Lets talk about motors
#1
Thread Starter
Tech Master
Joined: Jul 2018
Posts: 1,011
From: Florida
Im curious why we use brushless motors in RC? Is it important to have have maximum torque at 0 rpm with an decreasing slope? Isnt the point to understand your RMS velocity on a track and select a motor and FDR that aligns peak efficiency of the combination to the track?
An asynchronous motor is built for horsepower and the torque curve is the exact opposite of a synchronous motor. Wouldn’t that be a better place to start your hunt for speed?
I understand that if you calculate your rms torque for a track and its skewed towards low rpm synchronous is the way to go and Asynchronous for an application skewed towards higher rpm.
Isnt the point to align the efficiency curve of the motor to the application needs?
i think the critical thing about motors that doesnt get discussed much....that I would like to mention to create context is continuous torque is not what allot of people understand it to be.
you want torque to change velocity. Current creates torque and torque creates acceleration. If it takes 2 or 3 times the current to accelerate versus maintain a speed how come we cant dump 2 or 3 times the current into the motor continuously and just have more torque or more speed all the time? (Assume torque controls speed)
this is why most data that gets published (rarely in RC) shows 2 torque/speed curves and maybe an efficiency curve. One curve (peak or intermittent)is 2 to 3 times higher than the second one. A continuous torque curve (which is what you need to use when selecting the right stuff for an application) is really a peak curve restricted by the motors ability to dissipate heat continuously without exceeding its specifications. Its called a thermal time constant. If a motor is selected and the track you are running on requires and average speed/torque right in the middle of the efficiency curve isnt less heat being wasted and you are spending most of your time in the most responsive portion of the motors capability?
I only explain the torque curves because there is allot of anecdotal data people use to justify a selection....which is ok...and I even recommend things this way sometimes when i dont have something to measure. In robot design we use allot of basic physics that I dont think allot of people are exposed to this combination of information in RC or at least I havent seen it applied. im not here to really prove something to the world but understand if there is a difference. Also when I say robot I mean anything that is controlled by a servo with high/low inertia mismatch and a requirement for high/low dynamics. Robots is a catch all.
An asynchronous motor is built for horsepower and the torque curve is the exact opposite of a synchronous motor. Wouldn’t that be a better place to start your hunt for speed?
I understand that if you calculate your rms torque for a track and its skewed towards low rpm synchronous is the way to go and Asynchronous for an application skewed towards higher rpm.
Isnt the point to align the efficiency curve of the motor to the application needs?
i think the critical thing about motors that doesnt get discussed much....that I would like to mention to create context is continuous torque is not what allot of people understand it to be.
you want torque to change velocity. Current creates torque and torque creates acceleration. If it takes 2 or 3 times the current to accelerate versus maintain a speed how come we cant dump 2 or 3 times the current into the motor continuously and just have more torque or more speed all the time? (Assume torque controls speed)
this is why most data that gets published (rarely in RC) shows 2 torque/speed curves and maybe an efficiency curve. One curve (peak or intermittent)is 2 to 3 times higher than the second one. A continuous torque curve (which is what you need to use when selecting the right stuff for an application) is really a peak curve restricted by the motors ability to dissipate heat continuously without exceeding its specifications. Its called a thermal time constant. If a motor is selected and the track you are running on requires and average speed/torque right in the middle of the efficiency curve isnt less heat being wasted and you are spending most of your time in the most responsive portion of the motors capability?
I only explain the torque curves because there is allot of anecdotal data people use to justify a selection....which is ok...and I even recommend things this way sometimes when i dont have something to measure. In robot design we use allot of basic physics that I dont think allot of people are exposed to this combination of information in RC or at least I havent seen it applied. im not here to really prove something to the world but understand if there is a difference. Also when I say robot I mean anything that is controlled by a servo with high/low inertia mismatch and a requirement for high/low dynamics. Robots is a catch all.
#2
It’s been only brushless since I got in hobby in late 2010. I do own a few running brushed but not for any particular reason other then they came that way, liked how the ran. So I only know the stories of the old brushed days. Simple reason I’ve come to understand it’s they are overall simpler and far less maintenance then the old brushed motors. Hopefully those with experience from the brushed days will respond.
Most of your question I know I don’t understand what’s being asked.
Most of your question I know I don’t understand what’s being asked.
#4
Thread Starter
Tech Master
Joined: Jul 2018
Posts: 1,011
From: Florida
Yea, sorry. i do appreciate the intel but i should have used English. i said synchronous and Asynchronous. The best way to understand it in the real world is an AC motor versus a DC motor. The torque curves are polar opposites. The DC motor creates an electromagnetic field that is synchronized to the rotor of the motor and can be very dynamic but typically has a low rotor inertia so its better for full stop and start. An Asynchronous motor has a higher rotor inertia and the pulses that rotate the rotor maybe faster or slower than the poles on the motor. (Slip)
#6
Yea, sorry. i do appreciate the intel but i should have used English. i said synchronous and Asynchronous. The best way to understand it in the real world is an AC motor versus a DC motor. The torque curves are polar opposites. The DC motor creates an electromagnetic field that is synchronized to the rotor of the motor and can be very dynamic but typically has a low rotor inertia so its better for full stop and start. An Asynchronous motor has a higher rotor inertia and the pulses that rotate the rotor maybe faster or slower than the poles on the motor. (Slip)
The 1st brushless motors were sensorless which means the stator coils will act as sensors to sense the rotor position. At low rpm there is not enough conduction voltage produced to act as a sensor signal and that is why cogging starts. People did not want this and so later the sensored systems were introduced where HALL sensors pick up the rotor position.
I always said the sensored motor is the worst for stockracing. tolerances from factory as also "made" by the users on the sensors can make a wide range on performances on the motor production. Now these days sensorless controlers are pretty good in low rpm but no one will use them. Sensorless motors are much easier to make and to seal with an equal performance because there is no need to open them and there is no outlining of sensors.
The idea of an asynchronous motor I like. If more toque is needed it will run a lower output RPM than the RPM is given by the controler so no matter the FDR, the overal performance will probably more equalized + no sensors are needed and a full control from 0 RPM because the ESC has only to produce a rotating field w/o sensing the rotor position. But that will give a more demand on low air resistance and drive train resistance compared with normal brushless systems.
I always had the thought to use a sensorless system in low RPM range as a asynchronous system. In that way the startup is full controled w/o cogging and at a certain RPM then switching over to sensing rotor position. To be sure the rotor stays synchronized with the rotating field a soft start could be programmed so the rotor will follow the rotating field.
#7
Joined: Mar 2011
Posts: 6,410
From: Austin,TX
brushless motors are vastly more efficient and vastly more powerful, they also offer some very sophisticated mechanical timing features which can be read digitally using a motor analyzer to help you fine tune the readings across all 3 phases to help you optimize efficiency to power settings when shimming the rotor. I don't think this level of tuning is available for brushed motors, yeah maybe you can get some RPM readings, but I don't look at those numbers when I'm tuning my motors on the bench. I set timing based on current draw and check the phases to ensure optimal power/efficiency.
#8
Thread Starter
Tech Master
Joined: Jul 2018
Posts: 1,011
From: Florida
im glad it doesnt sound insane. if your track keeps your motor at an average speed closer to max than minimum having a torque curve that is fat at max makes sense and vice versa of course. Cogging should be an issue in a race car at least its level of importance should be lower than the conversion of current to torque and minimizing heat.
I know its easier to minimize cogging with a very high speed switching frequency but then you get more heat waste. Have you seen a drive or an ESC that maybe switches from 16k to 4k? if you can change the switching frequency without a noticeable bump you may be able to stay asynchronous and just close the velocity loop faster at low speed and cooler at high speed.
I know its easier to minimize cogging with a very high speed switching frequency but then you get more heat waste. Have you seen a drive or an ESC that maybe switches from 16k to 4k? if you can change the switching frequency without a noticeable bump you may be able to stay asynchronous and just close the velocity loop faster at low speed and cooler at high speed.
#9
I have seen brushless ESC's that starts in low RPM in sensored mode and at a certain RPM switches over to sensorless, The GM120 can do that. Switching fequencies I have not seen but I do not think that is the solution to prevent cogging. Cogging happens when the ESC can not reed the rotor position and basically that happens at low speed because the slow movement of the magnetic rotor along the stator will not create much voltage to sense. It is like an altenator which its voltage is depending the rpm speed.
With a magnetic rotor (synchrone brushless) you do not want to happen that the rotating field by the stator is out sync with the rotor speed. That can give huge current peaks and a kind of back fire That is why the position of the rotor is sensed to be sure that the rotor and stator rotating field stays synchronized.
With a magnetic rotor (synchrone brushless) you do not want to happen that the rotating field by the stator is out sync with the rotor speed. That can give huge current peaks and a kind of back fire That is why the position of the rotor is sensed to be sure that the rotor and stator rotating field stays synchronized.
#10
brushless motors are vastly more efficient and vastly more powerful, they also offer some very sophisticated mechanical timing features which can be read digitally using a motor analyzer to help you fine tune the readings across all 3 phases to help you optimize efficiency to power settings when shimming the rotor. I don't think this level of tuning is available for brushed motors, yeah maybe you can get some RPM readings, but I don't look at those numbers when I'm tuning my motors on the bench. I set timing based on current draw and check the phases to ensure optimal power/efficiency.
#11
#12
Thread Starter
Tech Master
Joined: Jul 2018
Posts: 1,011
From: Florida
I have seen brushless ESC's that starts in low RPM in sensored mode and at a certain RPM switches over to sensorless, The GM120 can do that. Switching fequencies I have not seen but I do not think that is the solution to prevent cogging. Cogging happens when the ESC can not reed the rotor position and basically that happens at low speed because the slow movement of the magnetic rotor along the stator will not create much voltage to sense. It is like an altenator which its voltage is depending the rpm speed.
With a magnetic rotor (synchrone brushless) you do not want to happen that the rotating field by the stator is out sync with the rotor speed. That can give huge current peaks and a kind of back fire That is why the position of the rotor is sensed to be sure that the rotor and stator rotating field stays synchronized.
With a magnetic rotor (synchrone brushless) you do not want to happen that the rotating field by the stator is out sync with the rotor speed. That can give huge current peaks and a kind of back fire That is why the position of the rotor is sensed to be sure that the rotor and stator rotating field stays synchronized.
agreed. The synchronous motor has great 0 speed torque but how many people are competing at low speed? I guess the first place to start is by looking at the RMS power for a standard track. If the average speed is above 50% just go simple AC motor (or asynchronous non permanent magnet brushless motor).
do you know how sinusoidal the waveform is? Is it more trapezoidal? I know that if the power circuit creates a waveform other than what the motor is designed for you get waste in the form of heat.
#13
brushless motors are vastly more efficient and vastly more powerful, they also offer some very sophisticated mechanical timing features which can be read digitally using a motor analyzer to help you fine tune the readings across all 3 phases to help you optimize efficiency to power settings when shimming the rotor. I don't think this level of tuning is available for brushed motors, yeah maybe you can get some RPM readings, but I don't look at those numbers when I'm tuning my motors on the bench. I set timing based on current draw and check the phases to ensure optimal power/efficiency.
#14
Tech Master
Joined: Jul 2009
Posts: 1,416
From: Deep South Texas
Fantom dyno with laptop computer
3 different powered brush cutters...
motor lathe with diamond cutters.
4 to 8 different type/kinds of brushes... most with silver added.
brush juice... choice of the month.
canned Freeze. when allowed.
hours of bench testing.
several types of motor bearings.
2 or 3 different types of speed controllers.
never ending replacement of high end batteries.. some every month... and special ones for bigger races..
Turbo 30 and or Turbo 35.. added upgrades
30 to 50 amp power supplies..
10 motors minimum... every gear known to man ... spur and pinion.
3 different powered brush cutters...
motor lathe with diamond cutters.
4 to 8 different type/kinds of brushes... most with silver added.
brush juice... choice of the month.
canned Freeze. when allowed.
hours of bench testing.
several types of motor bearings.
2 or 3 different types of speed controllers.
never ending replacement of high end batteries.. some every month... and special ones for bigger races..
Turbo 30 and or Turbo 35.. added upgrades
30 to 50 amp power supplies..
10 motors minimum... every gear known to man ... spur and pinion.
#15
Joined: Mar 2011
Posts: 6,410
From: Austin,TX
There was much more tuning opportunity with brushed motors than there is with brushless (windings, timing, shimming, air-gap, brush hardness/cut, spring stiffness, commutator diameter, armature stack, etc..). It was common for a guy who liked a 12-turn to have 3-5 different windings (single through quad strand windings) of just 12t. Some used actual Dynos to tune their motors. It was more than a little crazy. They were also just as powerful (modified, stock is on a different level now) even with nimh batteries. Ever since batteries went up over 3,000 mah, we’ve had an excess of power available. Brushless has pretty much saved electric racing for its ease of use and maintenance.
I'd like to see a break down of each tuning attribute and see how things have changed with the migration to brushless racing dominating the market. If brushed motors were so much better, then I don't understand why they are no longer marketed for racing anymore?
Windings - Did folks actually make their own stators for brushed motors? Current brushless manufacturers sell replacement stators, is this not the same thing?
Timing - With brushless motors, this is adjusted by rotating the sensor board and measured within 1° of accuracy across all 3 phases with a motor analyzer, how was this adjustment made with any accuracy on brushed motors?
Shimming - Pretty much the same, you want the bulk of the rotor centered on the stator, the motor analyzer shows you how perfectly centered you are as the reading of all 3 timing phases get closer together... how does this work with brushed motors?
AirGap - I don't even know what this means, please explain more on this and how it can't be adjusted with brushless
Brush hardness/cut, spring stiffness, commutator diameter - These are elements of gross inefficiency that were eliminated with brushless

Armature Stack - I have no idea how this is tuned either, please help me understand this better too.
Something that I see very common with brushless motors today is the ability to swap out various rotors to tune magnetic strength to get more torque/acceleration or transfer to a different rotor which provides higher RPM depending on track conditions. Was this done the same with brushed motor racing too?
When did ceramic bearings first come out? That is another tuning option that I've seen with brushless motors.
I've also seen non-magnetic titanium timing plates marketed as a tuning option for brushless motors to improve performance.



12Likes