox reaction
Lithium-based batteries have a higher energy density compared to nickel cadmium or nickel metal hydride batteries, which means they can provide more energy for less weight. LiPo batteries rival Li-Ion batteries in terms of energy density, but are especially popular because they are less likely to leak.
The energy density of LiPo batteries ranges from 140 - 200+ Wh/kg in terms of weight and 250 - 350+ Wh/L for volume. Volume energy density is important to consider when building a drone so the battery fits on the frame, but for performance calculations, the energy density by weight is more relevant. With higher density comes higher cost, so your budget may also be a limiting factor.
Figure 3: Relative energy densities by volume and weight of common battery types (Image: Open Impulse)
A technology that may soon rival LiPo batteries as the drone go-to are Sion Power’s Licerion batteries. These batteries boast an energy density up to 500 Wh/kg and 1000 Wh/L. They also have a 50% lower liquid electrolyte volume compared to other Li-Ion batteries. They were designed specifically for unmanned applications, notably high-altitude pseudo satellites (HAPS) and high-altitude long-endurance (HALE) drones.
Further reading: Types of Drones and UAVs
LiPo batteries have a greater risk of fire and swelling than older technologies due to their internal chemistry. Operating the battery at or beyond its limits can lead to the accumulation of oxygen atoms and build up of Lithium Oxide (Li2O), which creates greater internal resistance. More internal resistance leads to more heat, and the thermal runaway cycle begins.
Once the battery pack starts to swell, this is a good indication that the battery is damaged beyond repair or has reached the end of its life cycle. Using it beyond this point will further increase its temperature and potentially lead to a fire.
For common LiPo batteries, the nominal or average voltage is 3.7 V/cell with a maximum voltage of 4.2 V/cell. After the cell is fully charged, it will briefly provide 4.2 V before dropping to 3.7 V for most of the battery life. It becomes dangerous to discharge the battery after the cell voltage has dropped below 3.2 V because the resistance in the battery increases, causing it to heat up and swell, resulting in damage.
Figure 4: LiPo battery on fire (Credit: alishanmao)
To avoid this, many motor manufacturers have added a low voltage cutoff (LVC) to their controls, which stops them from drawing charge after a certain threshold, usually in the range of 3.2 - 3.4 V. Overcharging a LiPo battery is equally dangerous and can result in overheating and even an explosion.
Lithium ion battery fires are classified as Class B flammable liquid fires, so a type ABC or BC fire extinguisher should be used to put them out. These extinguishers stop the chemical reaction from occurring and eventually put out the fire.
For more LiPo battery safety tips, check out The Drone Girl’s article on “15 things every LiPo battery user should know”.
LiPo batteries are labeled with a few important pieces of information, including: battery capacity, voltage, cell configuration and discharge rate (figure 15):
Figure 5: Common LiPo battery labels
Battery capacity is given in mAh or Ah and can be used to estimate your flight time (more on this later). Battery capacity is more specifically defined as the number of hours of current or power the battery can provide. Common units are the ampere-hour (Ah) and the watt-hour (Wh). If a battery has a capacity of 1 Ah, you can draw 1 A of current for one hour. If the capacity is 1 Wh, the battery would provide 1 W of power for one hour.
The voltage rating of the battery will allow you to determine your motor speed and amperage. Since motors are rated in Kv with the unit RPM/Volt, the number of volts your battery can supply will determine how fast your motor will spin.
Further reading: How to Calculate Brushless Motor Kv
You can cause damage to your circuit or even cause a fire if your voltage rating is too low or your current drawn is too high, so it is important to choose your battery voltage carefully. To determine the maximum current that your drone will draw, set up your motor and propeller with a propulsion test stand and run them at maximum throttle. The current recorded at 100% throttle tells you the maximum amperage your motor will draw, so multiply this by the number of motors to get the total current draw for your drone. The battery should be able to provide at least this amount of current to avoid overheating.
Another way to determine the maximum amperage you can draw from the battery is by multiplying the capacity in Ah by the C rating. For a battery rated for 5800 mAh/5.8 Ah and 25C continuous, the maximum current you can safely draw is 145 A (5.8 x 25 = 145).
The cell configuration is sometimes present on the label and describes the number and layout of LiPo cells in the battery. Recall that one LiPo cell has a nominal voltage of 3.7 V and several LiPo cells can be connected in series. A 4S battery would have four LiPo cells in series (S), giving a 14.8 V battery (4 x 3.7 V = 14.8 V). A battery might also have a code like 4S2P, which tells us that there are four cells connected in series and two cell sets connected in parallel (P), for a total of eight LiPo cells.
The discharge rate or the C rating is a measure of how quickly the battery can safely discharge. If a battery has a C rating of 25 and a capacity of 5800 mAh/ 5.8 Ah, you could safely discharge it at 25 times the capacity of the battery, 25 x 5.8 = 145 Ah. With continuous power at that rate, the 5.8 Ah battery could be discharged in 2.4 minutes ((5.8 / 145) x 60 = 2.4). Batteries may also have a range or ‘peak’ discharge rate, where the battery may exceed its constant power output for a short period of time without overheating, such as during a sudden climb or correction. A higher C rating is great for applications like drone racing that require bursts of speed, since the battery can deliver the charge needed very quickly.
The best battery for your drone is the one that best suits your application. If flight time is your main concern, you will want to reduce your mass and maximize battery capacity (equation 1), which is dependent on energy density and mass (equation 2).
Where:
E = capacity
σ = energy density
M = mass in grams (g)
If power and speed are your top priorities, you will want a battery that can deliver high amounts of charge quickly and without overheating, so you're looking for a high voltage and C rating.
Here is a summary of how each battery variable affects your performance:
Battery capacity
Further reading: How to Increase Your Drone's Flight Time and Lift Capacity
Figure 6: Drone flight time vs. Battery capacity
Voltage
Discharge/ C rating
Current draw
In order to demonstrate the process of choosing the right battery, we will use the example of building a drone for agriculture - it must have a long flight time, but we don’t need fast power bursts. We will use data we’ve collected on our propulsion system to inform our decisions.
We want to choose a battery with the highest possible capacity that keeps our drone’s total mass at a manageable 20 kg. The mass of our other components is 12.5 kg, so we have 7.5 kg of mass available for our battery.
We need a total of 196 N of thrust to keep the drone at hover where it will conduct most of its operations, which is equal to about 50 N per motor/ propeller combo: (20)*(9.81) / 4 = ~49 N. For our design, we want the drone to be able to operate at up to double its hover thrust, so we will require up to 100 N / propulsion unit.
Our motor manufacturer tells us that our motor’s maximum continuous current is 100 A. Multiplied by 4 motors, our battery would need to be able to provide at least 400 A of continuous current if operated at maximum capacity.
However, we can look at our motor’s thrust data to see that at double our hover thrust (100 N), each motor only draws about 27 A (figure 7). Multiplied by 4 motors, our drone will be drawing no more than 108 A total for the majority of its mission.
Figure 7: Current drawn vs. Motor thrust
In terms of voltage, the motor manufacturer did not provide any data, but we can return to our propulsion test results to see that the motor drew between 45 - 49 V in our operating range (figure 8).
Figure 8: Voltage vs. Thrust in drone operating range
We don’t need explosive speed for our drone, so we will look for a 14S / 51.8 V battery with a minimum discharge rating of 25C, weighing less than 7.5 kg.
Here are three candidates that fit this description:
To calculate the estimated flight time values in the table we used a revised version of equation 2. We converted Ah to Wh by multiplying the Ah value by 47 V and we used a fixed value of 10 grams/Watt for propeller efficiency:
Based on these results, battery C will give us the longest flight time (33 mins), and won't exceed our weight limit. The maximum current we can draw from this battery is 550 A (22 Ah x 25C = 550), which exceeds the maximum draw of our motor. Therefore, we will go with battery C for our build.
Choosing the right battery for your drone can help you increase your performance, flight time, and endurance. This process starts with understanding the different characteristics of batteries such as battery capacity, discharge rate, and voltage.
You can then determine the priorities for your build - do you need the longest possible flight time or the ability to deliver strong bursts of power? This all comes down to how you intend to use your drone.
Finally, testing your propulsion system can help you narrow down your candidates and find the best battery pack for the job.
Further reading: An eBook on Drone Building and Optimization: How to Increase Your Flight Time, Payload and Overall Efficiency
If you have any more questions, don't hesitate to leave us a comment below.
]]>By Lauren Nagel
So you want to make your drone fly longer. Fortunately, there are many ways to do so.
This article has useful tips for drone users of all kinds - whether you’ve bought a ready-to-fly drone or you’re building one yourself.
Some of the suggestions could add minutes to your flight time whereas others may only give you an extra fraction of a second, but combined they can certainly make a difference.
This list also takes into account that some swaps will be possible for some drone applications but not for others. Cost is another important factor for many operators that will make certain suggestions easier to implement than others.
Propellers with fewer blades are generally more efficient. In theory, the most efficient propeller would actually have a single blade. In practice, however, having at least two blades is much more practical from a balancing perspective.
Propellers with a low pitch can also reduce energy consumption, as long as your drone is flying slowly. If you are flying quickly, it often makes more sense to have a propeller with a higher pitch.
Testing different propellers to find the most efficient ones is a great way to significantly increase your flight time.
Brushless motors that have low resistance, good cooling properties and high quality components are generally most efficient. Unfortunately, motor datasheets are not always reliable when it comes to reporting true efficiency figures, so it’s important to test them yourself to validate the numbers given. Including highly efficient motors in your drone is another way to get a good return on investment in terms of flight time.
Further reading: How to Test Your Drone's Motors
Testing various motors and propellers with a thrust stand like the Series 1585 can tell you which combination of components has the highest overall efficiency, giving you the highest possible flight time. Efficiency doesn’t stop at the individual component level, it is also important to make sure your components work well together.
Image: test drone motors and propellers with the Series 1585 thrust stand
Did you know that hovering draws more power than stable forward flight? Therefore flying your drone at a steady forward speed instead of hovering can help reduce power consumption and extend flight time.
There are so many things to consider when it comes to choosing the right batteries that this could be an article of its own. Actually, we did write an article on that, which you can check out here. Here’s a brief summary of the key points:
Avoid flying in windy or rainy conditions as the wind resistance can significantly decrease flight time. Extreme heat and cold are also hard on your battery and reduce performance, not to mention the possibility of total failure.
Generally speaking, flying at lower altitudes is where you will find lower wind speeds and less wind resistance, meaning less work for your drone and an increased flight time. Here are a couple resources for checking the weather conditions before your flight:
Vibration can cause unnecessary stress on your drone’s components and make them work harder than necessary, burning more power. Try using vibration dampeners or shock absorbers to reduce vibration and increase flight time.
Further reading: 16 Reasons to Test Your Drone's Motors and Propellers
LEDs and cameras can consume a significant amount of energy, so try to limit their use if possible, such as during transit portions of your flight. Depending on your needs, you could also consider using a smaller or lower resolution camera with lower power consumption.
Regular calibration can help your drone fly more efficiently and extend your flight time. Check your drone’s user manual for instructions on how to do so. Regular maintenance is also crucial to keep your drone in good condition, and this can extend its lifespan and improve its performance. You should clean your drone after every flight, inspect it for any damage, and lubricate the motors and propellers.
Using a high-quality charger can help you charge your battery more efficiently and extend battery life. Charging your battery too quickly can cause it to degrade faster over time, leading to gradually shorter flight times.
Drones with advanced power management systems can extend battery life and flight time by efficiently delivering the power needed to all components. Some power management systems include a voltage regulator, which can also help increase the lifespan of your components.
Keep a close eye on your battery life and avoid draining it completely as this can shorten its lifespan. Over-discharging puts stress on the battery cells and can lead to irreversible damage that may reduce the overall battery capacity. Listen to your low battery warnings!
Certain drone designs are superior for reducing drag and increasing flight time. Aerodynamic features include streamlined arms and frames, adjustable propellers and ducted rotors. With any of these components there may be a trade-off between drag and weight, so testing various options is ideal to get the best performance.
Certain ESCs are built for extra efficiency with different power-saving features. Some examples include having low resistance circuits, operating at higher frequencies, using advanced control algorithms, and good heat management.
Further Reading: How Does an ESC work?
If your drone’s center of gravity is not properly balanced, it can affect the drone's stability and handling, which in turn can increase your power consumption and reduce your flight time. When your drone is well balanced, it takes less power to maintain stability and altitude. Make sure to rebalance your drone if you add or remove components.
Features like sleep mode and low-power standby mode can help your drone use less power when not in flight, reserving more power for your next flight. This is ideal for applications that involve frequent starting and stopping.
There are several adjustments you can make to your flight controller settings that can reduce power usage and increase flight time. A few examples include: operating the controller in low-power mode, lowering the frequency of GPS updates, using stable mode in calm weather, reducing the max throttle etc.
Low resistance wires can reduce the amount of power lost as heat and ultimately increase flight time.
Products such as a heat sink or a cooling fan can be helpful in keeping your drone's components cool so they work more efficiently, thus maximizing your flight time. Again, it’s a trade-off between weight and efficiency.
When a drone is further from the controller, it may require more power in order to maintain a stable connection with the receiver. This is especially relevant if there are obstacles between the drone and the controller. To save on power and improve your flight time, operate your controller from as close to the drone as possible.
We hope this list has been useful and has given you some ideas for how to make your drone fly longer.
Motor and propeller testing is one of the most effective ways to increase your flight time, and we have several tools that can help you do so:
Testing your motors and propellers is the most effective way to evaluate performance and find areas for improvement in your design.
It allows you to make improvements in efficiency, which translates directly into increases in endurance, flight time, payload capacity, and more.
In this article we will cover both the how and the why of motor and propeller testing.
You must first ask yourself, what are your, or your end user’s needs? This question is important as it will help you know which parameters to optimize.
The final choice of power system depends not only on the airframe and payload, but also on your application.
Figure 1: NTM propdrive 35-30 brushless motor and Quantum 13-4 carbon fiber prop.
To fully characterize a motor and determine its efficiency, you need to measure the following parameters.
The parameters listed above are important because they allow you to determine the following key performance indicators:
The output speed is a function of the throttle, in %, and of the load or torque, in Nm. If you want to completely characterize a motor, you will need to test it with multiple input voltages and different loads. The throttle is changed with the controller, and the load is changed with the type and size of propeller.
Further reading: How Brushless Motors Work and How to Test Them
To find propeller efficiency, you will need to measure the following parameters:
These parameters allow you to determine:
Note that the mechanical power is the same for the motor and propeller. That is because all the motor’s mechanical power output goes into the propeller, since it is directly coupled to the motor’s shaft.
The overall efficiency of your system depends on how well your motor and propeller work together. Even with a motor and propeller that are highly efficient on their own, your system can be very inefficient if the two parts don’t match well.
Because these parts have a common link (the shaft), the overall system efficiency is expressed as:
System efficiency (g/watt) = Propeller efficiency (g/watt) * Motor Efficiency
where the system efficiency is in grams per watt of electrical power. Changing the motor, propeller, or even switching to another ESC will all contribute to changing the system efficiency.
Moreover, the efficiency value will only be valid for a specific command input and mechanical load. In practice, this means that you will need to test your motor over a range of command inputs and with various propellers (to vary the mechanical load) in order to get a full efficiency characterization.
In summary, to obtain motor and propeller efficiency you need to simultaneously record voltage, current, torque, thrust, and motor speed. By combining these readings you can extract the electrical and mechanical power, which in turn will give you the efficiency values.
The easiest and most effective way to perform these tests is with a thrust stand, a piece of equipment specifically designed for characterizing motors and propellers.
We used the Series 1585 thrust stand (5 kgf/ 2 Nm) to collect the data discussed below, but we also offer thrust stands measuring up to 500 kgf.
Further reading: Why You Should Test Your Drone's Motors and Propellers
In this article we will only cover static tests (we won’t talk about dynamic tests involving angular acceleration, estimating stall torque, etc…).
Before starting your tests, we recommend to:
A simple but effective test consists of ramping up the throttle in small steps, and recording a sample after every step. Before taking the sample at each step, allow the system to stabilize for a few seconds.
In the video at the beginning of the article, we demonstrate how we manually varied the throttle from 0 to 100% in 10 steps. This procedure could also have been performed using the RCbenchmark software’s automated test feature, which is covered in another tutorial.
Further reading: Automated Propulsion Test Scripts for Your Drone
The results obtained from our manual step test are shown in this CSV file.
You can analyze your output data using any plotting software. Below is an example of a plot we generated obtained using the data from the CSV file linked above.
Here we have compared propeller thrust and mechanical efficiency, to see at what points our propulsion system operates most efficiently.
You could then compare this data with data from other tests using the same protocol but different propellers to find the one that is most efficient with your motor.
For more examples of the types of analysis that you can do with this data, check out our articles on motor and propeller testing.
Liked this tutorial or have comments? Let us know below.
By Lauren Nagel
When it comes to drone performance, there is a necessary trade-off between two important factors: payload and flight time.
Having a good payload capacity is increasingly important for delivery drones, cargo drones, search and rescue drones, and more.
Being able to predict how payload will affect your flight time can help you set realistic goals and expectations.
In this article we will cover:
Figure 1: Agricultural drone carrying a payload
We can start with the assumption that adding payload decreases flight time.
Most people assume this intuitively, but why is it true? For the purpose of this article, we are assuming that the payload is increased while the drone’s components stay the same. Note that decreasing the airframe’s weight is a good option too and it has the same effect as reducing the payload’s mass.
It ultimately comes down to this: to overcome a higher weight, more thrust is required from the propellers, which requires higher RPM, which draws more power from the battery, thus decreasing the available battery life and flight time.
Let’s look at it in a bit more detail:
When you add mass to a UAV, you increase the amount of thrust required to lift it off the ground.
When a drone is hovering, it is in equilibrium so the thrust is equal to the weight:
Figure 2: Drone hovering in equilibrium
For most drones, you want their max thrust to be about twice as much as is required to hover, which allows them to take-off, climb, and operate smoothly. Ultimately, greater mass = greater thrust requirements.
Further reading: How Much Payload Can a Drone Carry
There are a few ways to calculate thrust, but it is generally dependent on several factors, including propeller diameter and pitch, air density, velocity, and rotation speed (RPM), as demonstrated by this equation for dynamic propeller thrust:
Based on this equation, derived by Electric Aircraft Guy, there are a couple ways we can increase thrust: we can change the dimensions of the propeller, or we can increase the RPM. This equation is quite generic. For a more accurate thrust, consider testing with a thrust stand.
Since we are keeping our propellers constant, we will therefore have to increase the RPM.
To increase the RPM, we have to draw more electrical power from the battery, specifically more voltage.
Learn more: watch this video to see how voltage affects motor speed.
Battery capacity is often expressed in Watt-hours (Wh). If you have a capacity of 70 Wh, you can draw 70 watts of power for 1 hour or 1 watt of power for 70 hours.
The more power drawn, the shorter the life of the battery. And this is where we come full circle.
Since we will be drawing higher voltage (and therefore greater power) to reach a higher RPM, the battery life will deplete faster, and our flight time will be lower.
Next let’s look at this in practice. We have developed a drone flight time calculator that allows you to input your drone’s information, such as drone and battery weight, battery capacity, and the # of propellers. It combines this information with propulsion data and estimates your flight time for you.
For this article, we will borrow data from this propulsion system with a T-Motor MN4014 motor and an APC 12x6e propeller in order to demonstrate the relationship between payload and flight time.
Figure 3: Drone Flight Time Calculator with sample data
With zero payload and the mass of the drone at 1.6 kg (battery included), we get a flight time of 39.6 minutes.
As we increase the payload by 0.2 kg increments, watch what happens to the flight time:
Figure 4: Drone Flight Time vs. Payload graph
As payload is gradually increased, flight time decreases.The equation below shows how battery capacity and power draw relate to the flight time of the drone. The more power you draw, the shorter the flight time will be:
There is one simple way that you can increase both the flight time and payload capacity of your drone, and it all comes down to one key variable: efficiency. Increasing the efficiency of your propulsion system reduces the power drawn from the battery, thus increasing your flight time.
Improving efficiency is not the subject of this article, but we have a full guide on How to Increase Drone Flight Time and Lift Capacity, which you can read if you are looking to improve your drone’s performance.
In this article we have demonstrated how drone payload affects flight time. The key variables are thrust, RPM, electrical power, and battery capacity.
If you have any outstanding questions or topics you’d like us to cover, leave us a comment below.
]]>By Lauren Nagel
If you are building a drone, RC vehicle, or any machine that uses a brushless DC (BLDC) motor, you have likely encountered the concepts of motor Kv and motor poles. Both of these parameters are useful for characterizing your motor and estimating its performance.
In this article we cover what these metrics mean, how to calculate and measure them, and finally how they are related.
After performing the calculations, we will use the Series 1585 thrust stand to confirm our results.
Motor Kv provides a way to describe the relationship between the peak voltage and rotation speed in a brushless motor in a no-load condition. The unit for Kv is RPM/V and it can be estimated by dividing the rotation speed of the unloaded motor with the applied voltage.
In truth, the voltage we should consider is in fact the back electromotive force (EMF) and not the applied voltage.
In a brushless motor, the back electromotive force (EMF) is a voltage that occurs in the opposite direction of the current provided by the power source, induced by the movement of the coils through the magnetic field in the motor.
This can be explained by Faraday’s law, which shows that a coil moving inside a magnetic field creates a flow of electrons inside the coil, called a voltage or emf.
When the motor turns, the back EMF generated is proportional to the speed of the rotor: as RPM increases, back EMF does as well. At full throttle with no load (ignoring the motor’s own inertia), we can describe the relationship with this formula:
The Kv rating provides an estimate of how many rotations a motor will undergo for every volt applied to it. This rating can be helpful for comparing motors that are of the same size physically, but have different performance characteristics due to their inner workings.
As a general rule, as the number of windings in the coils increases, the Kv of the motor decreases. Mechanically speaking, low Kv motors have a higher number of windings of a thinner wire and the thin wire carries more volts at lower current.
High Kv motors have fewer windings but a thicker wire that can carry higher current with fewer volts. The equations in the next section provide a mathematical demonstration of this concept.
Further reading: BLDC Motor Power and Efficiency Analysis
Applied to drones, low Kv motors tend to operate at lower RPMs and produce more torque, ideal for larger propellers and large drones. High Kv motors operate at higher RPMs and are ideal for low torque, small, and fast spinning propellers.
Therefore, lighter, fast moving drones such as racing quads are best served by high Kv motors and heavier, slower moving drones are better served by low Kv motors.
Figure 1: The EMAX Hawk Pro racing drone uses a 2400 Kv motor whereas the Foxtech Gaia 190MP heavy lift drone uses a 100 Kv motor
Another practical application is for inrunner and outrunner motors. An inrunner typically has a higher Kv value than an outrunner of the same size. The larger diameter of the rotor in the outrunner allows for more permanent magnets.
More magnets (poles) → lower speed → lower Kv. The small diameter of the inrunner also means it has a smaller circumference to cover for one rotation, thus more rotations for the same voltage.
It’s important to note that Kv is not a definitive way to evaluate a motor’s performance as the brand of motor can also affect its efficiency. For the same Kv value, a motor from one brand may perform better or worse than another, which will only be revealed through testing.
If we go back to our earlier formula:
We can rearrange it to show that:
To approximate Kv, you can swap ‘Back EMF’ with input voltage in the equation, which will give you a pretty good estimate of motor Kv.
For a more accurate estimation, you can measure the voltage between two leads in the circuit to obtain your root mean squared (RMS) voltage, then multiply it by sqrt(2) or 1.414 to obtain peak voltage, which you can plug into this equation:
0.95 is a value that accounts for deviations from the theoretical model to provide a value closer to that which is experimentally observed. This factor is commonly accepted for Kv calculations.
Measuring Kv experimentally is perhaps the best way to get an accurate idea of your motor’s true Kv. We have regularly observed differences between the stated and measured Kv of the motors we’ve tested, which you will likely observe as well.
We provide a pre-written script in our RCbenchmark software that allows you to measure Kv experimentally by running your motor through a no-load test.
All you have to do is hook up your motor to an RCbenchmark test stand, input the number of motor poles in the Setup tab, then run the test (without a propeller). The software will calculate your motor’s Kv automatically once the test is finished.
Even if your motor is labelled with a Kv rating, it can be interesting to test it to see if the experimental Kv matches what is advertised.
Figure 2: The automatic script that calculates Kv using the RCbenchmark software
Within a brushless motor, there are a number of magnets lining the circumference of the rotor (figure 1). These magnets are also referred to as the ‘poles’ of the motor. When an electric current is delivered to the coils in the stator, the rotor starts to turn as its magnets repel the like electromagnets.
In the diagram below, the blue rotor contains the permanent magnets in an inrunner motor (left) and an outrunner motor (right). The green stator holds the electromagnetic coils.
Figure 3: Motor magnets / poles in inrunner and outrunner motors
The magnets lining the rotor each have their own north and south end, with just one of those ends facing and interacting with the stator. There are an equal number of north and south magnets facing out from the rotor, and each set of N and S magnets is referred to as a ‘pole pair’.
For each pole pair in the motor there are two poles, so if the motor has 8 poles / magnets, there are 4 pole pairs. For this same reason, you will almost always have an even pole number (2, 4, 6, 8, etc.), as each magnet needs an opposite pole.
Ex.
The speed of the motor is inversely related to the number of poles in the rotor. The relationship is explained by the following formula:
Where:
Ns = synchronous speed
f = frequency of three-phase power supply
And #p = the number of poles
Alternatively, it can be written as:
Where:
Ns = synchronous speed
f = frequency of three-phase power supply
And #pp = the number of pole pairs
Since the number of poles is equal to the number of permanent magnets in the rotor, the simplest way to determine the number of poles is to count the magnets manually.
This is possible if you can see into your motor or if you are able to remove the outer casing. If you are unable to remove the casing this can be a challenging task as the magnets can be quite small and numerous.
Further reading: How BLDC Motors Work
Figure 4: Counting the magnets / poles inside a brushless motor
Other simple procedures have been suggested, like waving a magnet around your rotor to see where attraction and repulsion occurs. This method is most feasible for outrunner motors where the magnets are outside the stator near the surface.
Be careful if you attempt this technique as the magnets inside your rotor can be demagnetized by the influence of a stronger magnet.
Another option that requires a bit more effort and equipment involves finding the motor speed and power supply frequency experimentally, then rearranging the equation below to solve for the # of motor poles:
To measure your motor speed, you will need a piece of equipment to back-drive your motor (a drill, for example), plus a rotation speed-measuring device to measure the number of rotations, such as a tachometer.
You will also need a device to measure the frequency of the back EMF generated, such as an oscilloscope. Measure the speed and frequency simultaneously, then input the values into the equation above to determine the number of poles in your motor.
Perhaps the simplest way to measure the number of poles in your motor is to have a test stand determine it for you.
Several of our test stands come with automatic, pre-written scripts for calculating the number of poles in your motor, including the Series 1585.
Simply enter the Kv rating of your motor in the script, then run the automatic test with your motor installed on the test stand (without a propeller, as the motor will run at full speed).
The software will provide you with the pole count for your motor as soon as the test is done, which takes just a couple of minutes.
Figure 5: The automatic script that counts motor poles using the RCbenchmark software
As shown in the previous sections, motor Kv and the number of poles are both related to the speed of the motor / RPM. We know that as rotation speed increases, Kv increases as well. Alternatively, a higher number of poles corresponds to motors that operate at lower RPMs.
The relationship between the number of motor poles and Kv is thus inversely proportional. This makes sense when we think about it practically.
A larger motor with a greater number of poles will require high torque and operate at a low operating speed. The motor used would thus have a high number of poles and a low Kv.
A smaller motor with fewer poles will operate at high RPM and produce relatively low torque. The motor will therefore have a low pole count and a high Kv.
The concepts covered in this article, brushless motor poles and motor Kv, are two concepts that are key for understanding the properties and performance of BLDC motors. Knowing how to interpret these numbers can help you choose the best motor for your drone or electric aircraft and improve your performance and efficiency. Efficiency is directly related to flight time, payload and range, so this information is very valuable.
If you would like to be able to measure the number of poles in your motor and your motor’s Kv automatically, we suggest checking out our motor testing tools:
If you have any questions about the contents of this article, don’t hesitate to leave us a comment and we will be sure to respond.
Further reading: How to Test a Brushless Motor with a Thrust Stand
]]>The drone engineering process often operates as a ‘design loop’, which refers to the circular nature of the design process. Building the first version of the drone relies on certain assumptions, many of which will change as components are selected and the design comes together.
In this article we will cover:
We will be using the Series 1585 thrust stand to gather data for this article.
The design loop begins when the designer looks at how the first version of the design differs from the assumptions, then goes back to the beginning with the new information (figure 1).
Figure 1: The drone design loop illustrated
In our previous article, How to Increase a Drone's Flight Time and Lift Capacity, we covered the first stage of the design process and reached a first version of our design. In this article we will start where we left off, looking at how our assumptions held up.
We started our design process with the assumption that our drone would weigh 777 g and would be able to fly on its own. Following these assumptions, we predicted we would need 1.9 N of thrust per propeller for hover flight, so we looked for the motor-propeller combination that would be most efficient at 1.9 N. Once we found the most efficient combination, we had the tools needed to estimate our flight time, which is where we will start off today.
For this article we will be more precise with the mass of our components. We will assume the following mass breakdown of our 777 g drone:
Our goal is to maximize our drone’s flight time so that it can hover as long as possible. In our previous article, we modelled the flight time of our drone with varying battery capacity (figure 2).
Figure 2: Flight time vs. battery capacity for the original drone design
We presumed our design would include a Turnigy nano-tech 1300 mAh 4S battery and included its mass in our overall calculations. The battery’s capacity is just over 19.2 Wh (14.8 V * 1.3 Ah = 19.2 Wh), which occurs within the growth phase of the graph and gives us only about 4.5 minutes of flight time.
If we increased the battery capacity, we could also increase our flight time, but the trade off would be increased weight. This is where the design loop begins, as we swap components to try and build the drone that best meets our needs.
Up to the 0.2 hour mark there is an increase in flight time with increased battery capacity, but after about 100 - 125 Wh the marginal gains become less significant. For this reason, we will start by swapping our old battery with a new battery that has around 100 - 125 Wh of capacity in order to increase our flight time. The Turnigy 5000mAh 6S LiPo pack nicely fits our criteria with 111 Wh of capacity (figure 3).
Figure 3: Turnigy 5000 mAh/ 111 Wh LiPo battery (Photo: HobbyKing)
This new battery weighs a whopping 655 g compared to our old battery that weighed just 155 g. Assuming all of our other components stay the same at 622 g, the total mass of our drone is now 1,277 g.
We will therefore need to produce at least 12.5 N of thrust for the drone to hover (1.277 kg * 9.81), just over 3.1 N per propeller. We would also like to achieve at least double that thrust to have a good control authority, so we will be looking for the propeller that is most efficient at 3.1 N, but can also achieve up to 6.2 N of thrust.
To review, we have three propellers in our list of candidates:
We will work with the assumption that our drone frame is set and we cannot exceed 6” in diameter for our propellers. We can learn about our three propeller candidates by looking through the RCbenchmark database of electric motors, propellers and ESCs. Test data such as thrust, torque, RPM, power, efficiency and more is collected using one of our propulsion test stands, and for this drone the RCbenchmark Series 1585 would likely be the best fit.
For our candidates, data from the database tells us that all three propellers reach our hover thrust of 3.1 N, but only the 6040R King Kong nears the maximum thrust of 6.2 N (0.63 kgf) (figure 4).
Figure 4: Thrust performance of the propeller candidates
These results suggest that either our battery is too heavy or our motor/ propeller combination was not producing sufficient thrust. We are aiming to have the longest flight time possible, so rather than looking for a smaller battery right away, let’s explore some more propellers that fit within our frame limits, but produce more thrust.
Further reading: Automated Tests for Your Brushless Motors and Propellers
Our frame limits us to propellers that are 6” or less in diameter, but we can still experiment with our pitch, material, and brand. We will use the drone component database to filter for propellers that are 6” in diameter and produce at least 6.2 N (0.63 kgf) of force. This search provided several good options, but for simplicity we will narrow it down to three candidates that produce the most thrust:
Figure 5: Thrust vs. RPM for the new propeller candidates
As you can see in figure 5, all of our propeller candidates can produce 10 N (1 kgf) of thrust or more. For this reason, we can aim for a hover thrust of 5 N and a max thrust of 10 N, which will allow us to lift a larger battery with the same propulsion unit.
As shown in figure 6, at our original hover thrust of 3.1 N (0.32 kgf) and at our new hover thrust of 5.0 N (0.51 kgf) the efficiency of propeller 1 and propeller 2 is very similar, separated by only about 0.1 gf/W. Propeller 2 is slightly more efficient, but it is also heavier. This increased weight could lead to a shorter flight time and leaves less mass available for the battery. In a quadcopter, the total difference would be 3.76 g ((4.32 g - 3.38 g)*4).
Figure 6: Propeller efficiency vs. thrust for the new propeller candidates
After a quick look at the online marketplaces, it is evident that 4g makes no difference in terms of capacity for batteries of this size. For this reason, and the negligible effect of 4 g of mass for our drone, it makes sense to use propeller 2 due to its higher efficiency.
Our next step will be to find the brushless motor that is most efficient with this propeller at our new hover thrust of 5 N. In general, we are looking for a motor that can exceed our max thrust of 10 N, but not by too large a margin. We don’t want to operate the motor at its maximum speed for too long, but we also don’t want to haul a motor that produces more thrust and torque than we need.
Of the two motors we tested previously, MultiStar Elite 2306 2150 Kv and EMAX RSII 2207 2300 Kv, only the 2300 Kv motor meets our max thrust requirement (figure 7). We will therefore have to use the motor database to find a new candidate.
Figure 7: Motor efficiency vs. thrust for 2150 Kv and 2300 Kv motor candidates
From the database we find the Hypetrain Blaster 2207 2450 Kv, which meets our criteria. We ran a test with each of the two motors paired with propeller 2, and the results are shown in figure 8. Motor 2, EMAX RSII 2207 2300 Kv, is the most efficient with propeller 2 at our hover operating point of 5 N (0.51 kgf) and it also happens to be more efficient at our max thrust of 10 N (1.02 kgf). The efficiency difference at hover thrust is about 2.2% (55.6% vs. 53.4%), but the 2300 Kv motor is also lighter (32.37 g vs. 36.96 g), so it makes our decision easy.
Figure 8: Motor efficiency vs. thrust for 2300 Kv and 2450 Kv motor candidates
Now is a good time to summarize the mass of our components since the mass of our propellers and motors has changed as well as our hover thrust. Here is the new breakdown:
Based on these new values, we have 1392.7 g of mass available for our battery.
Since we also have our motor and propeller picked out, we can also determine our discharge (C rating) needs, which will also be a consideration for picking out the battery. We want to be sure that our motor will not draw more current than our battery can provide, or else the battery could rapidly degrade or overheat. The formula for determining current draw for a battery is: Current (A) = C rating * Capacity (Ah).
Further reading: Brushless Motor Power and Efficiency Analysis
There is no information on continuous or burst current for the EMAX RSII 2207 2300 Kv online, but we can look at data in the RCbenchmark database and compare all tests done with this motor. As we can see in figure 9, the max current reached during various tests was about 42 A.
Figure 9: Current vs. Rotation speed for EMAX RSII 2207 2300 Kv motor
The Turnigy High Capacity 16000 mAh 4S 12C Lipo Pack has the highest capacity in Wh of all the batteries in our weight range, giving us 4 * 3.7 * 16 = 236.8 Wh. It weighs 1,366 g, has a 12 C discharge rating and 16 Ah of capacity, so it can handle a current draw of 192 A, which is more than we need.
The main consideration for choosing an ESC is that it can deliver the motor’s peak current. In our case we do not expect our motor to exceed 42 A, so an ESC like the HobbyKing 60A ESC 4A SBEC will work great. It can deliver a constant current up to 60 A and a burst current up to 80 A, while providing 4 A to the BEC. This gives us a bit of a safety margin, so this ESC will be a good choice for our drone.
Figure 10: HobbyKing 60A ESC 4A SBEC (Photo: HobbyKing)
As we learned in our previous article, flight time is dependent on the capacity of the battery and the power drawn by the propulsion system. Many factors thus come in to play, summarized in the formula below (see previous article on increasing flight time for more details):
Where
E = capacity
σ = energy density
M = mass in grams (g)
We can copy+paste our propulsion test data into this handy flight time calculator, plug in our weight and battery capacity, and it will give us the best estimate of our flight time based on our data. Our estimated flight time is 15.2 minutes (figure 11), which is a significant improvement compared to our original design, which had only about 4.5 minutes of flight time.
Figure 11: Using the flight time calculator to estimate our drone’s flight time
As we have seen, the drone design process is cyclical and there’s almost always room to improve a design. Collecting propulsion data is one of the best ways to determine where there is room for improvement in your drone, and we offer many test stands and tools to help you do so:
If you enjoyed this article, you will also enjoy our free eBook: Drone Building and Optimization: How to How to Increase Your Flight Time, Payload and Overall Efficiency.
]]>By Lauren Nagel
Today's largest drones and eVTOL are capable of carrying thousands of pounds, in the form of humans, cargo and more. But how can we determine how much weight a drone can carry?
The amount of weight a drone can carry first and foremost depends on the amount of thrust produced by its propellers.
To say it in one sentence: the amount of weight a drone can carry is equal to the difference between the drone's total thrust and the thrust required for the drone to fly.
This number will differ based on the type of flight the drone is intended to complete, which we will get into below.
In this article we will cover:
Figure 1: A cargo-carrying drone
When a drone is hovering in place (in a no-wind condition), the thrust produced is equal to the weight of the drone:
So if the mass of the drone is 35 kg for example:
The drone’s propellers therefore need to produce 343 N or 35 kgf of thrust for the drone to hover. If we’re building a quadcopter with four propellers, that comes to 86 N or ~9 kg per propeller.
But it’s not that simple, because in most cases the drone will also need to take off and maneuver, which requires acceleration and greater thrust.
There is a basic rule that says that for most drones, if your propellers can generate twice as much thrust as is required to hover, you’ll have sufficient control for most operations. So if you’re building a drone for surveillance, videography or basic flying, this is a pretty simple rule to follow.
For our 35 kg drone, each propeller would therefore need to produce about 18 kgf of thrust for stable control.
Figure 2: IF1200 Heavy Lift Drone
Let’s say that we want our drone to carry a load in addition to its own weight. We just have to figure out how much extra mass we need to carry, then make sure that the propellers can produce enough thrust.
Let’s assume we want our 35 kg drone to carry a 5 kg load, so it will weigh 40 kg in total.
If we use the same formula as before:
We figure out that we need our quadcopter to produce 40 kgf of thrust to hover, or 80 kgf of thrust for stable control. Divided by four propellers, this is 20 kgf per propeller.
There are several methods we can use to confirm that our propellers can produce 20 kgf of thrust each, covered in this article on How to Calculate and Measure Propeller Thrust.
In conclusion, the amount of weight a drone can carry depends on a few factors, but most importantly the amount of thrust generated by its propellers.
There are several ways to estimate whether your propellers can produce enough thrust, but the only way to be sure is with proper testing.
If you’d like to learn more about drones, propulsion systems, and how to improve performance, we encourage you to check out our collection of articles.]]>When building a drone, one of the first steps is choosing the right motor. To get the best performance, it is essential to test multiple motors and select the most efficient one for your design. There are plenty of motors on the market, including many varieties of electric motors that are specifically built for use in drone design. In this article, we will focus solely on electric motors, touching on the following topics:
Note: We will be using the Series 1585 thrust stand to gather motor and propeller data for this article.
Before getting into the principle of the brushless mechanism, let’s first take a look at it’s components.
Electromagnets and Permanent Magnets:
A conductive wire coiled around a metallic base will not act like a magnet, but when a current flows through the wire, it will induce it to behave as a magnet would. This is commonly referred to as an electromagnet. If a negative current flows through that same wire, the magnet now has the opposite effect, it would attract another magnet instead of pushing it away.
The parts seen on the inner circle of figure 3 are the electromagnets, while on the outer circle we have the permanent magnets. To turn on the motor, you activate one of the electromagnets by delivering an electric current to its coils. This will make the rotor start to spin as the permanent magnet experiences repulsion from the like-electromagnet and tries to align with an opposite permanent magnet on the stator.
This will only make it spin for a short period of time as the electromagnet and opposite permanent magnet align. The next step is to power another electromagnet to keep the rotation from stopping, followed by the next electromagnet, and the next, and so on.
By delivering a three-phase current at a given frequency, the motor will spin at a speed equal to the frequency of that signal. The throttle on a drone’s controller is used for changing the motor’s speed, with a higher throttle input sending a higher frequency signal to the drone. The Electronic Speed Controller (ESC) controls the signal delivery to adjust the motor’s speed according to the throttle’s input.
An ESC or Electronic Speed Controller controls the electric motor by delivering electric signals that are translated into changes in speed or even a reversal of the rotation. It uses direct current coupled with a switch system to achieve an alternate three-phase current. This output current can then be modified by changing the rate at which the switches open and close in the circuit.
Brushless ESCs need information on the current position of the rotor to be able to start the motor and choose a direction for the rotation. To determine its position, the ESC uses information from the last unpowered electromagnet to measure its induction. This induction varies depending on where the closest permanent magnet is and the closer it is to the electromagnet, the stronger is the magnetic field induced.
The throttle is used to vary the speed of the motor. To do so, the ESC has to adjust the switching frequency based on the throttle’s signal. There are several signal delivery protocols with different performances, the most common ones being Oneshot, Multishot and Dshot.
The difference between them is the frequency of the signals delivered. Shorter frequencies allow a faster reaction time. Furthermore, the Dshot protocol is different from the two others because it sends a digital signal instead of an analog signal. This makes the signal more reliable since it is less sensitive to electrical noise and is more precise with its higher resolution.
Further reading: List of Brushless Motor Manufacturers
Every brushless motor is made of two main parts, a stator and a rotor. The stator is static, it does not move, and it holds the electromagnets. The rotor is the rotating component that holds the permanent magnets. There are two types of Brushless DC motors: Inrunner and outrunner models.
For inrunner motors, the rotor rotates 'inside' the stator, or further inwards relative to the motor casing. Outrunner motors have the opposite set-up as the rotor rotates 'outside' of the stator or further outwards, see figure 6. Both models have their pros and cons.
When comparing an inrunner and an outrunner of the same size, it is easy to see that the diameter on which the forces are applied is different. This happens because the electromagnets take a lot more space than the rotor carrying the permanent magnets. If the electromagnets are located inside, the diameter is bigger compared to if they are located on the outside (figure 6).
In addition to size, the consequence for the motor’s performances are also important to consider. A larger diameter means more torque because the force is applied further from the center of rotation, while a smaller diameter would be better for high RPM. Thus inrunners run best at high speed but generate less torque while outrunners work best with larger propellers because they can output more torque, but spin at slower speed.
Brushless Motors in the Drone Industry:
For the reasons described above, eVTOLs often use outrunners for vertical thrust due to their high torque. Inrunners are more commonly used for ducted fan jets or fixed wing aircraft and for horizontal movements requiring high RPM. This is easy to understand through the Kv value of a motor, also known as its RPM/Volt value.
This value determines at which speed the motor will spin if 1 V is sent through it. Therefore, an inrunner typically has a higher Kv value than an outrunner of the same size because, as stated above, the inrunner has a smaller rotor which spins faster for the same voltage. You will often see the Kv value displayed first when browsing brushless motors.
Further reading: Why You Should Test Your Drone's Motors and Propellers
When talking about the performance of a drone, we often think about the flight time. Unfortunately, the most advanced battery technologies can’t compete with gas motors in terms of the weight/energy ratio. The only solution is to carefully select the electrical components that will use the battery charge most efficiently.
Many factors can optimize the energy consumption of a battery. One of these factors is the motor’s efficiency. A higher efficiency helps increase the weight/energy ratio, thus getting higher performance from the motor and the drone itself. Testing a motor is crucial to be aware of the differences between a manufacturer’s specifications and the motor’s quality. To demonstrate this idea, we present a comparison of two similar BLDC motors with different price ranges.
To compare the motors' performance, we used the RCbenchmark Series 1580 thrust stand. It is capable of measuring thrust, torque, voltage, current, RPM and efficiency. The data acquisition is done using the RCbenchmark software. The two motors tested are T2407s of 1500 Kv and 2300 Kv respectively. Both motors spin the same 7-inch propeller with a pitch of 4.
The first graph in figure 7 displays the efficiency of both motors compared to the throttle. The second graph shows the efficiency at specific speeds. Clearly, the smaller 1500 Kv motor requires more throttle to run efficiently. It won’t achieve the same efficiency as the 2300 Kv motor in this set-up with the same propeller. The efficiency range is also better on the 2300 Kv motor, meaning the motor keeps a high efficiency at both low and top speed.
On the other hand, the smaller motor can operate at higher speed and would probably be more efficient with a smaller propeller. A drone used for transport would use a smaller motor to be able to stay idle and stable at low speed.
To get a full comparison, we would need to know the purpose of the drone and the RPM it will operate at. It is possible that the 1500 Kv motor is more efficient with a different propeller suited for a different function. Only through testing can we find out.
Finally, let’s talk about what to look for when choosing a motor. The most important characteristic is motor Kv, which describes the rotation speed you get for the power input to the motor. A motor with 2000 Kv will spin at 2000 rotations per minute for every volt sent to the motor.
Size is closely related to the Kv. A wide and large motor will often have higher torque but lower Kv and use a bigger propeller, while a thin but longer motor will have high Kv, low torque and would be best suited for a small propeller. This makes sense when you consider that a larger motor operates at lower RPM while a small motor operates at relatively high RPM.
The maximum power of a motor is also something to take into consideration. Exceeding the motor's power restriction would result in the motor heating up and thus drastically lowering its efficiency or even damaging it. The brand can also change a motor's performance. For the same Kv value, a motor from one brand may perform better than others. That’s why you should test your motor before building a drone around it.
Further reading: How to Calculate Brushless Motor Kv and Motor Poles
As a general rule, it is wise to use the smallest (lightest) motor possible without being in danger of overpowering and overheating it. The reason for this is to reduce the excess weight of a motor that is larger than needed. We have compiled a list of large motor manufacturers with products suitable for heavy-lift and manned drone operations.
The brushless DC motor is a genius invention that has been a game changer in the world of electric propulsion. Simple and streamlined in design, they allow vehicles like RC cars and drones to operate at high efficiency with maximum control.
If you’ve enjoyed this article you can check out our other content. We strive to create quality, educational material for drone designers and engineers. If you have any questions or ideas for other content you would like to see, leave us a comment below.
If you'd like to test your own motors, check out our test equipment:
Many UAVs are designed with dual motor or coaxial propulsion systems, where two rotors operate in the same axis of rotation, but in different directions.
There are several variables to consider when designing such systems, such as relative propeller size, speed, and inter-rotor distance. Making the right design decisions allows you to build the highest performing UAV possible - in terms of thrust, torque and efficiency.
We recently completed a study comparing various dual motor and coaxial rotor configurations.
We looked at how each variable would affect performance, including:
We believe that these results are pertinent to all unmanned aircraft designers, as they can inform better vehicle designs resulting in higher efficiency and performance.
This report summarizes our findings on physical interactions between rotors during coaxial tests, with a focus on the influence of the front rotor on the rear rotor. We quantified the effects of distance between rotors, rotation speed, and propeller diameter.
We used two Flight Stand 50 thrust stands for the tests, which measure up to 50 kgf of thrust and 30 Nm of torque, as well as RPM, voltage, current, mechanical power, electrical power, propeller efficiency, motor efficiency, and overall propulsion system efficiency.
The Flight Stand software controls the thrust stand and allows you to test up to 8 powertrains simultaneously.
During the coaxial tests, two Flight Stands were placed back to back on a rail system, with a brushless motor and propeller mounted on each of the stands’ force measurement units (FMUs).
The two propellers were set-up to rotate in opposite directions so that the torques of the motors would compensate one another and balance the structure. In a UAV, this coaxial set-up helps to avoid in-flight vertical axis rotation caused by torque.
The rear propeller was inverted to make sure that the airflow generated was in the same direction as the airflow generated by the front propeller (figure 1).
For the tests where distance between rotors was varied, the distance was measured between the two FMUs.
We first studied the effects of rotation speed and rotor separation distance on the thrust and torque produced by rotor 2 (R2) in the rear position. We anticipated that the performance of R2 would be affected by rotor 1 (R1) since it is located directly in the airflow generated by R1.
We completed the first set of tests with R1 rotating at two different speeds: 1600 and 2200 RPM. We also repeated the tests at four different separation distances: 10, 30, 50 and 70 mm. The results are shown in figure 2 and 3.
Figure 2: R2 thrust vs. speed at four coaxial separations and two R1 speeds
Figure 2 shows the evolution of R2 thrust as a function of R2 speed, at various R1 speeds and separation distances. As you can see, two groups of curves form, each representing a different R1 speed. The thrust of R2 was lower at all separation distances and R2 speeds when R1 was rotating at the faster 2200 RPM.
Figure 3: R2 torque vs. speed at four coaxial separations and two R1 speeds
Figure 3 shows the evolution of R2 torque as a function of R2 speed, at the same R1 speeds and separation distances as the previous test. Again, two groups of curves form, defined by R1 speed. As with thrust, the torque of R2 was lower when R1 was rotating at 2200 RPM.
For both thrust and torque, it does not appear that the distance between rotors has a significant effect on the thrust generated by R2. In general, we can conclude that an increase in the front rotor (R1) rotation speed reduces the amount of thrust and torque generated by the rear rotor (R2).
Further reading: Brushless Motor Power and Efficiency Analysis
In this section we investigate how R2 thrust and torque generation is affected by changing the size (diameter) of its propeller. Two rear propeller sizes were used: 40” (matching the size of the front propeller), and 47”.
Rotation speed was varied while the separation distance between FMUs was held constant at 20mm.
Figure 4: R2 thrust vs. speed with 40” (top) and 47" (bottom) rear propeller
Figure 4 shows the thrust generated by R2 as a function of R2 speed, at four different R1 speeds. The top graph shows results for a 40" rear propeller and the bottom graph a 47" rear propeller. As anticipated, the 47” propeller generated more thrust than the 40” propeller at all R1 and R2 speeds.
A decrease in R2 thrust was observed with both propeller sizes as R1 speed increased from 1600 to 2200 RPM. We observed a similar decline in torque (data not shown). The decline in thrust appeared to be greater for the 47” rear propeller, demonstrated by the greater spread between thrust data points at 1600 RPM and 2200 RPM.
We have two hypotheses for why this might be happening.
Hypothesis 1: the airflow generated by the front propeller applies a force on the rear propeller. This resistive load, which is oriented opposite to the force generated by the thrust of R1 (figure 5), is measured by the FMU. As a consequence, the resistive load is subtracted from the raw thrust value that is recorded.
Figure 5: The thrust and resistive forces experienced by R1 and R2
Hypothesis 2: R1 generates turbulence, and the induced drag counters the thrust and the torque generated by R2. We know this drag exists because of the windmilling effect observed when throttle is applied to R1 but not R2. The faster R1 rotates, the more drag is generated and a more significant loss of thrust will be observed.
When throttle is applied to the front rotor, causing it to rotate, the rear propeller will turn without any throttle. The torque is negative because what is measured is the resistive torque in the motor, which aims at braking the rotation of the propeller. It is induced by the drag generated by the airflow from R1.
Such a reaction can be seen when there is no electrical power input: R2 self-rotates and becomes a kind of wind-turbine (figure 6).
Figure 6: Aerodynamic forces causing windmilling in a rotor blade (source)
The 3D plot in figure 7 shows how R2 torque evolves as R1 and R2 speeds are varied. The blue and yellow sheets were interpolated from the data points shown.
The blue plot shows the evolution of the torque at a 10 mm separation distance between rotors whereas the yellow plot shows torque at a 25 mm distance. This chart demonstrates the presence of negative torque when little or no throttle is applied to R2.
Figure 7: Evolution of the R2 torque depending on the rotors speed - 10mm separation (blue) and 25mm (yellow)
Propeller efficiency is one of the main indicators of rotor performance as it measures the amount of thrust generated relative to the power input.
The propeller efficiency is calculated as the thrust generated divided by the mechanical power of the propeller (propeller rotation speed multiplied by the torque).
We studied how propeller efficiency evolved in relation to rotation speed and separation distance between rotors. In figure 8, R2 propeller efficiency is plotted against R2 speed. Each curve represents a different combination of R1 speed (1600 vs. 2200 RPM), and separation distance (10, 30, 50 or 70 mm).
Measurements were taken at R2 speeds 1200, 1400, 1600, 1800, 2000 and 2200 RPM, and the curves were interpolated based on these points.
Figure 8: R2 propeller efficiency as a function of speed and separation distance
As we can see, two groups of curves form, separated by R1 rotation speed. At all R2 speeds and separation distances, R2 has a higher propeller efficiency for the higher R1 speed of 2200 RPM.
For the curves where R1 rotates at 2200 RPM, there is a consistent decline in R2 propeller efficiency over the full range of R2 speeds. When R1 is at 1600 RPM, there is an increase and peak in R2 efficiency when R2 speed reaches ~1450 RPM, followed by a gradual decline parallel to the other set of curves for the higher R1 speed.
At high speeds, inter-rotor distance plays a negligible effect on R2 propeller efficiency. At lower speeds for both R1 and R2, the smaller separation distance is associated with higher efficiency. This difference is more prominent at low R1 and R2 speeds compared to when R1 speed is high and R2 speed is low. The effect becomes visible when R2 rotates at and below ~1750 RPM.
In conclusion, this study allowed us to observe the effects of several variables on the performance of coaxial rotor systems. We learned that the front rotor has a considerable impact on the rear rotor.
Our study provided several key takeaways:
Further reading: How to Test Brushless Motors
Please let us know in the comments what you would like us to study next!
]]>Understanding efficiency is the first step to improving the performance of devices and vehicles containing brushless DC (BLDC) motors. This article explains the key formulas for calculating brushless motor efficiency and demonstrates how they can be used in experimental situations.
This article will help you gain an understanding of how a brushless DC motor operates by providing examples with real data. If you are a drone designer, this knowledge will also help you develop strategies to maximize the flight time and payload of your vehicle.
In the experiments described, we used a Series 1585 thrust stand for data acquisition and powertrain control. The theoretical analysis is done in Octave.
There are two main efficiencies to consider in a brushless system: motor efficiency and propeller efficiency. In order to focus on motor efficiency, we will simplify propeller efficiency and say that the bigger the propeller, the higher its efficiency. Motors, on the other hand, have a high efficiency when spinning at high speed with relatively low torque.
Operating at high torque results in the motor heating up and losing efficiency. That said, the reduced efficiency might be a worthy compromise in order to use a larger, more efficient propeller and to avoid using a gear box, which adds complexity and leads to heat losses.
The theoretical model we use assumes that motor Kv does not change, though this is not exactly true due to the electronic speed controller (ESC). We also simplify heat losses as a simple resistance for use in the calculations, while in a real circuit there is a more complex impedance.
These assumptions and simplifications lead to differences between the theoretical and experimental results, which are interesting to observe through testing.
Further reading: How to Measure Motor Torque and Why You Should
There are several key formulas for understanding brushless motor performance. You can see how motor power and efficiency are closely related and dependent on factors such as torque, RPM, current and voltage.
The torque-speed plot helps us to understand how these two factors are related, as well as what happens when one of them is intentionally set to '0'. Let’s first look at formula 2, which relates the two parameters and describes how to calculate mechanical power:
Mechanical power is the product of torque and RPM, so when there is either no torque or no rotation, no power is produced. Figure 1 provides a visual reference for what this looks like in practice, where we present data from several motor tests.
Figure 1: Mechanical power for a 1500 Kv brushless DC motor
Along the horizontal axis you’ll see several blue circles, these data points were recorded during a no-load motor test. As you can see, when the motor is spinning freely with no load (i.e. no propeller), no torque is generated. An increase in voltage increases the motor’s speed along the horizontal axis, but the mechanical wattage remains at zero. The dynamometer used for the test confirmed the results, also indicating zero torque and zero mechanical power.
The other extreme is to have no rotation speed while slowly increasing the torque, achieved by sending power to the motor but not letting it spin. In theory, this mode of operation would produce a vertical line of data points along the vertical axis with increasing torque but no speed. In practice, motors used by UAVs and drones are not designed to work at very low speeds and would overheat with high torque. We did not want to subject our motors to this much stress, so we only performed this test at low torque, which is why there is only one data point above zero on the vertical axis.
In reality, your motor will operate between these two extremes. The six other power curves in figure 1 demonstrate the results of tests performed with unobstructed motors with propellers. Each dataset represents a different propeller, with the corresponding sizes listed in the legend on the right-hand side (diameter, pitch). As you can see, these propulsion systems occupy the middle of the graph between the two extremes, where regular operation occurs.
Electrical power can be determined experimentally or theoretically, and it can be a fun exercise to compare your test results to your theoretical calculations.
When calculating electrical power we can use formula 3, where electrical power is the product of current and voltage. We can also calculate electrical power using formula 4, where it is the sum of mechanical power (RPM*torque) and heat losses.
Figure 3 shows how electrical power relates to torque and RPM. The data points represent the test result from motors equipped with propellers of different sizes and pitch. Similarly to the 3D display of mechanical power in figure 1, the electrical power is proportional to the motor’s torque. However, even when there is no rotation (RPM = 0) and torque is applied, there is still production of electrical power. This is unlike mechanical power, where for all torque values at 0 RPM no power is produced.
Further reading: Drone Design Calculations and Assumptions
Each curve represents a different load (propeller) with speed and torque values plotted. The theoretical values are displayed as a 3D plot represented as a coloured surface, the brighter the shade, the higher the electrical power at that point.
Figure 3: Electrical power for a 1500 Kv brushless DC motor
To calculate a motor’s electrical power theoretically we use formula 4. To allow us to input values into the formula, we replace ‘mechanical power’ with the right side of formula 2, and ‘heat losses’ with the right side of formula 5, to give formula 6.
Here, R represents the resistance in the circuit generated by the motor (inversely proportional to the torque). The theoretical model assumes that the Kv does not change, which is not exactly true due to the electronic speed controller (ESC). We also simplify heat losses as a simple resistance, while in a real circuit there is a more complex impedance. These distinctions lead to differences between the theoretical and experimental results.
Using the test results we are able to make a 3D plot and fill in most of the variables in formula 7 (torque, RPM, current, and voltage), then we can use this information to find the R value closest to that observed in the experiments. The software Octave makes the 3D representation so the error between the estimated power usage and the real power usage is as small as possible, giving the R value.
Technically, Octave performs a linear regression with the cost function defined as the error of the estimated electrical power used. Formula 7 can also be rearranged to solve for R. Now that we have the R value for our motor, we plug in the values to formula 6 to calculate electrical power.
The efficiency of a motor is determined by dividing the mechanical power output by the electrical power input (formula 1). A goal for vehicle design is therefore to maximize this ratio and optimize the overall system efficiency by using the biggest propeller possible without overloading the motor.
This is why testing is so important. Using a giant propeller would drastically increase the propeller efficiency, but the motor would struggle so much to make it spin that its efficiency would be ridiculously low. Doing the opposite isn’t great either as making a tiny propeller spin very fast would lower both the motor and the propeller efficiency. The best solution is to balance both efficiencies to get the highest overall performance and system efficiency.
To achieve the highest efficiency the motor has to operate in the yellow zones of Figure 4 and 5 where there is a balance between torque and RPM. Testing various motors can help you figure out which size, Kv, and brand of motor will get you to that sweet spot for your operations.
Figure 4: Motor efficiency for a 1500 Kv brushless DC motor
Figure 5: Motor efficiency for a 1500 Kv brushless DC motor in 3D
The three graphs in figure 6 demonstrate how motor parameters vary and how to optimize them to achieve higher performance.
Figure 6: Parameters contributing to motor performance
The trend lines on these graphs can help you understand how to optimize your motor’s performance and operating conditions, whether you’re aiming for maximum flight time for your transport drone or maximum power for your racing drone.
In the left-most graph, the horizontal lines represent heat losses, which are dependent only on torque and current, not RPM. The lower torque and current your motor operates at, the lower the heat losses.
The graph in the middle shows the maximum power curve originating from the origin. If the motor does not operate along this line, it is operating at less than full power potential. A max power condition is ideal for winning a race, but not for increasing flight time or endurance.
Finally, the right-most graph shows the maximum efficiency curve. A motor operating on this line will use it’s battery more efficiently than anywhere else on the graph, thus increasing the flight time and endurance.
The location on the graph where you choose to operate depends greatly on the type of operations you are performing. The objectives of your flight should guide your design process and help you make decisions on which components to use.
When modifying the operating range of your motor, you will be looking at changing either your motor size, propeller size or both. In other words, you will be trying to optimize the load of the motor for its size. An easy way to tell if a motor isn’t efficient is to measure it’s temperature. A motor heating up too much is probably overloaded, AKA it is too small for its load. Furthermore, a motor not warming at all is probably too heavy or too big for its load, thus inefficient in that setup. Your needs will help you determine if you should start by modifying your motor or propeller size first.
When evaluating the performance of a motor, using both theoretical and experimental methods will give you the best understanding of your device. The difference between theoretical and experimental results comes from incorporating theoretical assumptions in your calculations, and also from imperfections in your measuring devices. For example, the ESC isn’t perfect and might sometimes activate at the wrong time. Bearing friction is also not taken into account and the measuring instrument’s precision comes with an expected error.
The theoretical model is a good tool to understand intuitively how a change will affect the motor performance. To confirm your assumptions and optimize your system, you need to perform actual measurements in a systematic way. Motor performance can be improved in several ways and the best way to determine them is to compare theory with test results to get a better understanding of your whole system.
We have several tools that can help you test your motor's performance:
The challenges explained throughout this article highlight the need for both approaches to achieve a well-rounded understanding of the components you’re working with and a respect for their limitations. Having several reiterations of a design is very normal as the drone building process operates as a loop: Assume → Build → Test → Repeat.
For a practical extension on how to apply these concepts to your drone design, we recommend reading another one of our articles on How to Increase a Drone's Flight Time and Lift Capacity.
If you have any questions about the contents of this article or the tools we used, don’t hesitate to leave us a comment and we will be sure to respond. If there are any concepts you would like us to cover that we haven’t already, just leave us a note below.
]]>Our Flight Stand thrust stands allow you to export data at any resolution from 1 to 1,000 Hz. This is a powerful feature that opens up a lot of testing and analysis possibilities.
But what does data sampled at 1,000 Hz look like? And what’s really the difference between 10 Hz, 100 Hz and 1,000 Hz?
In this article we will demonstrate the different levels of precision and analysis that are possible with data measured at each of these resolutions.
Table of Contents:
Dynamic testing is a test in which the values measured change relatively quickly. Acceleration tests, vibration measurement, torque ripple measurement are examples of dynamic tests. Mathematically, those are tests in which the value measured changes as a function of time, and an average value is not sufficient.
To achieve these results, dynamic testing requires a minimum sampling rate that is twice the frequency of the signal being measured. This is known as the Nyquist frequency, and it is the minimum sampling rate necessary to avoid aliasing and ensure that the signal being measured is accurately represented.
Static testing is a test in which the measured values are changing slowly. There are fewer requirements when doing static testing: the measurement synchronization is not as demanding, filters with higher time constants can be used, and the sampling rate can be lower.
Real-time testing describes a test where the timestamps associated with the measured values exactly match the timestamps at which the analog measurements are taken on the circuit board.
Filtering, especially low pass filtering, is a technique used to remove high frequency data from the signal. In a thrust measurement test, a low pass filter could remove the vibration, the noise in the current caused by the fast switching of the transistors, or the torque ripple. A high sampling rate is not sufficient to measure a dynamic test. The low pass filter applied on the data should have a cutoff frequency of at least twice the sampling rate so the filter does not remove important data.
To learn more about high speed testing, check out our article on Dynamic Testing in Data Measurement.
The amount of information you are able to obtain depends on the sampling rate of your data collection.
Here’s a few examples of what you can analyze at different sampling rates:
To demonstrate the differences in data at different resolutions, we performed a simple acceleration test (0 to 1436 µs) at 1,000 Hz, then resampled and exported the data at 10, 100, and 1,000 Hz using the Flight Stand software.
We plotted the measured variables on the graphs below, showing their behavior at three different resolutions. Note: the PWM and RPM values were adjusted to fit on the axes below while demonstrating the evolution of those two parameters.
As the sampling rate becomes higher, we can see a much more detailed picture of the evolution and variation of each variable.
We can look at the data from several angles to demonstrate the different levels of analysis possible at each resolution.
First, we will look at the torque and current fluctuations in this data, since torque is directly related to current.
To begin, at 10 Hz, we can see the major fluctuations in both of these variables. It appears that there are two major spikes in current, and one major spike in torque. The torque peaks around the 3.9 second mark at just about 4 Nm. The current also peaks at the 3.9 second mark to just under 6 A, then again at about 4.4 seconds, just crossing the line at 10 A. At this resolution, the data suggests that there may be a relationship between torque and current, but it is not obvious.
At 100 Hz, you can see that torque and current have a similar waveform and fluctuate at a similar frequency. The torque peaks at 3.68 seconds at about 5.5 Nm, and starts to resemble a torque spike. The current peaks at 4.46 seconds at just over 10 A.
At 1,000 Hz, we can really start to see the co-evolution of torque and thrust. The torque peaks at 3.670 seconds at just over 6 Nm and it is now evident that it is a torque spike. The current peaks at 4.410 seconds at about 10.7 A. At this resolution, we can determine which torque peaks are associated with peaks in the current, and which ones are not.
In this analysis we uncovered several torque peaks, including a torque spike at the beginning of the test when the motor’s acceleration began. This spike in torque occurs due to the extra effort required for the motor to start spinning from a stationary position. The spike in torque is very short and only visible with a high sampling rate.
The true max torque (>6 Nm) was much higher than what was measured during the 10 Hz test (~4 Nm), a 40% difference. This is important for designers to know, as the strain on the propulsion system may be much higher than anticipated.
The other torque peaks that are not associated with current peaks can be indicative of misalignments or inefficiencies in the propulsion system, and may warrant further investigation.
This type of analysis can be crucial for identifying weak points in your design.
Further Reading: How to Measure Torque and Why You Should
Vibration is an important variable when it comes to testing propulsion systems as it can affect the security of the vehicle's structure and the longevity of its components.
One way to quantify vibration, measured as the fluctuations in acceleration, is with a Fourier Transform (FT) analysis. A limitation of this technique is that the Nyquist limit requires the minimum sampling rate must be 2x the given frequency to be measured accurately. For this reason, it is important to start with data measured at high frequency.
For example, a 50 Hz data measurement will allow you to measure events that happen at a frequency of 25 Hz and lower. Same for 1,000 Hz, you can only measure events that happen at frequencies lower than 500 Hz.
The graph below shows the results of an DFT analysis performed on data from two different propulsion systems, called configuration 1 and configuration 2. A constant thrust test was run with both systems for 30 seconds, maintaining ~3180 RPM, equivalent to 53 Hz. The top graph shows the results for thrust and the bottom graph for torque.
Here’s what we can see in these results:
There are several other vibration peaks whose source is less easily identified.
They may be attributable to several other sources, which can be confirmed with further analysis:
Further Reading: Reducing Vibration in Propulsion Testing
The speed of the motor and ESC are not perfectly constant, and the fluctuations are visible in data measured at higher sampling rates. At 1,000 Hz we can look at their reaction time and dynamic response.
For the motor, thanks to the precise timing of the control, the thrust stand can characterize reaction time and the dynamic response of the motor.
An example is the negative torque that becomes visible as the sampling rate is increased. In the acceleration curves below, at 50 Hz we can see a negative torque peak of about 1.6 Nm become visible around the 3.64 seconds mark. At 1,000 Hz, we can see that it is in fact 2 peaks, the larger of which peaks at about 2.2 Nm.
This type of analysis can also be useful characterizing ESCs, which use transistors that pull a high amount of current quickly. You may decide to vary the wire length or capacitors used to change the response time and minimize current spikes. Data measured at 1,000 Hz can show you the difference in reaction time created by those changes.
Additionally, most ESCs use a frequency higher than 50 Hz, so a higher sampling rate is required to see variations in the current delivered and subsequent changes in the parameters like speed and thrust.
In summary, data collected at 1,000 Hz offers many possibilities for in-depth analysis of your motors, propellers, ESCs, and overall system.
In this article we have looked at how data collected at 1,000 Hz can help study torque and current fluctuations, system vibration, and reaction time. This information can help you gain a better, more precise understanding of your system’s performance in order to make improvements in your design.
For more information on how to improve drone performance, check out our eBook on Drone Building and Optimization.
If you’d like to start testing your motors, propellers, and ESCs at 1,000 Hz, check out our propulsion thrust stands:
These days it seems like drones are everywhere - delivering packages, inspecting buildings, filming movies, and much more.
As they become a more common sight, it can be useful to be able to identify the different types of drones you see.
This guide will help you differentiate between the different types of drones and UAVs.
In this article, we will explore:
Figure 1: Mavic Pro quadcopter drone in action (Photo by Pedro Henrique Santos)
A ‘Drone’ is defined by the Merriam Webster dictionary as “An unmanned aircraft or ship guided by remote control or onboard computers”.
It is used interchangeably with ‘Unmanned Aerial Vehicle’ (UAV), which shares the same definition.
Saying the aircraft is ‘unmanned’ strictly refers to the fact that the pilot is not on board, as passenger-carrying drones are already proving quite feasible. This becomes a bit more complicated when we start to discuss eVTOLs, but we will cover that in a later section.
An additional point to note is that ‘drone’ and ‘UAV’ generally refer only to the aircraft itself, whereas there are other terms that include all components that make the drone fly: UAS and RPAS.
A UAS, unmanned aerial system, is defined as “an aircraft and its associated elements which are operated with no pilot on board”.
The RPAS definition is more detailed and refers to “a remotely piloted aircraft system, its associated remote pilot station(s), the required command and control links and any other components as specified in the type design”.
The main difference between the two, according to the ICAO, is that in practice, UAS must receive special airspace accommodations and be kept away from other aircraft, whereas RPAS can be integrated into airspace alongside manned aircraft.
The technical difference between the two is that RPAS can only be piloted from a remote pilot station (RPS) whereas UAS may be piloted from an RPS or a ground control station (GCS).
Similar to manned aircraft, drones can be categorized based on the structure of their lift-producing surfaces.
Figure 2: eBee X fixed wing UAV
First off, as the name suggests, fixed wing drones have wings that do not move, they are bound to the body of the drone. They may have control surfaces that turn and rotate, such as ailerons and rudder, but the wings themselves are fixed. Figure 2 shows one such example, the eBee X by AgEagle, which has just one forward thrusting propeller located behind the fuselage and wings. In this system, lift is generated by the forward thrust of the aircraft coupled with the aerodynamic shape of the wing.
Pros of fixed wing drones are that their aerodynamic shape allows them to remain airborne for long periods of time, so they can cover large areas and are energy efficient. This is ideal for missions that require the drone to be airborne for hours or days, such as surveillance and climate monitoring.
Further reading: Comparing UAV Power Systems
The main cons are the steep learning curve for operators, the ample space required for take-off and landing, and the fact that they cannot hover in place.
For rotary-wing drones, the rotor-blades rotate around a central mast, forcing air downwards and creating the vertical lift required for the aircraft to become airborne. All vertical take-off and landing (VTOL) aircraft fall into this category, including small and large helicopters and multicopters. Rotary-wing drones may have a single rotor or even up to 16+ rotors generating thrust.
Pros of rotary-wing drones are that they are simpler to operate initially and can hover in place, allowing them to fulfill a wide variety of roles. Basic models are also relatively inexpensive, so small rotary drones are a great place to start for those wanting to get into the industry.
The main cons relate to their short flight time, as generating upward and forward thrust requires a lot of energy, thus restricting the range and endurance of the vehicle.
Powered-lift drones occupy a fascinating middle ground between fixed wing and rotary drones, using elements of both to complete a flight.
Rotors are used for VTOL-style take-offs and landings, then once airborne, the aircraft transitions to forward, fixed-wing-style propulsion. As such, they benefit from the low space requirements of rotary drones, while also having the improved aerodynamics and efficiency of fixed-wings.
There are several types of powered-lift drones, namely tiltrotors, tiltwings, and drones with two perpendicular sets of rotors.
For tiltrotor drones, the rotors themselves rotate in order to transition from vertical to horizontal thrust during flight. The Skywalker X 8 VTOL is a great example, with three rotors transitioning between copter and fixed wing mode.
Slightly different in design, tiltwing aircraft undergo a more dramatic in-flight transition as the rotors are fixed to the wings themselves, so the entire wing undergoes a rotation during the conversion from lift-off to forward flight.
Tiltrotors and tiltwings are also referred to as convertiplanes as they are converted from one configuration to another.
The third type describes drones that use perpendicular sets of rotors for lift-off and forward flight, neither set rotating or changing position. In this scenario, one set of rotors is operated only for take-off and landing while the perpendicular set is used only during portions of the flight requiring forward thrust.
While seeming to make the best of both worlds, powered-lift drones are quite complicated to design, which explains why they are not more common. Designing a system that can transition between configurations while remaining stable is challenging, though once achieved, the result is a highly versatile aircraft.
The majority of rotary drones on the market today are fueled by electric power from lithium-ion polymer (LiPo) batteries, though many new power sources are being explored.
LiPo batteries are rechargeable and lightweight, thus ideal from a design perspective. This is the most common power source for drones, which we cover in detail in our Guide to LiPo Batteries.
Solar power has also been employed as a source of electricity, with solar cells mounted on the upper surface of the drone. This concept could turn the industry on its head as it would effectively eliminate flight time limitations as long as the sun keeps shining.
That said, solar cells require significant surface area and have almost exclusively been used with fixed-wing models.
One fully solar-powered quadcopter was built by an engineering team at the University of Singapore. The design carried little more than the solar cells and rotors themselves, but it was still an impressive vehicle.
At the other end of the spectrum, gas-powered drones have become major players for carrying out long-distance missions.
While electric motors are more efficient than combustion motors, gasoline has about 50x the energy density of LiPo batteries, so maintains the energy advantage.
There are several possible fuel options for gas-powered drones, such as unleaded gasoline, two-stroke motor oil, nitro-based gasoline, and diesel.
Most gas-powered drones have a fixed wing configuration since they can more easily accommodate larger engines. Their aerodynamic shape produces so much natural lift that little gas is needed to achieve flight times of many hours or even days.
On a record breaking flight, Vanilla Unmanned’s diesel-powered VA001 drone not only stayed airborne for five full days, but when it landed, they found it had an extra three days worth of fuel on board.
Thus far, gas motors have been uncommon in rotary drones, as a heavier motor and fuel system is required.
Figure 5: VA001 diesel-powered drone (Photo: vanillaunmanned.com)
Furthermore, they can be noisy and require ignition of the motor to start. That said, some models are in development, such as the Goliath quadcopter, which uses a single gas-powered motor to power four propellers.
A third category of drones has emerged that boasts not one, but two power-sources. Hybrid-powered drones are changing the game and can come in different configurations, such as gas/battery-powered or battery/solar-powered.
The performance of these vehicles is unparalleled and they have been breaking all kinds of records with both rotary and fixed-wing aircraft.
Skyfront’s Perimeter 8 octocopter is a great example, using electronic fuel injection to convert gasoline into electricity in-flight. The novel system provides the aircraft with a record 5 hours of no-load hovering flight time, 10 times longer than the best battery-powered drones on the market.
Also to come is proof of concept for the PHASA-35, a high altitude long endurance (HALE) hybrid drone advertised to have a flight time of up to one year in the stratosphere. The drone operates on solar power during the day and battery power at night, with a 35 m wingspan to provide ample lift. The drone’s developers believe the drone could take over many of the current functions of satellites, such as delivery of telecommunications networks, weather monitoring, and surveillance.
Two other power sources to mention are hydrogen fuel cells and laser-charging of batteries in-flight.
Hydrogen fuel is a clean and energy-dense power source with a lot of potential in the drone, aviation and general transportation industries.
South Korean company Doosan claims their DS30W drone is the "world's first mass manufactured hydrogen fuel cell drone". It is truly unique as it has a rotory-wing octocopter design, yet it can achieve 2 hours of flight time.
In April of 2020, the DS30 drone delivered over 15,000 protective masks to residents on the remove Gapa, Mara, and Biyang Islands off the Southern cost of South Korea.Figure 6: DS30 drone by Doosan delivering face masks to remote South Korean islands (Photo: business wire)
Laser-charging of drones is a technology that has been on the scene since 2012. Recently, a team of researchers at Northwestern Polytechnical University in Xianyang, China demonstrated their laser-charging drone that can stay airborne indefinitely. They use an "intelligent visual tracking algorithm" to keep the laser beam targeted on the drone so it never loses power.
Now that we’ve covered the ways that drones can be classified, let’s look at some specific types of drones available.
Multi-rotor drones are perhaps the most recognizable form of UAV, complete with a compact body, multiple arms, and high RPM propellers.
They come in a variety of sizes and are named for the number of rotors they have: quad = 4 rotors, hexa = 6, octo = 8. Following the same rule, a drone with 16 rotors would be called a hexadecacopter.
These drones can take on a single or coaxial rotor configuration, which describes the layout of the rotors.
Single rotor configurations have just one rotor in a given vertical section whereas coaxial configurations have two, either completely or partially overlapping.
Many drones are piloted by visual line of sight (VLOS), where the pilot has the vehicle in view for the duration of the flight, and controls it based on observation. Other drones are piloted via first-person view or FPV, whereby the cameras on the drone deliver footage to a set of FPV goggles or a monitor so the pilot can fly the drone remotely from an onboard perspective.
Further reading: Drone Design Calculations and Assumptions
Though it was once a niche segment of the drone industry, demand for helicopter drones is growing, with the market projected to be worth $11 billion USD by 2027.
Operating on the same principle as manned helicopters, heli-drones use a single or coaxial rotor to generate vertical lift, partnered with a tail rotor to counter torque. Like multirotor drones, a perk of helicopters is that they can take-off and land virtually anywhere.
That said, drone helicopters are unique in that many are powered by gas, except for very small models that are powered by batteries. There are exceptions to this rule, of course, such as AeroVionment’s VAPOR 55 helicopter, which is all electric and manages a payload of up to 10 lbs.
A significant societal benefit of helicopter drones is their potential to replace humans in dangerous missions.
Manned helicopters play a big role in operations such as search and rescue, disaster relief, fire fighting support, etc., jobs that involve flying in challenging terrain with potentially poor flight conditions.
Unmanned helicopters could change this by delivering supplies to inaccessible areas, flying in smoky or windy conditions, and even evacuating humans from hazardous situations. All this could be accomplished without having to send a human crew into the field. Designed precisely for this purpose, Laflamme Aero’s LX300 helicopter can fly in severe weather and has a payload of up to 90 kg, enough to carry the average adult.
The electric VTOL aircraft (eVTOL) category encompasses all electrically powered VTOL vehicles, but for the purposes of this article, we will focus on a large and highly competitive segment of the industry: urban air mobility (UAM).
This new mode of transportation has the potential to reduce transportation-related emissions and cut down on commute times, acting as a sort of aerial Uber.
As such, there is an ongoing race to see who can design and manufacture a safe and efficient eVTOL vehicle for transporting people around urban areas, with companies such as Archer, Joby Aviation, Lilium, and Vertical Aerospace leading the pack.
Archer's Midnight eVTOL aircraft is designed to carry a pilot plus four passengers and is capable of flying up to 100 miles.
If all goes well, many major cities could have eVTOL taxi programs by the end of the decade.
Figure 9: Archer Midnight eVTOL (photo: Archer website)
Within the eVTOL category there are purely electric vehicles and also hybrid eVTOL vehicles, which most commonly use a combination of battery power and gas power. Hybrid aircraft are important players in the UAM movement as battery technology is still evolving to support extended flights independently.
The Bell Nexus 4EX for example is built to be able to run on battery power or hybrid gas-electric power, offering flexibility and greater flight time between charges.
That said, a major driver in the industry is the mandate to offer more sustainable transportation solutions, so many companies are still focusing their development efforts on going fully electric.
eVTOL vehicles can also be hybrid in nature in that they are designed for pilot-optional flight, meaning they can be operated manually onboard, piloted remotely, or pre-programmed for a flight. This versatility will come in handy as society gets used to the idea of inner-city travel via drone, appreciating a certain amount of redundancy.
HALE drones are aptly named for their purpose of flying high altitude, long endurance (HALE) missions.
To give you an idea of what this means, ‘high altitude’ drones are expected to fly above 60,000 ft and ‘long endurance’ flights are expected to be days, months or even years long.
Many HALE drones are intended for military use, designed for long-term surveillance, intelligence and reconnaissance operations. Others may be used as sub-orbital or pseudo-satellites, performing functions such as weather monitoring and telecommunications network delivery.
Figure 10: RQ-4 Global Hawk (Photo from the U.S. Air Force website)
The most widely used HALE aircraft to date is perhaps the Northrop Grumman RQ-4 Global Hawk, powered by a Rolls-Royce AE 3007 turbofan engine and most notably used by the U.S. Air Force (USAF).
Capable of flying up to 60,000 ft for over 32 hours, it has mainly been employed for the military purposes described above.
We hope that this article has provided you with a greater understanding of the diversity of drones in the industry. The field still faces many challenges such as obtaining government certifications, overcoming design limitations, and building trust with consumers.
With much development underway, it won’t be long before these aircraft become bigger and better, offering a more permanent service to society.
Further reading: The Future of Wind Tunnel Testing for Drones
]]>In the world of drones, UAVs and eVTOL aircraft, having an optimized motor-propeller configuration not only allows your aircraft to fly, but to perform optimally.
Manufacturers’ data can give you an idea which motors and propellers will work in your design, but testing is not standardized, so it is impossible to compare parts across manufacturers.
Here are a few ways that a thrust stand can help you improve drone performance:
A major reason for testing your motors and propellers is to increase your drone’s flight time. Increasing your vehicle’s air time will allow you to shoot longer videos, collect more data, maintain visual contact on a target, and fly farther on a single charge. Simple tests and modifications can add precious minutes to your flight time, giving you a competitive advantage over competitors. A great example comes from two mechanical engineering Masters students at the University of Ottawa who were able to more than double the flight time of their reconnaissance helicopter drone by testing various motor-propeller combinations.
Throughout the testing process, the students were able to determine that lighter electrical components would perform equally well and found a more efficient motor than the one they were previously using. The result was that they increased their helicopter’s flight time from 3 minutes to 7 minutes without compromising on noise or payload (and they came up with an idea for a great aerospace company). This example goes to show that flight time can be improved through basic modifications to your design that centre around increasing efficiency.
Further reading: How to Increase Your Drone's Flight Time
Many up-and-coming drone applications require vehicles to carry all kinds of payloads longer and farther than ever before. Meeting demands for payload capacity often requires testing multiple motors and propellers, but the initial investment will almost always pay off due to the improved operation of your UAV. Maximizing a drone’s payload capacity is important for industries such as eVTOL design, shipping and delivery, videography, cargo carrying, human transport, and more.
Hobby drones frequently have a payload up to 2kg, while drones in the “heavy lift” category may carry hundreds of kilograms of cargo. Whatever your payload requirements, testing multiple motor/ propeller configurations can ensure you’re getting the most bang for your buck. SkyDrive, a Tokyo-based aerospace company, makes use of such testing to optimize the geometry, size, and components of their drones. Most recently, they were able to build a heavy-lift drone capable of flying 15 minutes with a 30kg payload. Their final product, “CargoDrone”, contains 4 coaxial rotors for a total of 8 propellers and motors.
Improved antennas and range extenders have greatly enhanced our ability to fly UAVs into uncharted territory. The limiting factor is no longer how far we can communicate with the drone, but how long the vehicle can stay airborne on a single charge. Testing a drone’s propulsion system can help to extend its range by maximizing powertrain efficiency, contributing to a longer air time. This is especially important when flying missions into inaccessible terrain or over water.
If a drone cannot muster the power for the return trip, it could be lost completely. Ambitious videography projects and reconnaissance flights require a guarantee that the vehicle and footage collected will not be lost. With this guarantee, fantastic footage can be collected with confidence, such as this stunning tour of the Matterhorn in the Swiss Alps. Testing and optimizing your propulsion system can make these flights possible with the added benefit of knowing exactly what to expect from your UAV.
While many eagerly await the future of drones and eVTOL aircraft, one of the biggest societal concerns remains the increased noise levels and their impact on our environment. With the potential for drones to be flying overhead delivering packages, inspecting buildings and taking people to work, this is a fair and valid concern. For many drone applications, the amount of sound produced will be an important factor in deciding whether or not they are put to use. This is true not only for everyday applications, but especially for surveillance and reconnaissance operations that demand silence. Testing your drone’s propulsion system allows you to anticipate the noise levels produced and resolve any issues before it takes its first flight. This insight can ultimately lead to a more effective and competitive UAV solution.
An inspiring example of this technology at work is in wildlife monitoring and conservation efforts. Ocean Alliance’s “SnotBot” program utilizes modified consumer drones to collect organic samples from whales in order to better understand the health of the population. The drones used in these expeditions are exceptionally silent so as to not disturb or frighten the whales, a paramount requirement for the sustainability of their research. Motors and propellers produce the majority of the noise in a UAV, so testing and comparing motors and props is your best bet for conceiving the quietest version of your design.
All powertrains generate some degree of vibration, but excessive reverberation can cause damage to your components and is generally indicative of a lack of efficiency. Running a vibration test for your propulsion system is a great way to balance your propeller, detect inefficiencies and streamline your design. In doing so, you will likely notice that parts last longer and you get more performance out of a single battery charge.
Reducing vibration is especially important in the world of drone videography, where vibration can cause shaky or blurry videos, symptoms of the Jello effect. Stabilizers and post-production editing can improve the quality of your videos, but reducing the amount of vibration you are contending with in the first place can save time and money. Realizing these smoother flights and videos can easily be achieved with a bit of testing.
There is great incentive to increase reliability in the drone industry as the drone failure rate is about two orders of magnitude higher than that observed in commercial aviation. For UAVs to take over functions currently occupied by manned aircraft, their reliability, or mean time between failures (MTBF), must increase. Testing your propulsion system can help to prevent and predict failures as the data can provide insight into the state of your components.
Performing a Reliability, Availability, Maintainability, and Safety (RAMS) assessment, for example, is a great way to prove the reliability of our drone as it is an industry-recognized test that consumers recognize and trust. Once a system has been optimized, data from reliability tests can be a useful resource to reference or even publish as part of a marketing strategy.
Testing your motors and propellers can help understand how environmental factors affect your drone, such as the risk and impact of in-flight icing. In-flight icing or ‘atmospheric icing’ can be a major hindrance to drone operations, as ice build-up changes the aerodynamic properties of the aircraft. Ice accumulation results in increased weight and drag leading to loss of lift, thus inhibiting the flight capabilities of your drone. The NRC’s wind tunnel experiments also demonstrated how temperature and propeller type impact ice accretion, and that torque increases linearly with time exposed to icing conditions. Cold temperature resistance is key for drones operating in cold climates, high altitudes, and in challenging weather conditions.
Knowing how a drone responds to cold temperatures can help to design it accordingly and determine whether its operations require an additional investment in de-icing products. UBIQ Aerospace, a world leader in de-icing technology, studies the patterns and effects of ice accumulation on UAVs by testing the aerodynamics of propellers in wind tunnel experiments. Their tests have allowed them to refine their D•ICE technology, which mitigates ice accumulation for fixed-wing UAVs. For any cold weather drone operations, propeller testing can be an invaluable source of information for anticipating performance.
Further reading: What Conditions Cause Drone Icing
One of the best ways to save time and money on drone maintenance, and any vehicle’s maintenance for that matter, is to repair wear and tear before it becomes a problem. It is infinitely better to invest in replacement parts than to have to replace an entire drone due to a fatal failure.
Testing your motors and propellers is an important part of this protocol as damage is not always evident. Minor erosion from water damage or debris can affect motor efficiency, thus limiting the drone’s performance. Propellers can also become unbalanced over time and take on increased vibration, wearing down the entire propulsion system. Testing your powertrain as part of your preventative maintenance schedule can help detect these inefficiencies, resulting in improved drone performance and great savings in time and money.
In addition to the design phase, motor and propeller data can be useful throughout your vehicle’s lifetime. Recording diagnostics at scheduled intervals, every 50 flight hours, for example, can help to monitor a drone’s performance over time. Such tests are useful for detecting wear and tear as well as lost efficiencies.
Diagnostic motor and propeller testing can also provide insight as to why a UAV hasn’t been performing optimally, or give warning that it may not achieve the performance you are expecting due to unforeseen damage. Drones operating in environments with high humidity, high temperatures, or dust and debris may wear out faster than expected and surprise you with shorter-than-expected flight times or less-than-expected max throttle performance. Diagnostic testing of the propulsion system can detect these weaknesses before a failure occurs, preventing uncomfortable or potentially dangerous situations.
Safety is perhaps the number one concern of investors and regulators in the drone and electric vehicle industry. Without certain guarantees, aerial vehicles will not be permitted to enter the market and serve their purpose. Achieving a safer design can be accomplished relatively painlessly by testing the system’s components, especially the motors and propellers that complete thousands of revolutions per minute.
Better understanding these elements can help prevent overheating, engine failure, power loss and more, thus preventing accidents and injuries. As our skies become increasingly populated by aerial vehicles, citizens must feel confident that they will perform as prescribed, posing no threat to human activities. Safety tests can significantly increase the credibility of individual vehicles and the industry as a whole, guaranteeing performance and promoting peace of mind for investors, government officials and the public alike.
Special permission is required for many common drone operations in the form of a waiver or exemption from a regulative authority. In the USA, for example, a waiver is required for any operation not included in Part 107 of the FAA’s Small UAS Rules. The list of operations requiring a waiver includes flying at night, beyond visual line of sight (BVLOS), over people, more than 400ft above ground level (AGL), or more than 100 miles per hour. This list naturally overlaps with the activities of many hobby and commercial drone operations.
Obtaining a waiver is largely dependent upon the applicant’s ability to convince the FAA that their operation is safe, and extensive testing of a UAV’s propulsion system is an important part of the evaluation of a safe vehicle. Performing and demonstrating the replicability of propulsion tests can greatly support a waiver application while providing the designer with valuable information. While regulations vary worldwide, many designers will find propulsion data useful or mandatory for drone certifications all over the world.
Improving a drone’s efficiency is a circular process that begins with certain assumptions. These assumptions can include the total weight of the drone and the weight of individual components, as well as its intended use. Once these initial assumptions are made, propulsion testing can help determine whether your design will meet the requirements for its proposed purpose. For racing drones, you can determine if your design will meet speed requirements or for a delivery drone, whether it will achieve flight time minima.
The best part is that if your initial design doesn’t meet your needs, you can swap in new motors and propellers to find an ideal configuration. Once you have found your ideal powertrain set-up, you can also go on to try new batteries or modify your frame. The initial assumptions guide the optimization process, but once you have completed one round of review, you can make informed decisions about other modifications to your design.
Further reading: Drone Design Calculations and Assumptions
Propulsion testing can strengthen and legitimize the quality assurance offered to potential investors and clients. Backing up a marketing pitch with rigorous performance data is an excellent way to foster confidence in your product and give your sales team a competitive edge. In addition to sales, testing your motors and propellers can provide valuable information for data sheets and allow designers to breathe easy knowing the vehicles will perform as advertised.
Propulsion testing will also allow you to further standardize products, ensuring each unit performs equally. This translates to less time spent on customer support requests as the additional level of security will ensure a lower number of defective products leaving the facility. Whether your operation is large or small, testing motors and propellers is a great way to legitimize your quality guarantee and ensure consistency between products.
It is often necessary to learn how quickly a propulsion system can react to a change in control input. Data from such tests provides insight into how quickly the UAV can react to disturbances such as wind gusts. A typical way to physically test the reactivity of a propulsion system is to subject it to a frequency sweep control signal. A frequency sweep is a sinusoidal signal whose frequency is constantly varied to cover the whole spectrum of frequencies to be tested. With the data collected, the UAV designer can determine how fast the propulsion system can react to sudden changes in throttle.
Another method to learn about the reaction time of a powertrain is to subject it to a step input. While data is recorded at high speed, a sudden change in throttle is applied. After some time, the propeller stabilizes to a new rotation speed. The time it takes to stabilize the speed is known as the settling time. Finally, a proportional integral derivative (PID) test can measure your propulsion system’s consistency over an extended period of time by commanding a consistent, targeted thrust. These tests can be performed using a dynamometer with sufficient scripting capabilities, allowing for a complete understanding of a UAV’s throttle response.
Propulsion testing can contribute to simpler, more efficient flight replication for applications such as agriculture, surveying, research and more. The ability to record flights using propulsion testing equipment allows designers to plan and save a route automatically, then reproduce the throttle data later on. Without even leaving the lab, designers can optimize their flight plan based on data collected on motor and propeller performance. Agricultural applications of autonomous UAV technology are often based on the principle of efficiency, minimizing labour and input costs with the added benefit of protecting human health.
Airboard Agro’s agriculture drone is a great example, as it sprays crops on pre-programmed routes in challenging terrain, conducting the laborious and repetitive work previously performed by humans. To truly reap the benefits of this technology, drones should be optimized for the specific task they will carry out, taking into consideration the speed, thrust and stability required. Propulsion testing is a great resource for improving these flight replications and ensuring UAVs will attain their maximum efficiency for autonomous missions.
One of the most common causes of drone failure is engine overheating leading to engine failure. Maximum temperature and voltage ratings are often provided with electric motors, but it can be unclear when your motor is approaching these limits. Additionally, despite the fact that engine cooling depends on current, current ratings are not standard in the industry. One way to test a motor’s limits is to measure its temperature at various speed intervals using thermal probes, a useful strategy for circumventing failures.
At Kent State University in Ohio, Dr. Blake Stringer’s lab is performing such thermal tests with electric motors to investigate their thermal properties and to look at management of sUAS eVTOL motors under high-power conditions. This video produced in their lab shows the unfortunate consequences of thermal runaway, resulting in the overheating and destruction of the motor. These studies are increasingly important as drone operations become longer and move into harsher, hotter climates. Testing motors early in the design process can prevent overheating and ultimately save paying the cost to replace them.
Further reading: How to Automate Propulsion Tests for Your Drone
We hope that this article has demonstrated the multitude of benefits that come with testing your drone's propulsion system. We offer several tools that can perform these testing functions and help you take your design to the next level:
Our test stands measure thrust, torque, RPM, current, voltage, mechanical power, electrical power, motor efficiency, propeller efficiency and overall efficiency. Check out our online store to get yours or contact our sales team for a quote on our bigger tools.
]]>By Dominic Robillard and Lauren Nagel
A thrust stand is a great tool for testing your brushless motors and propellers.
It can save you time in many ways, and one such way is with automated testing.
Automated testing means hitting 'Start' and watching the test run independently as the data comes in.
In this article we cover 7 ways to test your brushless motor with our thrust stands, plus how to automated those tests.
Top 7 Brushless Motor Tests:
This article references two of our softwares that pair with our thrust stands:
The Series 1585 test stand was used to collect the data shown in the videos in this article.
Figure 1: Propulsion systems tested in a wind tunnel
A step test takes your propulsion system through a sequence of signal steps in a regular or irregular pattern. For example, you could set your throttle to start at 25% then increase at 5% intervals through 50% (25%, 30%, 35%, 40%, 45%, 50%).
At each of these ‘steps’ there is a defined settling time when the system holds the throttle steady to let the system stabilize before the data sample is captured.
There are a number of reasons to perform a step test as it is one of the simpler tests available.
First off, a step test is easily repeatable and can be run in exactly the same way as many times as needed. This makes it easy to compare different configurations of your propulsion system as the data collection points are the same each time.
It is also useful for looking at a system’s performance at various operating points. How does efficiency compare at 55% throttle vs. 60% throttle? What is the power consumption at 80% vs. 85% throttle?
A step test can help you answer these questions with a simple protocol.
Further reading: Why You Should Test Your Motors and Propellers
Designing a step test requires you to determine what operating points you want to learn about. Are you wanting to look at the whole range of signal inputs or are you focusing on a specific phase of flight?
Once you have determined this, you can input these points into a test script, like the one provided in the RCbenchmark software (figure 2). Set your minimum value, maximum value, the number of steps, and the settling time between steps.
Your steps may be manually input or imported from a spreadsheet.
Figure 2: Automated test script for a step test
During a sweep test (AKA a ramp test) data is continuously recorded while your propulsion system undergoes a smooth ramp from one throttle value to another.
Unlike a step test, there are no stops along the way, just continuous data collection during the transition between points.
For example, you could set your starting value at 40% throttle and your max value at 80% throttle, and have data collected continuously during the increase.
A sweep / ramp test allows you to test a full range of ESC operating points between two values. It is the equivalent of a step test with an infinite number of steps, so one advantage is that you don’t have to enter the steps manually.
Among other end goals, this can be useful for throttle curve analysis and observing signal aliasing effects.
One consideration is that the number of data points for the test is not fixed. It is still very possible to compare experiments, but data from different tests will not line up at precise values as they would for other tests such as a step test.
Additionally, the generated data files may become very large, making analysis more challenging.
In the RCbenchmark software there is a pre-written sweep test script that allows users to input their own custom values. The user enters a minimum starting value, a maximum peak value, plus the desired time to transition between the two, which controls the speed of the sweep.
One thing to keep in mind is that the sweep must be performed slowly enough to minimize the torque of acceleration of the propeller so that this does not affect the readings. If the ramp is sufficiently slow, this effect will be negligible.
The endurance testing category includes any tests that have a long duration and the end goal of testing the limits of the propulsion system or the components within it.
A few examples:
Further reading: Static vs. Dynamic Testing for UAV Components
Endurance testing is useful at many stages of the design process. It can help you select the best components and create accurate technical documents.
Endurance testing is also a big factor in safety and reliability. It can help with drone certification by demonstrating the useful life of components and the safe operation limits of the aircraft.
Endurance testing is also useful for setting up a drone maintenance schedule, as it can tell you when components will need a tune-up.
Any throttle pattern can be used for endurance testing, such as step, sweep, sinusoidal, flight replay, etc. The parameters for the test are very much based on your goals as a designer.
A few examples:
The RCbenchmark software can help you design your endurance tests using the pre-written script, “Custom steps sequence” (figure 3). The user creates a test sequence by defining the number of steps, the throttle value at each step, and the duration of each step.
The sequence can be repeated as many times as required. A data sample is recorded at the middle of every step in the test sequence.
Figure 3: Editable script for endurance testing
During a closed loop control test AKA a constant thrust test, the system constantly reads sensor data and adjusts the throttle to reach a constant thrust/power/RPM, as defined by the user.
An example of constant loop control is a PID controller, which senses an error between the target and measured value, then applies a correction based on proportional control with integral and derivative adjustments.
A closed loop control test is configured so that your desired variable stays constant, even as other areas of the system change.
For example, as a LiPo battery depletes and the voltage drops, the throttle will be adjusted to maintain the thrust/power/RPM that was programmed.
This is useful in many scenarios, such as when you are designing a flight, testing your battery or testing system endurance.
Running a closed loop control test with the RCbenchmark software requires a custom script, which you can create within the software using standard Javascript. Here is an example script of a closed loop test to get you started.
The script will have to be modified to achieve your specific goals. The main challenge is ensuring the system does not compensate too quickly or too much, which will cause the system to oscillate around your target value, so some tuning is needed.
A 90% settling time test measures how long it takes for a propulsion system to reach 90% of its final RPM value after a step input. This is a way of measuring the reaction time of a drone, for example.
It is a dynamic test that should be performed with a high sampling rate to ensure an accurate reading.
A 90% settling time test allows designers and control engineers to quantify the reaction time of a propulsion system as a whole - the ESC, motor and propeller together. This is useful for studying drone performance and resilience, for example how quickly it recovers from a disturbance such as a wind burst.
This test is easy to run as it is pre-programmed in the RCbenchmark software. The script delivers a step input signal to your propulsion system, causing it to go from 0 to 100% throttle. It simultaneously records the reaction time required to reach 90% of the final RPM.
Acceleration is also recorded to provide an estimate of the acceleration slope immediately following the step input. This data is included as an extra column in the output .CSV file.
Figure 4: 90% Settling time test script
A flight replay test is a way of recreating a past flight in the lab with your propulsion system. You can perform this test using the throttle data stored in your onboard flight computer from a previous flight.
This type of test is useful because it allows you to see the performance of your propulsion system at different stages of your flight. You can determine which stages draw the most power and which are the most efficient.
This testing method is one of the most accurate ways to reproduce flight operations, especially when combined with wind tunnel testing. It is especially useful for estimating battery life. Having access to this information can lead to changes in your drone design, your flight procedure or both.
Further reading: How to Calculate and Measure Propeller Thrust
The first step is to export the throttle data from a flight to your computer. You can then import the file into the Flight Stand software as a .CSV file to create a custom test script. If you use the RCbenchmark software, a custom script will need to be written.
Once you have connected your propulsion system to the thrust stand, the software will replay the throttle points through your system. You will be able to see how your thrust, torque, RPM, power and more change in real time.
During a sinusoidal test, the propulsion system is controlled with a smooth sine wave signal, forcing the motor to constantly change speed. The system goes through a whole spectrum of throttle points while continuously recording data.
During a chirp signal test, a smooth sine wave signal is sent to the propulsion system, but now the frequency progressively changes over time (figure 5). The signal frequency may increase or decrease, but the amplitude of the signal remains constant.
Figure 5: The structure of a chirp signal (Source: Wikiwand)
A sinusoidal test allows you to see the performance of your system at a whole spectrum of operating points. The continuous recording is similar to a sweep test and the pattern is useful for endurance testing.
A chirp test allows the user to test the complete range of speed changes of an ESC. This can tell you how quickly and to what extent the ESC responds to throttle changes.
It allows you to observe the system stability as it undergoes a higher frequency of commands, and it is often one of the tests performed by engineers to validate that a control system is stable at all frequencies.
For sinusoidal tests, we have a pre-written script for generating a sinusoidal function in the ESC output: https://cdn-docs.rcbenchmark.com/scripts/sinewave.js. This can be used with the RCbenchmark software.
Performing a chirp signal test requires you to write a custom script, and the sinusoidal script above serves as a good starting point.
Note that the update rate is limited by the ESC protocol and the USB communication rate. For this reason, we don't recommend setting a sine wave frequency above 10 Hz with the RCbenchmark software.
The automated brushless motor tests discussed in this article play an important role in optimizing propulsion systems. The data can help you identify practical areas for improvement in your system.
If you would like to perform these tests in your own lab, try our Series 1585 for small drones, our Flight Stand 15 for medium drones, our Flight Stand 50 for large drones, and our Flight Stand 150 for extra large drones and eVTOL.
As always, our powerful softwares are provided free of charge.
We are always looking for ways to improve and meet the needs of our customers. Are there any tests we missed? Let us know in the comments.
]]>By Lauren Nagel
A question we get fairly often is: “Why do you need to measure torque when doing motor or propeller testing?” This is often followed by: “How can I measure torque?”
These are both important questions for drone designers who want to get the most out of their designs. It ultimately comes down to measuring your motor’s efficiency by comparing the input to the motor with its output.
Note: Our small thrust stands and larger Flight Stands are both capable of measuring torque.
There are two key variables when it comes to the propeller: the first is the rotation speed and the second is torque. When you multiply rotation speed and torque together, you obtain mechanical power.
If we look at how a propeller and motor are connected, we see that the only connection or “information” sent from the motor to the propeller is RPM and torque.
Figure 1: Drone motor connected to three blade propeller
At the other end of the motor, electricity enters from the battery or power source. We can therefore consider the motor a machine that transforms electricity into RPM and torque or electrical power into mechanical power.
This brings us to our key efficiency formula. When we measure torque, we’re able to obtain mechanical power, which we can divide by the electrical power to obtain efficiency:
There are design trade-offs that come with increasing torque and RPM, and testing multiple propellers is the best way to find the most efficient motor-propeller combination for the type of flight you want to do.
Further reading: Brushless Motor Power and Efficiency Analysis
If we can keep the same propeller efficiency, increasing the ratio of mechanical power to electrical power means that air vehicles will be able to fly longer and carry more payload.
There are a few ways to measure torque, and in our test stands we use a steady-state solid system, which means that there are no moving parts. This is great because it reduces hysteresis and vibration.
In our Series 1780 test stand we have three load cells that each measure two forces, so there’s a total of six forces measured (figure 2). The image below shows our coaxial test stand, so there are three load cells for each propulsion system, or six load cells total. After our calibration procedure, we’re able to measure the exact torque and thrust that is applied to the motor mounting plate by the motor.
Further Reading: How to Calculate Motor Torque Using Formulas
Figure 2: The RCbenchmark Series 1780 test stand with three load cells
Now that we have measured torque, we also need to measure the motor’s RPM / rotation speed to fill in our equation for mechanical power.
This is achieved by using a small infrared RPM sensor that can sense when a piece of reflective tape passes in front of the sensor. The accompanying electronics use a counter to determine how many times the reflective tape passed the sensor, which allows it to calculate the rotation speed. The rotation speed can also be measured electrically, using the ESC signal. This method is simpler mechanically, but it is more sensitive to the motor load and size.
Multiply this figure by rotation speed and divide the product by electric power to get mechanical power.
Further reading: Drone Design Calculations and Assumptions
Measuring the system’s torque is essential when designing a propulsion system, as it allows you to measure the motor efficiency separately from the propeller efficiency. Here we covered how you can measure torque and how to use this information to build a more efficient drone.
If you are interested in testing motor torque yourself, check out our range of test equipment:
By Lauren Nagel
The electronic speed controller (ESC) is an essential part of an electric propulsion system’s hardware. It acts like the brain of the system by telling the motor how fast to go based on data signals it receives from the throttle controller.
For smaller applications like drones and RC vehicles, this controller has the name ‘ESC’, whereas for larger manufacturing applications it may be called an electronic control unit, inverter, or motor controller.
Figure 1: Afro Race Spec 20A ESC
The mechanism within the ESC as well as its interaction with the battery and motor are quite fascinating. In this article we will cover the fundamentals on how ESCs work, the protocols they use, and how they are used to control brushless motors and drones.
The role of the ESC is to act as the regulating middleman between the battery and the electric motor. It controls the rotation of the motor by delivering timed electric signals that are translated into changes in speed. It uses the direct current from the battery coupled with a switch system to achieve an alternating three-phase current that is sent to the motor.
The vehicle’s throttle controller is used to vary the speed of the motor, whether it be an electric car, plane or drone. Increasing the throttle increases the output power, which modifies the rate at which the switches open and close in the ESC’s circuit.
Figure 2: The controller communicates with the drone’s onboard throttle receiver
There are several signal delivery protocols that are used to convey throttle information from the remote controller to the ESC. Each protocol has a slightly different performance, the most common ones being PWM, Oneshot, Multishot and Dshot.
The most important difference between them is the frequency of the signals they deliver. Shorter frequencies allow a faster signal and a quicker drone reaction time. Furthermore, the Dshot protocol is different from the others because it sends a digital signal instead of an analog signal. This makes the signal more reliable since it is less sensitive to electrical noise and is more precise with its higher resolution. We will cover this in more detail later in this article.
Browse ESC test performance with our ESC Performance Database
Within the ESC there are a number of important components, including the microcontroller, gate driver and MOSFETs (figure 3), as well as the battery eliminator circuit and device manager adapter in some cases.
Figure 3: The key components of an ESC
The microcontroller plays three key roles in the ESC’s operation: 1) housing the firmware that interprets the signal from the controller and feeds it in a control loop, 2) keeping track of the motor’s position in order to ensure smooth acceleration, 3) sending pulses to the gate driver to achieve the desired command
The firmware used in ESCs is often pre-installed by the manufacturer, but open source versions can also be obtained from 3rd party sources. In hobby drones, the pre-installed firmware is generally a variation of BLHeli (either BLHeli_S or BLHeli_32), though other softwares like SimonK and KISS are also available. The chosen firmware must be compatible with the hardware as it will determine the ESCs performance and what protocols can be used.
The microcontroller also determines the motor’s position through a sensored or sensorless system. Sensored systems use electronic sensors in the motor to track the rotor’s position, which is great for low speed, high torque applications such as ground vehicles. The more popular sensorless systems use back EMF to determine the location of the rotor relative to the stator. This works great at high speeds, though when the motor is turning at lower speeds with less back EMF, the sensorless system does not work as well. This is generally not an issue when driving a propeller. Overall, for high speed applications, the sensorless system is more efficient, cheaper and more reliable.
The gate driver’s job is to act as the middleman between the controller and the gate of the MOSFETs. Upon receiving a low-voltage signal from the microcontroller, the gate driver amplifies the signal and delivers a high-voltage signal to the MOSFETs. The driver has lower resistance than the microcontroller so can deliver higher current, which also amplifies the speed of the signal. This allows for faster switching and lower heat production. Some ESCs have insulation optical chips between the low voltage microcontroller and the high voltage transistors. Manufacturers may call those ESCs Opto-ESCs.
The Metal Oxide Semiconductor Field Effect Transistors or MOSFETs are the switches that strategically deliver power to the motor. The ESC has six of these transistors and each wire from the motor is connected to two of them. The MOSFETs receive signals from the microcontroller then deliver power to the motor so that each of its coils is in one of three phases: high voltage, low voltage, or off/ grounded
As the motor rotates, the signals from the MOSFETs switch the phases of the coils so the rotor keeps spinning. The ESC uses direct current coupled with the switch system to achieve an alternate three-phase current (figure 4). The higher the throttle input, the faster the switching frequency, leading to a higher RPM in the motor. There are several signal delivery protocols that control this process, each with a different performance and signal frequency.
Figure 4: Opening and closing of switches in an ESC circuit
ESCs often have a built-in battery eliminator circuit (BEC), which doesn't eliminate the need for a battery, but acts as a voltage regulator to eliminate the need for a separate battery for on-board electronics. The power going through the BEC is dropped to a lower voltage, usually 5 V, which safely powers the throttle receiver and any other devices on board (figure 5).
Figure 5: Electric propulsion system wiring including an ESC and BEC
The device manager adapter (DMA) allows the user to connect their ESC to their computer to download firmware updates and use advanced programming options to customize their device. This keeps the ESC up to date and allows for control of advanced settings such as voltage cut-off, throttle calibration mode, and motor direction.
This component is generally brand-specific and is not available for all ESCs.
Figure 6: DMA from KDE Direct compatible with their UAS ESCs
Pulse width modulation (PWM) was the first ESC protocol and it is still used to this day. PWM uses timed power pulses to tell the motor how fast to turn, based on input from the throttle controller. The throttle controller sends a signal to the ESC’s microcontroller which tells it how much voltage to draw from the battery and deliver to the rotor.
The signal is delivered as pulses, whose width determines for how long voltage is drawn. Voltage pulses (‘on’) are separated by ‘off’ periods where no voltage is delivered. The greater the ratio of ‘on’ time to ‘off’ time, the more power is delivered and the faster the rotor will turn. The ratio of ‘on’ to ‘off’ time is also called the duty cycle.
In a PWM system, the length of the pulses varies from ~1000μs to ~2000μs. Originally, pulses were sent every 50ms, but this has increased over time so the signal is sent every 2.04 ms (490Hz). If the frequency were 500 Hz, the signal could potentially be 100% ‘on’, which would be detected as a fault.
The gate driver takes the voltage from the microcontroller and delivers it to the MOSFETs, where it drives them to switch between its three phases. The more voltage arriving at the MOSFETs, the faster they switch phases, and the faster the rotor turns.
If you plot time on the x-axis and voltage on the y-axis, you can see how ‘pulse width’ is controlled or modulated in this system (figure 7).
Figure 7: Pulse length in ms at min and max throttle
You can estimate your RPM by taking the average voltage over time (for both ‘on’ and ‘off’ signals) and multiplying that by your motor’s Kv rating.
The ESC protocol is essentially the language that the flight controllers use to communicate with the ESC. They use unique signal patterns as a way of conveying throttle information while varying the speed of the signal to vary the motor’s rotation speed.
Prior to 2015, PWM was the only ESC protocol commercially used by small UAVs. Since then, several new protocols have been created and it is common for hardware developed after 2017 to support all or most of them.
The most commonly used protocols include Oneshot125, Oneshot42, Multishot, and Dshot300, Dshot600 and Dshot1200. The Oneshot and Multishot protocols use analog signals like PWM, whereas Dshot (Digital shot) uses a digital signal.
Analog protocols require calibration to ensure that the oscillators (clocks) in the flight controller and ESC are synced, while digital protocols do not require this step. Without this calibration, your drone might not respond as expected due to the ESC misinterpreting the length of signals.
Dshot1200 is the fastest protocol, delivering 1,200,000 bits of data per second. Dshot1200 has a fixed signal length of just 13 μs, which is almost twice as fast as Multishot, the next fastest protocol, with a 25 μs signal length (figure 8). While Dshot1200 is impressively fast, some say the difference between Dshot600 and Dshot1200 is negligible in practice.
Figure 8: Signal length for common protocols in microseconds
A lower latency means a faster reaction from the vehicle, but there are diminishing returns, especially for larger airframes, due to the inertia of the quadcopter and the propellers. An aggressive flight controller also consumes significantly more power, so it is not advisable to use one if the vehicle does not require it.
Proshot is a unique protocol that contains elements of both digital and analog signals. This protocol encodes a DShot signal into PWM pulses - each pulse containing 4 bits of data. This encoding means that you can fit 16-bits of data into just 4 PWM pulses. Similar to DShot, ESC calibration is not required when using Proshot.
Proshot1000 delivers 1,000,000 bits of data per second, slightly less than the fastest DShot protocol. There is debate as to which protocol has higher CPU usage in practice, with no across-the-board answer yet.
Since 2018, there have been ESCs on the market that can support Proshot through BLHeli_32 firmware.
As we’ve learned, the role of the ESC is to deliver power from the battery to the motor in a controlled manner. If you input 50% throttle on the controller, the ESC will deliver 50% ‘power’ to the motor. What is ‘power’ depends on the firmware used. Some will use the average voltage sent to the motor, others use the speed target, and some use a mix of both.
On one end, the ESC has two wires to connect to the battery, a red (positive) wire and a black (negative) wire (figure 9). On the other end are three wires that connect the ESC to the brushless motor. If the motor spins in the wrong direction after connecting it to the ESC, switching any two of the wires will make it spin in the right direction. The final extension connects to the throttle receiver, which is powered by the BEC.
Figure 9: ESC with wiring for the battery (left), throttle receiver (middle) and motor connections (right)
Within the brushless motor are two components: the rotor (containing permanent magnets) and the stator (containing copper coils). When a current is delivered to a coil of the stator, it becomes an electromagnet and develops a North and South pole. When the polarity of the electromagnet matches that of the permanent magnet it faces, their like poles repel and the rotor spins. The current is delivered by the ESC as a three-phase signal that constantly changes the polarity of the electromagnets, that way the rotor keeps spinning.
In order to start this process, the ESC needs to know the position of the rotor to be able to choose which electromagnets to activate. To determine its position in sensored motors, the ESC uses Hall Effect sensors. This information is used to precisely synchronize the phase output with the angle of the rotor in order to ensure a smooth acceleration. In motors without sensors, more commonly used on UAVs, the start process is a bit less robust. The ESC will send a predetermined sequence to the motor to make it start. As soon as the motor has enough speed, the back electromagnetic force (back EMF) will be sufficient for the ESC to obtain a precise position estimate and synchronize the pulses.
For more information, check out our article on How Brushless Motors Work
Choosing an ESC is an important part of the drone design process. You want to ensure that it meets the electrical needs of your aircraft without draining your battery more than necessary. Below are a few factors to consider when choosing an ESC.
The ESC's current rating should be 10 - 20% higher than the motor’s. This will prevent it from overheating and provide a bit of wiggle room when operating at max throttle. You do not want to go much higher than this range to minimize weight. The ESC should be tested in conditions similar to flight as the main limitation is thermal. High temperature and low air circulation will reduce the ESC rating and operating life. Some ESCs have two current ratings: continuous and burst. The continuous current is sustainable for prolonged periods of time and the burst current for short periods only.
ESCs have a maximum voltage limit that may be given as a voltage range or a cell range. For example, an ESC rated for 3S - 8S cells will support a voltage of 11.1 - 33.6 V. The ESC may let you set a switch-off voltage that will alert you when the battery voltage becomes too low (3.0 - 3.4 V per cell) to avoid damaging the battery. Those systems are called Low Voltage Cut Off (LVC) and they will reduce the maximum power that the ESC can provide. Eventually, the ESC will shut down the motor.
When wiring ESCs into a quadcopter you can have one ESC for each motor or use a 4-in-1 ESC with a single board and four motor connectors (figure 10). Having four ECSs can help spread the heat load if the motors have a high power draw while a 4-in-1 ESC is a great option for saving space and limiting weight from hardware.
Figure 10: 4-in-1 ESC from iFlight
In this article we’ve covered the ESC basics: how they work, the key components, the protocols, and how they work with brushless motors and drones. Having a good understanding of this essential drone component can help you improve your knowledge and your build.
For more information on optimizing your drone, check out our free eBook on Drone Building and Optimization.
If you have any questions, don’t hesitate to leave us a comment below.
]]>By Lauren Nagel
Wind tunnels are a valuable piece of equipment for characterizing object aerodynamics. There are several types of wind tunnels that work in different ways.
In this article we will cover:
Wind tunnels are an excellent tool for producing wind flows in a controlled setting to replicate flying conditions.
By using one or more fans to force air over an object, one can visualize the interaction between the object and the surrounding airflow in order to predict its aerodynamics.
Figure 1: Installing a model aircraft in the test area of a wind tunnel
In traditional wind tunnel testing, wind is generated by a fan and passed through a test area where the object of interest is installed. The tunnels themselves vary in size and shape.
Most wind tunnels have either an open or closed style return. For some supersonic testing, blowdown style tunnels may also be used, which rely on a pressure difference between a high pressure basin upstream of the test area and a low pressure reservoir downstream.
Wind tunnel tests may use a combination of air pressure sensors, force balances, and physical indicators like smoke, oil and paint to characterize how an object interacts with a wind flow.
Advanced methods include pressure sensitive paint, which changes colour with variations in pressure, and particle image velocimetry, which uses a laser sheet to track the velocity of particles passing through a plane in the test area.
Figure 2: Testing a full-size U.S. Navy aircraft in a wind tunnel
There are a number of ways to identify the different types of wind tunnels, such as speed, configuration, style, and output.
Traditional wind tunnels are classified by the speed of the air passing through the test section relative to the speed of sound (Mach 1). They are divided into four categories: subsonic (Mach <0.8), transonic (Mach 0.8 – 1.2), supersonic (Mach 1.2 – 5.0), and hypersonic (Mach >5.0).
The category limits are based on the magnitude of compressibility effects that must be considered for flights at that speed.
As objects move faster and faster, they compress the surrounding gases, thus changing the fluid density and altering the amount of force on the aircraft.
This effect does not occur at subsonic speeds, but has unique effects at transonic, supersonic and hypersonic speeds, hence the categorization.
Wind tunnels come in many configurations, but open and closed return tunnels are most common, differing in their shape and how the air is circulated.
Open wind tunnels have open ends on both sides of the test section and gather air from the environment where the tunnel is located.
Once the air passes through the test section and out one end of the tunnel, it is recirculated through the room to the tunnel entrance where it can re-enter the tunnel (figure 3).
An open wind tunnel with an open test section is also known as an “Eiffel tunnel”, named after Gustave Eiffel who used his Parisien wind tunnel to study the properties of lift on an airfoil.
An open tunnel with a closed test section is also known as an “NPL tunnel”, named after the National Physical Laboratory where they were first used.
Figure 3: Open return wind tunnel (grc.nasa.gov)
Closed wind tunnels use a self-contained circuit that recirculates air within the tunnel through the test section (figure 4).
The tunnel is not open to the test environment surrounding the apparatus, so the same air is circulated over and over again. Turning vanes help to create smoother changes in direction for the airflow throughout the tunnel.
Closed wind tunnels are also known as “Prandtl tunnels” after Ludwig Prandtl who developed the first theories on supersonic shockwaves. They also go by “Gottingen tunnels”, named after the German research laboratory where they were first used.
Figure 4: Closed return wind tunnel (grc.nasa.gov)
The benefits of traditional wind tunnels like these include their ability to produce high speed flows and their ability to restrict the test area.
That said, restricting the test area also makes it more challenging to test an aircraft in motion since space is limited. Models are often used in traditional wind tunnel testing since tunnels able to test full size aircraft are extremely costly to build.
The world’s largest wind tunnel, measuring 80' x 120’ and based at the NASA Ames Research Center in California, was constructed through the modification of a smaller 40' x 80’ wind tunnel. The project budget in 1979 was $85 million USD, equivalent to about $350 million USD today.
Open air wind tunnels, or Windshapers, are a promising alternative to traditional closed wind tunnels.
This option is especially relevant for free flight testing of aircraft and drones as it allows for tests without the boundaries of a physical tunnel. Such open air facilities can accommodate the same size aircraft for a fraction of the cost of a traditional wind tunnel.
Open air wind tunnels differ from traditional wind tunnels in that the test area is open to the environment and not enclosed within a tunnel (figure 7).
A wall of fans produces a 3D wind flow based on a three variable function (x, y, z). The function communicates information to the individual fans, which can produce all kinds of wind profiles, such as waves, turbulence, wind bursts, etc.
The wind flow is directed through an open test area where the aircraft is located. This allows the aircraft to fly freely without needing to be mounted in place.
Further reading: Why You Should Test Your UAV's Motors and Propellers
Figure 7: Drone testing with an open air wind tunnel (Windshaper)
Air flow visualization with an open air system is made possible through motion tracking and flow probe technologies used simultaneously.
Flow visualization software combines data from the two sources, then processes, interpolates and visualizes the data (figure 8).
The data collected can be visualized in real time or analyzed later using standard CFD visualization tools.
Figure 8: Drone airflow visualization with ProCap technology by Streamwise
The diversity of wind profiles produced with one or multiple walls of fans is great for simulating a flight through difficult weather or around buildings and terrain.
This video by the government of Canada provides a great demonstration of the challenges of urban flying with a drone, for example, and the kind of skills inner-city pilots must develop.
Because of the open of the test area, open air wind tunnels also allow you to test take-off and landing configurations.
Windshapers have a tilting function that allows them to rotate up to 90 degrees. This allows the user to practice the most critical phases of flight while simulating ground effect, thermal updrafts, crosswinds, etc.
In addition, climatic effects can be added to the test area, such as rain, snow or fog (figure 9).
Figure 9: Drone testing in rainy conditions
Open air wind tunnels are relatively inexpensive compared to traditional wind tunnels with the same cross-sectional test area. They therefore provide more feasible free flight testing.
So far no open air wind tunnels have been constructed that reach transonic speeds, which is one downside of the model.
Computer-based wind tunnel simulation, or computational fluid dynamics (CFD), can be used independently or hand in hand with physical testing to understand an object’s aerodynamics.
In budget-limited projects, simulation may be the only method used to evaluate an object, whereas in larger projects it may inform and compliment physical wind tunnel testing.
Figure 5: Visualization of airflow over a wing flap by AirShaper
Computer-based programs can start with as little input as a 3D model or 3D scan of an aircraft. The model is uploaded into the testing software and subjected to different simulated wind conditions.
The user can control the wind speed, angle of attack, aircraft movement, and the properties of the fluid.
The results of the simulation provide key information such as wind shape, surface pressure, and wind flow separation along portions of the aircraft (figure 6).
For newcomers to CFD, NASA offers several free wind tunnel software designed to teach the basics of aerodynamics and propulsion. For more advanced commercial uses, companies like AirShaper and SimScale offer solutions that allow clients to produce detailed models and comprehensive wind profiles.
Figure 6: Aircraft landing gear flow visualization with SimScale CFD technology
Overall, CFD is a cost effective way to collect theoretical aerodynamic data, especially as the technology becomes more advanced.
At the end of the day, however, the data is just that, theoretical. For certain applications it may be sufficient to go from simulation to field testing, but for many projects, physical testing in a wind tunnel provides greater peace of mind.
Wind tunnels can help aircraft builders improve their design by identifying areas where aerodynamics can be improved.
This allows designers to improve their lift: drag ratio, which can translate into improved flight times and improved resistance to turbulence.
More advanced wind tunnels can also simulate turbulence and sideways wind flows to study performance in diverse flight situations.
The size of the wind tunnel required depends on the size of the aircraft and the type of testing performed. Some wind tunnels are large enough to test full size aircraft while others employ models to make assumptions about the full size versions.
A larger tunnel will be required for free flight testing compared to testing a mounted aircraft model, for example.
Here is a comparison of three types of wind tunnels:
For more information on traditional wind tunnel testing and CFD, check out the companies and links mentioned in the text.
For more information on open air wind tunnel testing (Windshapers), check out our wind tunnel testing landing page.
Further reading: Wind Tunnel Testing at the University of Ottawa
]]>However, not all motors are created equal when it comes to efficiency. In this article, we will explore the differences between different types of motors and compare their efficiency.
Table of Contents:
First, let's define what we mean by motor efficiency. Motor efficiency is the ratio of output power to input power, expressed as a percentage.
It measures how effectively a motor converts electrical energy into mechanical energy. The higher the efficiency, the less energy is lost as heat or other forms of waste, and the more power is available for useful work.
Three common types of motors are DC motors, AC motors, and BLDC motors. Each of these motors has its own advantages and disadvantages when it comes to efficiency.
Further reading: Finding the Average Motor Efficiency with a Database
DC motors are the simplest type of electric motor, and they have been used for many years in a variety of applications. They operate by using a direct current to create a magnetic field that rotates the motor's armature.
DC motors are known for their high starting torque and controllability. However, they are also known for their low efficiency, typically ranging from 50-80%.
This is due to the energy lost as heat in the motor's windings and brushes.
AC motors operate using the same electromagnetic principle as DC motors, but they operate by using an alternating current to create a rotating magnetic field that drives the motor's rotor. This type of motor is more commonly used in industrial and commercial applications.
AC motors can be further subdivided into two categories: synchronous motors and induction motors (which can be further divided into single-phase induction motors and three-phase induction motors).
AC motors are known for their high efficiency, typically ranging from 75-90%. This is because they don't have brushes, which eliminates the energy loss associated with them.
However, AC motors are less controllable than DC motors, and their starting torque is usually lower.
BLDC motors are a newer type of motor that combine the best features of both DC and AC motors.
They operate by using a permanent magnet rotor and an electronic controller to switch the current in the motor's windings.
Are brushless motors more efficient? Compared to traditional DC and AC motors, the answer is generally yes.
BLDC motors are known for their high efficiency, typically ranging from 80-95%. This is because they don't have brushes and use electronic switching to control the current, which eliminates energy loss.
BLDC motors also offer high controllability and starting torque.
Further reading: Brushless Motor Power and Efficiency Analysis
In conclusion, motor efficiency varies depending on the type of motor and its application.
When comparing the efficiency of different types of motors, it's important to also consider the specific application and requirements.
DC motors have lower efficiency but high controllability and starting torque. AC motors have higher efficiency but lower controllability and starting torque. BLDC motors offer high efficiency, controllability, and starting torque, but are typically more expensive.
When choosing a motor, it's important to consider the trade-offs between efficiency, controllability, and cost to find the best option for your specific needs.
To learn more about how motors work, check out our library of articles.
]]>By Lauren Nagel and Charles Blouin
Electric motors are used in many tools and vehicles. In drones and electric aircraft, they are most often controlled with a throttle controller and ESC.
When you increase the throttle on your controller, what you are doing is increasing the amount of electrical power drawn from your power supply (i.e. battery) and fed to your motor via the ESC. If you want to slow down the motor, reduce power; If you want to speed up the motor, crank up the power.
But what is the mechanism behind controlling the motor’s speed?
In this article we will cover:
Figure 1: Xoar T110 motor mounted on thrust stand
Controlling electric motor speed is as “simple” as controlling the amount of electrical power delivered to the motor. In the following article, we are going to assume that we are looking at a steady state condition where there is no acceleration.
To be more precise, for a given mechanical load (or resistance), it is the increase in voltage that leads to the increase in speed. Alternatively, we can reduce the load on the motor.
If we look at this equation for angular velocity:
We can see that the angular velocity, (AKA motor speed, AKA RPM), is proportional to voltage and negatively proportional to torque.
Therefore, if we want to increase the motor’s speed we can either a) increase the voltage delivered or b) decrease the torque. Since decreasing the torque would require changing your design or load, it is much simpler to increase the voltage.
The equation also demonstrates that at a constant voltage, a higher torque results in a lower RPM.
We should also note that there are some minor losses within the ESC, so the input voltage is higher than the output voltage to the motor, but usually only by a few percent.
We can also describe this process of changing motor speed in terms of the ESC protocols used to deliver the power.
With PWM, for example, the value sent from the controller (in µs), corresponds to a percentage of the maximum voltage that can be sent to the motor and the ESC.
The PWM value in µs will be between 1000 (no throttle) and 2000 (full throttle). All of the values in between have a signal length in µs, which correspond to a duty cycle between 0 and 100%.
At 0%, there is no power delivered to the ESC and motor. At 100%, power is constantly delivered to the ESC and motor.
Figure 2: Xoar TA130 motor mounted on test stand
There is variation in this implementation by ESC manufacturers. However, the relationship between duty cycle and motor speed is usually roughly equivalent to the average output voltage of the ESC.
The ESC approximates a sine wave output on each phase. At 50% ESC output, the sine wave will have about half the amplitude of the sine wave at 100% ESC output. Note that the electronics drivers on most ESCs only have an on-off state, so the actual signal is not as smooth as a perfect sine wave.
There are many variations between motor and ESC manufacturers, and the electronics can introduce non-linearities. For practical applications, it is best to characterize the motor and ESC using a dynamometer.
Click here to learn more about the Internal Workings of an ESC.
Digital protocols like Dshot and Oneshot use the same duty cycle concept, but with much faster signal delivery and different signal delivery patterns.
You may also see the duty cycle written as a value between 0 and 1024. This is based on the fact that the technology was developed around 8-bit controllers. 2 options (on / off) with an 8 bit system → 2¹⁰ → 1024.
This system can also be used to determine the ESC’s output voltage.
For example, 796 → 78% ‘on’ relative to a 100% output.
In a no load condition, you can use the motor’s KV value to estimate the motor’s speed based on the voltage delivered.
As discussed in our article on how to calculate motor KV, we can replace Back EMF with input voltage in this equation to estimate the rotation speed of our electric motor.
Let’s use MAD Components’ M50C35 EEE 9 KV motor as an example. It has a max voltage of 400 V and a KV rating of 9 KV. Plugging these numbers into the formula we get a max speed of:
So at max throttle, this motor will rotate at 3600 RPM. If we applied only 75% throttle, corresponding to 300 V, it would rotate at a speed of 2700 RPM.
As demonstrated in this article, there are a few ways to increase motor speed.
The most actionable way is to increase voltage delivered to the motor by increasing the throttle. You can also decrease the torque, but this is generally more disruptive to the design or purpose of the UAV.
If you have any more questions about how electric motors and UAVs work, please check out our other articles or leave us a comment below.]]>In the six years since we started selling our drone test stand technology, we’ve learned a lot of lessons through trial and error. One of those lessons led us to develop our (patent pending) technology, the solid state load cell system for measuring thrust and torque.
If you are familiar with our products, then you know that our test stands combine thrust and torque measurements with voltage and current to derive the system’s power and efficiency.
Perhaps the most obvious concept for such a tool, and the cheapest to build, is a system involving bearings or bushings. (We would know, since this is what we used in our original Series 1585 design).
Figure 1: An early version of the Series 1580 thrust stand
What we’ve learned, however, is that using bearings and bushings has a number of disadvantages:
In addition to our early tools, we noticed that many hobby thrust stands and homebuilt designs used bearings and had the same issues. Once we figured out what was causing our issues, we released an upgraded version of the Series 1585 in 2018 with semi-rigid hinges to replace the regular hinges. This allowed us to reduce our measurement error rate to less than 1%.
Figure 2: The modern Series 1585 test stand
There are disadvantages to using a solid state system, such as:
The only way to address these issues was with simulations and many experimental tests. The tests we performed included calibration and verification with hundreds of points, endurance testing, and hundreds of hours of motor testing.
Overall, given our experience, we strongly recommend that any thrust and torque measurement tool use a solid state design without pivots or bearings. In a professional environment, losing months of tests due to faulty data can be extremely expensive. The tests have to be redone and the product may be delayed. There is also a recall risk if the faulty data is used in production.
Although they are of a simpler design, we have completely abandoned traditional pivots and bearings in our thrust stand designs due to the issues mentioned above.
If you’d like to learn more, contact us or check out our report on the measured differences between a solid-state system and one that uses bearings (view the report here):]]>By Lauren Nagel
Torque is an important variable when considering electric motor performance. It can be used to calculate mechanical power and derive electrical power. For drones and electric aircraft, knowing your torque also allows you to calculate motor efficiency separately from the efficiency of the rest of the system.
In this article we will:
Let’s say we want to know the torque of our motor, the Xoar 2407. Here’s what we know:
Nothing here is immediately helpful for calculating torque, so let’s look at some background information.
Motor torque is related to several variables, but the most important one for calculating torque is current.
‘Kt’ is the motor torque coefficient, whose units are N.m/A (Newton-meters per Amp). Kt is the ratio of torque:current, a relationship that is not perfectly linear.
By itself, this coefficient is not of much use to us, but we can use the assumption that Kt = 1/Kv, and Kv is a number that we do have.
Let’s say we want to deliver 20 A of current to our motor. We can use this number and our motor’s Kv to calculate the theoretical torque.
Motor Kv is generally given in RPM/V, but for it to work in our formula we need it in SI units, which is (Radians/ Second)/ Volt in this case. To make the conversion we will divide 2300 by 60 s and multiply it by 2π.
For a more precise explanation of motor Kv see our article: How to Calculate Motor Kv
Now we can plug our SI Kv value into our motor torque formula:
This figure seems reasonable, but the nonlinearity caused by the ESC and the motor can mean that theoretical results do not necessarily reflect reality. We generally expect a difference of 10 - 50% between the theoretical value and the measured value.
Note that this equation works on the assumption that the relationship between torque and current is linear, which is not the case, so the torque calculated will not be perfectly accurate.
Let’s hook up the motor to our test stand and see how close we get to our theoretical value.
We used the test stand to run a simple step test, taking a measurement when the system reached 20 A. The results of that test are below.
Figure 3: Torque vs. Current results
At 20 A, we measured 0.133 Nm of torque with the Series 1585, a 46% difference from what we calculated.
As you can see, contrary to what the formula suggests, the relationship between current and torque is not linear. This is due to losses from the ESC and the motor.
This data goes to show that torque calculations can only give us an estimate of the true torque produced by a motor. To obtain accurate torque data, you need a tool to measure it.
We’ve been aware of this dilemma for quite some time, which is why we developed our motor test stands to allow users to collect highly accurate motor data:
Further reading: Brushless Motor Power and Efficiency Analysis
If you have any further questions don’t hesitate to leave us a comment below.
]]>Windshapers are the latest technology in drone aerodynamic testing, but what are they exactly? In this article we explain what a Windshaper is, how it works, and how it's an improvement over other drone testing technologies.
]]>By Lauren Nagel
Windshapers are the latest technology in drone aerodynamic testing, but what are they exactly? The terms ‘open air wind tunnel’, ‘wind generator’, and ‘3D wind flow creator’ are all accurate, but they don’t fully capture what the Windshaper can do.
In this article we will explain what a Windshaper is, how it works, and how it's an improvement over other drone testing technologies. Here’s an overview:
Table of Contents
Figure 1: A fixed wing drone flying in front of a 12 x 8 module Windshaper
The key component of a Windshaper is its main wall of fans that produces wind profiles for testing drones in various conditions. The types of tests possible include wind resistance tests, free flight tests, and many more.
Windshapers are highly modular in shape, as they are composed of base units called ‘wind modules’ that can be stacked and combined in any combination. The size of the Windshaper is described by the number of modules it contains, i.e. a 6 x 3 Windshaper would be 6 modules wide by 3 modules tall.
Each wind module measures about 25 x 25 cm and contains 9 individual wind pixels. Each wind pixel is controlled individually and contains two fans. Fans can produce a wind speed of up to 16 m/s (36 mph), or up to 45 m/s (100 mph) with a convergent, depending on the size of the Windshaper.
Figure 2: A 2 x 2 Windshaper with a wind module and wind pixel labeled
Each fan can be individually programmed in terms of speed and timing, thus permitting the user to create 3D wind profiles. We will discuss this more in section 3.
In addition to the main wall of fans (1), users can add side walls (2) to create crosswinds, simulate corners, and diversify the testing possibilities. The other key components include the power distribution boxes (3), and onboard computer (4), all seen in figure 3 below.
Figure 3: The structure of a Windshaper
What you’ve likely noticed is that these Windshapers come in all shapes and sizes - some can fit on your desk while others require their own lab space.
The Windshaper is controlled with the WindControl software, whose GUI is pictured below (figure 4). The PWM legend at the far right indicates the power supplied and each pale blue square represents a single wind pixel (fan unit). The fans used on the machine comprise two layers of counter rotating rotors, called fan layers. By default, both layers are set to the same speed or PWM value. This achieves the best performance and lowest swirl level, but if desired, fan layers can be controlled independently.
In this particular Windshaper, there is a main wall measuring 12 modules wide by 8 modules high, plus two side walls measuring 2 modules by 8 modules, giving you a total of 1008 wind pixels, or 2016 individually controllable fans. All three walls can be controlled simultaneously using either manual control or the advanced Python API.
Figure 4: The WindControl software interface
Manually controlling the wall is as simple as typing in speed values for individual pixels (or groups of pixels), which are activated immediately. When the user is ready to change the power to a group of pixels, they simply select those pixels and type in a new value, offering dynamic control of the Windshaper. You can also input a mathematical function to have the machine reproduce any steady or time-variable wind profile.
A Python API is also provided with the machine that allows the user to program automated tests. The wind flow is designed ahead of time and runs independently without intervention. This allows the user to focus on controlling the aircraft / airfoil or observing the test. The Python API also allows users to generate more complex flow, and to interface the Windshaper directly with other systems (drone, tracking cameras, connected probes, etc.)
Once tests are completed, data from each wind module can be output into a zip folder on the home computer containing time-stamped information about the Windshaper's performance and status.
With a Windshaper you can create a number of wind situations, both constant and time-variable. Here are a few of the basic categories of wind shapes that can be used alone or in combination.
Steady Flow
This setting mimics the constant flow you are mostly likely to see in a traditional wind tunnel, which is great for evaluating the aerodynamics of a drone. A steady condition is generated by setting the Windshaper’s flow speed to the speed the drone would be traveling in still air, while the drone maintains its position above the ground. In this scenario, the relative wind speed, as seen by the drone, is equivalent to the speed the drone would be flying. A flow straightener may be used to achieve a wind flow with a lower turbulence level.
Figure 5: A drone flying in steady wind
Turbulent Flow
Turbulent flow is ideal for simulating the conditions a drone is likely to face in its work environment due to weather and topology. At the altitude drones typically fly, it is unlikely they will experience laminar flow. In this test setting, the level of turbulence can be controlled by modulating the power delivered to each wind pixel. The turbulence level can be equal across the Windshaper or different in each section of the test area.
Figure 6: A drone flying in turbulent wind conditions
Shear Flow
The term ‘shear flow' describes a wind profile where adjacent layers of fluids move parallel to each other at different speeds. This can lead to flow instabilities near walls, foliage or in regions with noticeable thermal effects. This can be simulated by setting wind pixels on one fan array to a slow wind speed and setting wind pixels on an adjacent fan array to a higher wind speed. This technology also enables users to easily generate a boundary layer without needing to use traditional boundary layer generators such as roughness blocks or spires that are found in traditional wind tunnels.
Figure 7: A group of drones flying in an environment with wind shear
Time Variable Flow
With time variable flows you can create unique wind profiles by changing the wind speed of each wind pixel over time. A given wind pixel may begin at 2 m/s, increase to 10 m/s, then return to 2 m/s and so on. With this level of control you can create wind shapes that simulate real-life flying conditions, like a vehicle passing a drone (figure 8).
Figure 8: The wind speed experienced by a drone changes as it flies behind a truck(truck.mp4)
Wind Gust
Sudden changes in wind speed (gusts) can be challenging for a drone to navigate. Gusts can be simulated with rapid changes in wind speed coming from the wind pixels. This allows you to study drone displacement or resistance to gusts, and the responsivity of the flight controller. Adding additional side walls to the Windshaper is another way to simulate gusts and crosswinds.
Figure 9: A drone experiences a gust of wind as it flies around a building
Vertical Wind / Landing Phase Optimization
While landing, drones experience a relative wind from below caused by their own wake (downwash), which leads to an unstable situation. To simulate this situation, the Windshaper is placed horizontally and generates a wind flow equivalent to the drone’s downwash.
Figure 10: A drone practicing take-offs and landings
As you can see, there is a great variety of wind shapes you can produce with these Windshapers, and this article doesn’t even cover them all. If you have a specific type of test in mind, contact our sales department to see if it’s possible.
There are several advantages and disadvantages to working with a Windshaper for drone testing and validation. Here are a few of the most important ones:
Figure 11: Testing a quadcopter with an 8 x 8 Windshaper
The Windshaper has several complementary systems that can enhance your drone testing. Contact our sales team for more information about any of these add-ons:
Propulsion Test Stand
Windshapers can be partnered with a measurement test stand to allow you to record performance data while your propulsion system experiences different wind conditions. The 2 x 2 Windshaper Station is specifically made for this application, allowing you to test motors and propellers producing up to 5 kgf of thrust that are up to 16” in diameter.
Figure 12: A 2 x 2 Windshaper Station
You can also use a larger test stand with a custom sized Windshaper for larger drones. For example, the Flight Stand comes in two sizes for measuring 15 or 50 kgf of thrust for single or coaxial motor setups.
A propulsion test stand lets you measure the thrust, torque, RPM, power and efficiency of your system at every stage of your flight.
Convergent
A convergent device allows you to increase your maximum wind speed from 16 m/s (36 mph) up to 45 m/s (100 mph). The convergent attaches to the front of the Windshaper and passes high speed air through an enclosed test area. Note that this device will reduce your test area size, depending on the size of the Windshaper. Adding a divergent after the test section (as in figure 14) can further increase wind speed past 55 m/s (135 mph).
Flow Filter / Turbulence Reduction
A flow filter can be added to your Windshaper to reduce unwanted turbulence and ensure a low turbulence intensity. This feature ensures an even flow of air and is a great option for more traditional aerodynamics testing using force or pressure measurements, which can benefit from a lower signal to noise ratio.
Figure 15: A 6 x 3 Windshaper with a filter to reduce turbulence
Tilting mechanism
The tilting mechanism add-on allows you to rotate your Windshaper 90 degrees, so it can be standing vertically or laying horizontally. This is a great setup for simulating the effects of thermals, ground effect, and practicing landings in different conditions. It is not only a tool for measuring your drone’s performance, but also for allowing your pilots to refine their skills.
If you have any remaining questions, leave us a comment below and we will be sure to get back to you.
You can find Windshapers among other products on Metoree's comparison list of Wind Tunnel Manufacturers.
If you’re interested in purchasing a Windshaper, you can request a quote from our sales team.
]]>By Lauren Nagel
Many electric motor manufacturers promise impressive performance up to and beyond 95% efficiency.
The numbers on the datasheet are impressive, but do the real values measure up?
We conducted a study using motor efficiency data to try and answer the question: What is the Average Efficiency of an Electric Motor?
There are no industry standards for testing and presenting motor performance figures, so it is nearly impossible to compare manufacturers during the purchasing process.
That’s why we are big supporters of testing your motors yourself, as it allows you to determine their true performance and find the best one for your design.
In this article we will cover:
Along with our testing equipment, we offer a free online database of motor performance data.
This data has been collected and uploaded by the users of one of our electric motor test stands, either the Series 15XX or the Series 1780.
These test stands collect data on thrust, torque, RPM, current, voltage, electrical power, mechanical power, motor & ESC efficiency, propeller efficiency, and overall propulsion system efficiency.
Click here to see the formulas used to determine these values.
At the time we are writing this article, the database contains motor & ESC efficiency data for 48 unique motors. In creating the short list, we were careful to account for duplicates and tests with missing data.
Once the motor list was finalized, we looked at the data for each of the motors. For some motors there was only one test performed whereas for others there were as many as 22 tests.
For each motor we looked at all the data points available and used the ‘maximum observed efficiency’ value for our analysis.
For the motor below, for example, there were 8 tests available, and we used the max observed efficiency value (indicated by the blue arrow) from all of the tests for our analysis. It is important to understand that because different propellers were used, the torque applied was different between tests. The motor efficiency is a function of RPM and torque.
While we cannot guarantee that this is the motor’s true maximum efficiency, it is the highest efficiency that is observable with the data available. Due to this limitation and others, we will not be naming the motors included.
Further reading: Brushless Motor Power and Efficiency Analysis (article)
In our analyses we compared the efficiency of all the motors, and also that of subgroups based on weight. The boundaries were created to try and have an equal number of motors in each group, while grouping them with motors of a comparable size.
There is a gap in weight between the second last and last group in order to more clearly represent the size of the motors included.
Before we get into the results, we would like to discuss a few limitations of the study.
This study is purely observational and based on the data available to us in our database. We did not control for the environment of the tests or the type of test performed, nor can we say whether the data uploaded is representative of the motors’ normal performance.
Therefore, this is not a precise comparison. This is simply data we have compiled for interest’s sake and for a general overview of the matter.
While there are certainly others to be considered, below are a few of the limitations of this study.
Study Limitations:
First off, let’s look at the motor efficiency data of the group as a whole.
The average maximum percent efficiency for all 48 electric motors included in the study is: 77.64%.
Considering that the data was not collected for maximum efficiency purposes, that’s not bad. This is likely a good estimate of typical electric motor efficiency.
The highest maximum efficiency observed was 97.61% and the lowest was 55.79%.
The graph below shows the maximum efficiency distribution:
When we break it down further into groups based on weight, this is what we observe:
Weight Category (g) |
Average Maximum Efficiency |
3.9 - 30.0 |
70.80% |
30.1 - 36.0 |
67.63% |
36.1 - 70.0 |
79.98% |
70.1 - 159.0 |
82.78% |
159.1 - 255.0 |
84.13% |
1093.0 - 5130.0 |
85.42% |
As you can see, there is an overall trend towards increasing efficiency as the weight of the motor increases.
If you include all the data points individually, the trend is similar though dispersed:
These graphs tell us that in general, we can expect to see a higher maximum efficiency as the size of the motor increases. This is in line with the common knowledge of the industry today.
We can also see that there is a wide range of efficiencies for motors of the same size, especially for motors around 30 - 50 g in weight.
For most UAVs, swapping a motor at the bottom of the range with one at the top of the range could increase your flight time by several minutes.
In conclusion, according to this observational study:
If you would like to get started with testing your electric motors and propellers, check out our propulsion testing tools:
Vibration is a common obstacle that needs to be overcome when building drones. Excessive vibration can cause issues in flight or during testing.
In this article, we are going to demonstrate two ways of approaching the vibration problem in drone thrust stands: simulation and experimental testing. The solution we used to change the fundamental frequencies using dampers is similar to what you would use with a drone.
We recently released the new Flight Stand 15/ 50, a thrust stand/ dynamometer made for testing drones that can handle loads up to 50 kgf of thrust. The stand has a hollow tube structure in order to offer better aerodynamics as well as improved cable management, with cables running inside the tube instead of along the outside.
An outstanding challenge in designing the thrust stand was limiting the amount of vibration during propulsion testing. As mentioned, the stand’s basic shape is a vertical tube that must house a Force Measurement Unit (FMU) and hold a motor and propeller mounted at the top (figure 1).
The stand is used to test motors and propellers, which inherently induce vibration, potentially damaging the tool. To optimize and protect the stand, we performed a series of vibration tests similar to those used for drone components such as quadcopter arms. Our experimental procedure and results are outlined in this article.
Figure 1: Early design concept of the Flight Stand.
The thrust stand is adaptable to the customer’s motors and propellers, and with this comes variation in the design parameters. One parameter with a strong influence on the frequency of the system is overall mass.
The equations below show how an increase in mass will decrease the natural frequency of the system. The stand can measure a large range of thrusts, therefore different masses of propellers and motors will be used. Depending on the user's propeller and motor mass, the natural frequency will change [1]:
Due to variations in the mass, propeller size and thrust of testable propulsion systems, the thrust stand has the potential to operate in a wide range of frequencies. With this, the resonant frequency of the structure may fall within the operating range of the stand.
Resonance occurs when the natural frequency of a structure is excited, resulting in a larger amplitude of vibration or what can be seen as greater movement in structures. Maintaining that frequency results in increased excitation and can lead to structural damage or failure.
A common example of the danger of resonance is the Tacoma bridge collapse of 1940. While there were several reasons for the collapse, a major contributor was the resonance that caused the bridge to twist and move uncontrollably.
When designing the stand, we noted that the natural frequency would fall within the range of operation. The thrust stand has the ability to measure anywhere from 1.5 kgf to 50 kgf, which corresponds to propellers that can be run anywhere from 1,000 - 10,000+ RPM, correlating to frequencies of approximately 16 Hz - 200 Hz.
We ran some vibration simulations in order to observe the stand’s reaction to various vibration modes, shown in figure 2. The vibration modes represent the motion that will occur if a certain vibration is maintained. It is important to refrain from operating the stand in these modes for extended periods of time to avoid resonance.
Figure 2 : Vibration modes of the thrust stand (*note that these are early tests and do not reflect the current figures for the Flight Stand 15 /50*)
A prototype stand was built to allow for a better understanding of how these vibrations would impact the stand in practice. While simulations are very helpful, it is important to see how the stand is affected by vibrations during real tests.
The stand in figure 3 was assembled and tested with an adaptable motor mount at the top instead of the FMU. A propeller trimmed on one side was used to create imbalance in the system.
Figure 3: Prototype thrust stand for vibration testing
The vibration dynamics were recorded using the Flight Stand software. One key observation was the importance of a stable base. When the stand is not securely attached to the ground, the base experiences more movement, resulting in a lower resonant frequency and higher vibration readings.
Figure 4 shows how the resonant frequencies can be determined through testing. After the motor is started and the propeller speed is slowly increased, large jumps in vibration can be observed throughout the test, which can be associated with resonance.
Figure 4: Vibration of the thrust stand with increasing propeller speed
Typical damping methods such as sandwich mounts were considered, however vibration dampers must be used in the same direction that vibration occurs. The locations available on the stand for mounting damping are in the vertical direction, but vibrations primarily occur laterally, posing a challenge.
The initial tests were run on a prototype that was designed and assembled in house. A second round of tests was run with an official prototype, machined in a shop with correct tolerances and fit. During these tests there was minimal noise and the stand was very rigid. As in the previous test, we observed that the stability of the base was very important.
With the stand secured directly into concrete, the stand experienced very little vibration and there was minimal noise at the highest vibration recorded. When mounted to the rails, there was still minimal noise and movement in the stand but the vibration reading was higher and the stand was slightly less stable.
If there are significant vibration or noise issues with the stand mounted on the railings, we recommend adding more concrete fasteners to limit movement in the rails.
The updated resonant frequencies of the Flight Stand 15 and 50 can be found in their datasheets.
Similarities can be found between the vibration observed in the thrust stand and that observed in drones. In both cases, vibration is caused by the movement of the propeller and motor, with added complexities for drones because they are designed to have constant motion and carry fragile electronics.
Drones are equipped with many sensors to monitor the system and provide useful data about the vehicle's position and speed, but additional testing in the lab can help predict and prevent vibration-based damage to the structure and to the fragile electronic components.
Figure 5: A drone in flight with a camera load
The vibration analysis of a multi-rotor frame is explained in depth in reference [2] with some similarities to the tests described in this article. First, a propulsion experiment is conducted to determine the rotational speed at which the propeller induces the most vibration.
With this information, a simulation modal analysis is completed followed by an excitation experiment to see how the frequencies impact the system.
As demonstrated, the vibration testing process is similar for thrust stands and drone components. Simulation and calculations are an important part of the process, followed by experimental testing.
These experiments can provide important insight for designers and can ultimately help reduce vibration in test tools and the aerial vehicles they test.
Further reading: How to Build a Thrust Stand (Bearings and Hinges)
[1] Engineering ToolBox, (2017). Beams Natural Vibration Frequency. [online] Available at: https://www.engineeringtoolbox.com/structures-vibration-frequency-d_1989.html
[2] J. Verbeke, S. Debruyne, “Vibration analysis of a UAV multirotor frame,” In Proc. ISMA2016, 2016, pp.2329-2338.
]]>By Lauren Nagel
We’ve added a simple but very useful new feature to our Flight Stand software - the ability to remotely control the software from any computer on the same network as the computer that is connected to your Flight Stand hardware.
This feature is possible with our Flight Stand 15 for small/ medium drones, our Flight Stand 50 for large drones, and our Flight Stand 150 for extra large drones and eVTOL.
This feature has two main advantages:
You can watch the video below for a tutorial on how to set it up, or you can follow the steps in this article. We’re using version 1.9.2 of the software.
Further reading: Static vs. Dynamic Testing with a Thrust Stand
Step 1: If you have the Flight Stand software open, close it, then find the shortcut for the software.
Step 2: Right click on it and select ‘Properties’, then find the Target Field.
Step 3: At the end of the line add a command line argument “--remote” → this tells the software to activate remote mode.
Figure 1: Editing the Target field in the Flight Stand software properties
Step 4: Find the IP address of the computer that you want to control → open command prompt then type the command “ipconfig” and take note of your IPv4 address.
Figure 2: Find your first computer's IP address
As you complete these steps, you'll see a warning message pop up in the software, saying “Remote mode activated”. This is to remind you that you are accepting remote connections from any other computer on the network.
This does have some safety implications since anybody in the network can make your propeller spin. Make sure to only use this feature on a private network, and ensure there is good communication with your team so you have full control of the environment.
Step 5: On the second computer that you will use for remote control, find the shortcut for the Flight Stand software and open ‘Properties’.
Step 6: Find the Target Field and add the command line arguments: “--target=”, then input the IPv4 of the first computer, followed by a colon and “50051”. This is currently the only port that we support. Here’s an example of what that looks like: “--target=192.168.86.57;50051”
Figure 3: How to edit the Target field on the target computer
Now, when you launch the Flight Stand software, the display will mirror your actions on both computers.
Remote control is now set up and you’re ready to start testing.
Step 1: Go to properties and delete the command line arguments on both computers
We hope you'll find this feature useful and if you have any questions feel free to let us know in the comments.
Further reading: 7 Tests To Do With Your Thrust Stand
]]>Our database of drone motors, propellers and ESCs is meant to be a resource for drone designers, making drone test data more available.
This article will show you how to:
It is accompanied by a 3-part video series available on our YouTube channel.
The database is community-driven and anyone can upload test data from their drone propulsion tests using our Series 1580 or Series 1780 thrust stands.
The database currently contains results from over 900 propulsion systems, including 71 brushless motors, 115 propellers and 35 ESCs.
Our landing page lists the largest and highest performing systems tested.
When you arrive at the database landing page you will see a summary of tests, a video on how to improve your flight time, and ‘top 5’ lists of the largest motors, propellers and thrusts measured.
Across the top of the page there are headings for ‘Motors’, ‘Propellers’, and ‘ESCs’. Clicking on one of these links will take you to a list of all the motors/propellers/ESCs with accompanying data in the database.
Figure 1: Headings on the database landing page (red squares)
Next to these headings is “Test data”, which is a list of all the tests performed in the database. This is where you can use filters to narrow by the equipment or performance you need, covered later in this article.
Clicking the ‘Thrust stands’ heading will take you to the main page of tytorobotics.com, where you can view and purchase the test equipment used to generate propulsion data.
Finally, ‘Login’ and ‘Register’ at the far right are used to manage your database account, which is required if you would like to upload data. This is also covered later in the article.
Clicking on the ‘Motors’ heading will take you to the list of all the brushless motors with propulsion test data in the database. There is currently data for 71 unique brushless motors in the system.
Figure 2: List of brushless motors in the system with sorting options
You can sort the motors however you like: by weight (g), shaft diameter (mm), Kv value (RPM/V) and number of magnetic poles.
To sort by one of the parameters, click on the up or down arrow next to that heading. By default they are sorted by the number of tests performed with each motor.
You can change the display units if you like, though this feature requires logging into an account.
Further reading: How Brushless Motors Work
Clicking on the ‘Propellers’ heading will take you to the list of all the drone propellers with propulsion test data in the database. There is currently data for 115 unique propellers in the system.
Propellers can be sorted by weight (g), diameter (in), pitch (in), and material (i.e. nylon, carbon fiber, plastic, polycarbonate, etc.).
The same default options exist for propellers as for motors and the same method is used to sort them (see previous section).
Clicking on the ‘ESCs’ heading will take you to the list of all the ESCs with test data in the database. There is currently data for 35 unique ESCs in the system.
ESCs can be sorted by max current (A) and BEC current (A). The same default options exist for ESCs as for motors and propellers and the same method is used to sort them.
The filters available on the ‘Test data’ page are a powerful way to narrow results and get exactly what you are looking for. They are divided into the categories, ‘General’, ‘Components’, ‘Data’, and ‘Aggregates’.
The example filter applies a filter in each of these categories to give you an idea how it works (figure 3). This is a nice way to get an introduction to the system, and Part 1 of our database video series takes you through the example.
Figure 3: The example filter demonstrates some of the data narrowing possibilities
Starting with ‘General’, you can filter for single motor or dual motor configurations, which can be helpful if you know whether you will have a single or coaxial motor set-up.
You can also narrow results for tests done with a specific test stand if you know the size of propulsion system and the type of data you want to look at. In general, Series 1580 = thrust and torque ≤ 5 kgf, Series 1780 = thrust and torque ≤ 75 kgf.
Under ‘Components’ you can specify the characteristics you are interested in for motors, propellers and ESCs.
For motors, you can narrow results based on a few options: shaft diameter, magnetic poles, Kv value, and weight. Whatever parameter you choose, you can look for results that are greater than, less than, or equal to the value you specify. For example, you could filter for motors weighing ≥ 50 g.
For propellers you can filter by diameter, pitch, and weight. You can apply a criteria for all three of these parameters at the same time by clicking the “ADD” button below the drop down menu.
For example, you could narrow results to propellers that are 5 inches in diameter, have a pitch > 3 inches and that weigh < 5 g (figure 4).
For ESCs you can filter by maximum current, BEC current, and weight.
Figure 4: Example filter for propeller diameter, pitch, and weight criteria
The next category is 'Data'. These filters allow you to narrow results based on performance data for individual or coaxial powertrains.
For individual propulsion systems, you can set limits for thrust, torque, RPM, voltage, current, electrical power, mechanical power, and various types of efficiencies.
If you want to look at multi-rotor propulsion systems, you can look at cumulative performance data for the system, such as total thrust, total electrical power, total mechanical power, motor/ ESC efficiency, propeller efficiency, and system efficiency.
This filter is most important if you know your performance requirements for your drone. For example, if you know you need to be able to produce 5 N of thrust in order to hover, you can search for systems where total thrust > 5 N, then sort the results from most to least efficient to find the best combination of motor, propeller and ESC for your purpose.
Further reading: Automated Brushless Motor and Propeller Testing
The final category, ‘aggregate data’ offers a few options:
Option 1 allows you to search for the maximum value of thrust, power or efficiency and also let’s you apply a cut-off to limit your results so they are greater than, less than, or equal to a certain value. You can then use the sorting arrows in the results table to order them lowest → highest or vice versa.
Clicking ‘exclude empty results’ will clean up your search results by removing any tests without data for the parameter you are looking at.
Option 2 allows you to look at the interpolated value of one parameter at a given value of another parameter. For example, you could determine the interpolated value of total thrust when total electrical power is equal to 3 watts for a set of propulsion systems.
Finally, option 2 allows you to look at the interpolated value of one parameter at the maximum value of another one. This is great for finding the conditions at which your motor is most efficient, for example, the total thrust at a motor’s maximum efficiency value.
When performing propulsion tests with your drone, you can upload your data directly to the database. You can also copy and paste that data into our flight time calculator to estimate how long your drone will stay airborne based on its performance data.
Watch Part 2 of our database video series to learn how to use filtered data with our drone flight time calculator (figure 5).
Figure 5: Our flight time calculator estimates your flight time based on data from the database
Uploading your own data to the database is a great way to visualize your test results for your own use, to share with colleagues, and to contribute to the greater testing community.
Uploading data requires a database account, an RCbenchmark thrust stand, and our free software, which you can download here.
Part 3 of our video series takes you through this process.
Creating a database account is free and simple, just click on ‘Register’ in the top right header on any database page. Enter your name, email address and a password, and click ‘Verify Email Address’ in the verification email you receive. Once your account is verified you are ready to start uploading test data.
The next step is to open up the RCbenchmark software and connect your test stand. In the ‘Setup’ tab, make sure that the box for ‘Activate Experimental Scripting Mode' is checked. This setting is required to upload data to the database.
Figure 6: Check ‘Activate Experimental Scripting Mode’ in the Setup tab
Next, click on the ‘Database Upload’ tab, then specify the name of your test and the throttle points it will cover. The naming convention employed by most users names and describes the motor and propeller used, though that is up to you.
Once you have entered the information, click ‘Run’ then ‘Start’, and the test will run automatically (figure 7).
Figure 7: Propulsion test in progress with automatic script
Once the test is completed, a browser window will open to the online database, prompting you to ‘Add new test’. Click this button then on the next page, fill in the information for your motor, propeller and ESC.
Typing in partial information will bring up a drop-down menu of options so you can select the equipment you used. If your components are not listed, simply add them to the list as prompted.
As you enter the test information it updates and saves it live, so you don't have to click save when you're done. Once all of the information is entered, the landing page for the test will list the components used, test stand information, data plots, and a data table.
The database allows you to compare test results for any tests listed in the database.
Here are a few examples:
We will follow-up with the motor example by clicking on 'Motors' in the header bar and then clicking on the first motor in the list, which is the DYS Samguk Shu 2306. This page includes all the tests done with this motor in our database.
We will use the drop-down menu to show 10 entries, select all of the ones we would like to compare, then click 'compare 10 selected’ at the bottom of the list (figure 8).
Figure 8: The procedure for comparing tests involving the DYS Samguk Shu 2306 motor
This will open up a new page with plots comparing all the data for the tests that we clicked on (figure 9). We can look at rotation speed over time, thrust or torque vs. rotation speed, voltage, current and power properties. And the last graphs look at motor & ESC efficiency, propeller efficiency, and propulsion efficiency.
If you hover over any of the data points it will give you the numbers associated with that data point as well as the motor and propeller used.
If we compare the propeller efficiency and propulsion system efficiency graphs, we can see that at lower rotation speeds it is the HQ Prop 5131 that is most efficient, but at almost all rotation speeds, the propulsion system that is most efficient uses the HQ Prop 5040 with the DYS Samguk Shu 2306 motor.
Figure 9: The plots generated by comparing data sets
This database can be an excellent tool during the drone building and optimization process. The database becomes more effective as more users upload results, so we highly encourage you to upload your own test data.
If you have any questions or would like us to cover any other topics related to the database, please leave us a comment below.
Further reading: BLDC Motor Manufacturers
]]>Every year we have clients come to us after having attempted to build a thrust stand to characterize their motors and propellers.
Despite some clients reporting they achieved their objectives, many also mentioned the resources invested were not worth it, as they lost focus of the main goals of their project.
If you are trying to decide whether to purchase or build a thrust stand yourself, we encourage you to read this article. It may bring to light some factors you haven’t considered, such as:
This industry is extremely competitive and we know that time to market is critical. Our test stands are generally in stock so you can start testing within a few days of placing your order.
(Please note that this article refers to professional drone builds and test stands.)
Figure 1: An early thrust stand prototype and the more recent Flight Stand 15
Motors generate a huge amount of EMI, which tends to add noise in measurements and can make USB dataloggers crash (as we have discovered through R&D). These problems are hard to identify and it is best to carefully design and simulate electrical connections early. Time needs to be set aside to diagnose EMI issues, and an electronics engineer with expertise in power electronics, EMI, and grounding should be consulted.
Based on our research, a commercial multi-axis load cell (that measures torque and thrust simultaneously) is difficult to source with the proper ratio of thrust/torque. This usually means one of the two measurements will not make use of the load cell’s full scale of measurement, resulting in lower accuracy. Furthermore, our load cells are designed to minimize airflow restrictions while maintaining a minimum thickness in order to test coaxial systems. Commercially available load cells will not have that level of specialization.
Finding the right load cells is just the beginning, then designing a way to calibrate them and verifying the calibration is an additional task. For powertrain testing, thrust and torque are never present alone; it is important to calibrate the load cells as close to its operational condition as possible, which means to calibrate both thrust and torque at the same time and to take crosstalk into account when generating the polynomial. Without a calibration machine, there is also no way to re-calibrate them after many hours of use. That is why we calibrate all of our load cells in house to ASTM standards and offer our clients the option of a yearly re-calibration for their force measurement unit.
When we developed our new generation of test equipment, redesigning the software took just as long as redesigning the hardware. The complexity of the software depends on what features you want, but controlling the test stand and data logging are the bare minimum you’ll need.
The Flight Stand software (figure 2) controls the test stand via manual or automated tests, displays live data on customizable plots, stores and displays saved data, prepares data for export, and has many additional features for data processing (see part 5). In addition, it does all this simultaneously at a 1,000 Hz sampling rate.
Figure 2: The GUI of the Flight Stand Software
Beyond recording data, the way the data is processed can have a large impact on your workload. We’ve added several data processing features to our software to make data handling and analysis easier:
These features take time to program, but they save time for the user and are highly practical.
Some unexpected challenges when designing a test stand can include identifying the stand’s resonance frequencies and finding ways to limit vibration. Without knowing the resonance frequencies, you are more likely to damage the stand and your propulsion system. With vibration, there is always a trade-off between stability and measurement accuracy, and it can be challenging to find the sweet spot.
After mastering a single-rotor testing set-up, there is the possibility you will want to test dual motors in a coaxial or offset configuration. Face-to-face propeller configurations are simple enough, but back-to-back configurations, as is seen in many quadcopters, are more of a challenge. Getting the rotors reasonably close to one another takes careful design and planning. With our Flight Stand 15/50 (figure 3), coaxial stands can safely operate in a back-to-back configuration with only 9 mm of separation, which comes to about 91 mm of separation between motors.
Figure 3: Flight Stand 50 in a back-to-back coaxial configuration
Another factor that takes significant development time is designing a compact test solution. If the test stand itself interacts too much with the airflow from your propellers, it will affect your readings. Building a compact solution that doesn’t compromise on stability is a lengthy endeavor.
One of our priorities when redesigning our test stands was to have a compact design that would interfere with the airflow less than or little more than the motor itself.
This is a consequence that all drone builders dread: you’ve built an aircraft based on data you thought was reliable, and now you’re facing a situation where your prototype’s performance is less than what was expected/promised.
Our test stands have an error rate of <0.5%, so you can be sure the results are highly accurate. Just ask Marc Stollmeyer from Inspired Flight.
Our test stands may not always fit your project requirements, but we hope after reading this article, you will have the knowledge to make an educated decision on whether building your own test stand or purchasing one is best for your project.
If you would like to use a reliable, professional test stand to get results you can have confidence in, we invite you to learn more about our products:
Click here for the Series 1580 - up to 5 kgf of thrust and 2 Nm of torque
Click here for the Flight Stand 15/50 - up to 50 kgf / 30 Nm
Click here for the Flight Stand 150 - up to 150 kgf / 150 Nm
]]>