Brushless motor efficiency

trialspower2

100 W
Joined
Dec 31, 2016
Messages
108
I have a 120100 brushless outrunner which is supposed to be rated at 25kw with a maximum 100volts. 50kv

I have had the motor running on 50v and the no load current is 5.5amp. When I run the motor on 75volts I am getting 10.5 amp no load current and the motor is heating up quite rapidly.

Does this indicate that 75v is to much for the motor and it is loosing its efficiency?

As the motor is for a bike application I could run a lower voltage and higher gear ration and let the motor pull more current. I am thinking this might improve efficiency? I don't think motor torque is an issue
 
trialspower2 said:
I have had the motor running on 50v and the no load current is 5.5amp. When I run the motor on 75volts I am getting 10.5 amp no load current and the motor is heating up quite rapidly.

Does this indicate that 75v is to much for the motor and it is loosing its efficiency?
What speed is it running at? Is it over its maximum? What speed will you use it at?

75 volts at 10.5 amps is 750 watts. That's a lot for a motor to dissipate.
 
That happens on some motors in regards to heat.

Those motors are designed to be air cooled basically. At 25kw, the RC airplane is flying at 200mph or something like that. :)

You will not get 25kw out of it at bike speeds.
25kw is probably a short peak and not a continuous rating anyway, based on how RC motors tend to be rated.
 
Ps, 750w of no load current on a 25kw motor isn't totally out of the ordinary. This figure will scale up as you add voltage and produce higher RPM. That's eddy currents you see.

It could be a lot better with thinner laminations though.. how thick are the laminations on that motor?
 
It's core losses.

Many high performance motors overheat sitting on a bench no-load at full RPM, despite resistive copper heating from the windings. They may stay plenty cool in the application of spinning a propeller, or bursts of high power/rpm like an RC car.
 
neptronix said:
That happens on some motors in regards to heat.

Those motors are designed to be air cooled basically. At 25kw, the RC airplane is flying at 200mph or something like that. :)

You will not get 25kw out of it at bike speeds.
25kw is probably a short peak and not a continuous rating anyway, based on how RC motors tend to be rated.

IS there some point to peak 3kw motor with 25kw. It saturates propably in 10kw. So anything after that is just pushing more amps that generates only heat with no added torque / performance?
 
It's core losses.

Many high performance motors overheat sitting on a bench no-load at full RPM, despite resistive copper heating from the windings. They may stay plenty cool in the application of spinning a propeller, or bursts of high power/rpm like an RC car.

Since my original post I have found that my sensor alignment has moved which is definately making my no load current worse, although I don't know to what degree. I am hoping this will account from my ramp up going from 50v to 75v.

This motor is in a trials bike, so I am only looking for short bursts of output. However I would like to be able to go down the road at a decent speed. Now, if the motor losses are more at higher RPM's, running a higher gear ratio and lower voltage would be more efficient?

I know on this motor the locked rotor current on 24v was 300a. With about 35ft/lb of torque on the motor shaft. On 50v it broke my torque gauge which was supposed to be 80ft/lb, however I drove it into it a bit so I have no idea what happened. So I expect locked rotor current would be 600a on 50v. So on the bike I have a map which only allows a certain output at a certain rpm. So at 75v I am probability only going to allow 35% duty cylce from stopped, which I expect will make a lot of torque! which will hopefully launch you up big rock steps. And as the motor speed increases the allowed duty cycle will increase keeping it at maximum output. So even though the motor is running on 75v, under normal trials riding I can't give it full voltage at all. So I dont know whether 75v offers much advantage over 50v and 50% duty cycle from stopped other than the increased top speed. (Of course the battery pack would need to be different depending on the high amp 50v option vs, lower ampage 75v option)
 
You should be able to achieve the same on most controllers using an adjustable current limit? If you set it at, say, 300A then you don't have to worry about PWM or settings at different speeds.

That some controllers will allow 0.5 or 1.0second over-current before limiting can provide a usefully burst of torque :)
 
This is running on my own controller so I scan write current limiting code also as a backup. I am hoping using a lookup table I will be able to get the throttle 'feel' how I want it. I can look at short bursts of even higher currents, but I am not sure if this will make more torque, or just warm the motor up?
 
trialspower2 said:
This is running on my own controller so I scan write current limiting code also as a backup. I am hoping using a lookup table I will be able to get the throttle 'feel' how I want it. I can look at short bursts of even higher currents, but I am not sure if this will make more torque, or just warm the motor up?

For hotrods or trials bikes needing bursts of torque, more phase current is unconditionally better to have.

Even after your motor iron reaches the "saturation knee", this just means you're no long linear with the phase amps to torque output relationship, but it still gives more torque with more phase current up around ~30-50% over saturation starting depending on the motor design.

For bursts of torque used in trials in the 1-2sec range that drive huge phase current shouldn't be a thermal concern, brushless motors can survive massive current pulses for short intervals.
 
Luke, if there is one thing I have learned from you it's that...if I have a big enough motor using enough amps, I won't be accelerating for long, I'll be up to cruise speed in a very short amount of time.
 
spinningmagnets said:
Luke, if there is one thing I have learned from you it's that...if I have a big enough motor using enough amps, I won't be accelerating for long, I'll be up to cruise speed in a very short amount of time.

Yep, you win burst power EV needs by finishing your brutal acceleration in a timely fashion prior to overheat.
 
Hey trialspower, ive had a reasonable dabble with that motor in its many forms.

They do have a fair bit of loss in the rpm department but are possibly one of the most torquey 120mm motors you are going to find, its possible to get them to 5000rpm at around 900w of rpm related losses with 30khz pwm and some timing advance (yes, they are very sensitive to hall timing) and at 33Nm of torque (that costs about the same 900w) you then have a little over 17kw output at 90% efficiency, not bad for 4.5kg!

The total max measured output (at the tire on a moto dyno) i have seen from this motor is 22kw and 45Nm of torque is a fairly saturated condition.

I would recommend 75v if your controller can do 25+khz pwm frequency just to give you a better speed range (it's not a disaster to lose say 800w at a seldom used top speed when you are likely giving away 2000w to wind drag and the rolling resistance of big soft trials tires at 5psi) but with 28 magnets you will need nearly 1200hz commutation frequency that not all controllers can do.

And maby 9:1 ish gearing to give you a solid 450Nm at the tire for really impressive launches at around the 250 phase amps -assuming the true Kv of your motor is 50?

Ask away, if you want more info i can have a dig.
 
Hello toolman,

It is really interesting to hear from someone with experience of this motor in this kind of application. How did you find the heat build up at higher outputs? I have got the alien power version of the motor which allows for limited water cooling of the core. I have not yet tried using this.

I am concerned with top speed and losses as I would like to do some competitions over here in the uk which involve some distance on the road and moors between sections. I aim to be the first e trials bike to finish these events.

The motor is running on my own controller, which currently has 100khz pwm. I could look to lower this, however I am finding that the controller builds up very little heat. I have just been running the bike at 30mph (top speed with 7:1 gearing at 60v) without heat sinks on the FET's, just a strip of copper bolted between them. They didnt even seem to get warm. The motor was just starting to warm up after say 500meters or so. This was taking I think just under 40a. This was with me standing and a flat rear tyre. (having issues with the tubeless sealing band!)

I might add an extra battery on and try it at 75v and see what current I get at 37mph or so. (will fix the tyre before this)

I could look at going on a higher voltage, my FET's are 150v rated, and I have got my switching spikes reasonble now I think. I think around 10v ringing at 60v. But if I get more losses at speed I dont think I will gain. Low down torque is what im interested in and the motor nearly names enough of this at 24v!.

Are there any good signs with regard to sensor alignment, or is minimum no load current the best measure?

I have written code to ensure that the sensor sequence is in the correct order, but it keeps tripping when setting off. The code currently only allows for forward direction. I am wondering if the chain drive makes the motor oscillate slightly and very briefly go back to the previous segment when on the change over. Or I maybe have to much sensor advance, or both. I think I will modify the code to allow backwards rotation also as this will always happen for sure at times when trials riding.

The true KV of the motor appears to be 50, but that is only going off the code I have written. I havent checked with the scope.
 
Lowering the PWM frequency might help motor efficiency. The fact that the controller isn't heating is a good sign and indicates low switching losses in the FETs, but the iron is going to have higher losses due to hysteresis at the higher frequency. Most motor iron isn't really good at high frequency. Unless the motor has extremely low inductance, I'd go with a lower frequency. It just has to be high enough so doesn't go discontinuous during PWM. Even if it is discontinuous at very low duty cycles, it will probably be better than running the frequency too high.

Most factory controllers run around 15-20khz.
 
fechter said:
Lowering the PWM frequency might help motor efficiency. The fact that the controller isn't heating is a good sign and indicates low switching losses in the FETs, but the iron is going to have higher losses due to hysteresis at the higher frequency. Most motor iron isn't really good at high frequency. Unless the motor has extremely low inductance, I'd go with a lower frequency. It just has to be high enough so doesn't go discontinuous during PWM. Even if it is discontinuous at very low duty cycles, it will probably be better than running the frequency too high.

Most factory controllers run around 15-20khz.

I thought the higher pwm frequency was just the voltage and assuming the inductance of the motor is enough the current would be smoothed and not pulsed once it gets to the motor, so no hysteresis with higher frequency. No?
 
I don't know, you might be right about that. There will always be an AC component to the current (ripple). The magnitude of the ripple will be less at a higher frequency. But the transition takes place more often, so might sort of cancel that out. For motors with lower inductance, the optimum frequency will be higher.

What I do know is guys that are a lot smarter than me that design large motor systems tend to go with lower switching frequencies. Often just high enough so you don't hear it audibly. Whether this is to reduce switching losses in the silicon or the losses in the iron I'm not sure.

The real test is to put the motor on a dynamometer and measure the overall system efficiency under realistic loads at different switching frequencies.
 
My understanding is that high switching frequency is inefficient for the controller only and high commutation frequency was a problem for (high pole count) motors (and probably the controller as well).
 
How can you have high motor current and lower battery current when running no load at like 95% duty cycle? So almost full pack voltage to the motor at such high duty cycle (won’t go over 95%), but the motor current still higher than battery by so much.


Is motor current more inefficient than battery in the motor? It shows running lower duty/ rpm being less efficient on the grin simulator but it incorporates the whole system. At least that would be an easy test and could run at a fixed rpm With different batteries on this live data tool. That would just be a no-load test though
 
Motor current at no load is just the torque to overcome core losses normally. If you have high motor phase current it means high motor torque output if that current is being applied at the right commutation timing.
 
liveforphysics said:
If you have high motor phase current it means high motor torque output if that current is being applied at the right commutation timing.

this image is using the foc program and will try doing it in bldc program right now and post it here to compare, but this image I would consider not high motor torque output since it has no load spun to 94% duty on the table, foc.
 
The current values seen on that software is well inside the current measurement noise expected. If you load it, things will look much different.
 
Without a dyno this is as exciting as I get

Using the bldc program while increasing speed/duty cycle current also goes up even at the last step to 95% duty cycle but with the foc program you can see in the pic the current almost drops by half when making the step from 94 to 95 percent duty cycle. Why is that? I’m not sure if the current measurement is including the Vesc or if it’s even accurate9EFC2273-0F06-4243-B949-6E96B35CB14A.jpeg
 
You're looking at current values in Amps that are around and amp with instrumentation thats +-2amps or more.
 
Back
Top