Sanity check: outrunner with large no-load current, low inductance

L

lerneaen_hydra

100 µW
Joined
Apr 8, 2019
Messages
8
Hey all, this is just a quick sanity check to make sure that I haven't made some obvious mistake. I have a 12090 140kv outrunner (28-poles) I am planning to use for an electric outboard rebuild. What I've found so far (that I didn't expect) is:

- Running unloaded at a pack voltage of 40V (giving a speed of ~6500 rpm) draws around 25A. This gives ~1kW of no-load losses. This is very noticable in that after about 20 seconds the entire motor is heated to around 70 degrees, despite a very powerful cooling fan.
- The measured phase-phase inductance is only 1.5uH. This seems very low, I had expected at least 10uH (and maybe even more). I can only assume that the ripple current must be extremely large, especially at full speed when the high-side FET is kept on continuously.

Is there any obvious conclusion that can be drawn here? Is the motor defective (wound incorrectly)? Am I making some other trivial mistake?
 
most likely the phase / hall wiring combination is incorrect.

next most likely is the halls are the wrong phase angle for the controller (60 vs 120, usually).

next most likely is the halls are not producing a clean timing/position signal, either because of magnetic interference or poor placement in the stator.

could be other things too
 
Completely reasonable suggestions, but (as I forgot to mention :oops: ) the motor controller is running sensorless. I've measured the BEMF and all phases are similarly rough sinusoids. Also, for what it's worth, when running at low speeds (unloaded, around 50 rpm) the current consumption is low, maybe around a total of 10-20W to drive everything. Is it common for eddy losses to be so ridiculously large for outrunners?
 
lerneaen_hydra said:
I've measured the BEMF and all phases are similarly rough sinusoids. Also, for what it's worth, when running at low speeds (unloaded, around 50 rpm) the current consumption is low, maybe around a total of 10-20W to drive everything. Is it common for eddy losses to be so ridiculously large for outrunners?
Click to expand...

Try plotting rpm vs current. If you see a lot of non linearity magnetic loss is probably the culprit.

I recall reading that one of the large 150kv hobby king outrunners (rotomax?) drew something like 700w at 12s. Your motor is even larger.
 
I'll definitely try sweeping rpm (for eddy losses) and voltage (for ripple current losses) and measuring the unloaded power draw. Have I maybe oversized my motor? I was aiming for around 3kW continuous mechanical power and around 5kW peak (10-second) power at around 5-6kRPM. Would a smaller motor have sufficed? My thought (naive as it seems now) was that a larger motor would tend to be more efficient. Welp.
 
Welp, looks like its a poorly designed motor all around. Others have found the same issue (https://endless-sphere.com/forums/viewtopic.php?f=30&t=20840&start=100#p307898). Seems to be mostly core losses. Thanks for the help though, now I'm fairly sure it's not just me making a stupid mistake. Back to the drawing board...
 
Probably has oldschool 0.5mm laminations or something stupid like that.
Not a surprise for a motor "rated" at 18000W with thick lams.
Under full load, the motor might be in the upper 80% efficiency though.

Let me guess.. alien motor?

Now you know not to buy things without dyno graphs or at least proper specs on the stator. Almost everyone here has to learn that lesson at some point. Unfortunately the market is full of garbage like this... so for me.. no dyno, no buy.
 
Well, serves me right for being naive and just assuming a not-horrible design. Yeah, alien power.

Anyone happen to know what the general consensus on the 80-100 outrunner variants is? From what I've haphazardly found they seem at least passable: halved pole counts -> no more than 1/4 the iron losses assuming the same horrible lamination thickness, and at least 10uH phase-phase inductance, giving around 1/100 as much ripple current power loss. I'm aiming for around 5kW for 10 seconds, 3 kW continuous, so around 130 A short-term phase current and 75 A continuous phase current at around 5kRPM.
 
Does your 120/90 motor have sufficient torque to maybe gear it down and run it slower? Eddy current losses are proportional to the square of the frequency.

They're also proportional to the square of the lamination thickness, which explains why 0.5mm thick lams suck on anything but a hubmotor! :D
 
Punx0r said:
Does your 120/90 motor have sufficient torque to maybe gear it down and run it slower? Eddy current losses are proportional to the square of the frequency.
Click to expand...

+1 yes, running this motor to 6500rpm unloaded would use 1kw as would producing about 40Nm (that it can do), so loaded to that level is fully maxed out and realistically making about 22kw output and around 90% efficient. -not a bad result for a 4kg motor despite the nay-saying..

-The lams are going to be 0.35mm.

And yes the inductance would be about 1.5uH for a 140kv wind, i reckon you would be right with more like 70kv and would then have around 6uH.

Or another way to run this motor (very) efficiently for about 5kw max would be on 20volts and use a bigger prop?
-You would ideally want a 400a+ capable controller set for 250a and 20-30khz pwm.
 
toolman2 said:
Punx0r said:
Does your 120/90 motor have sufficient torque to maybe gear it down and run it slower? Eddy current losses are proportional to the square of the frequency.
Click to expand...

+1 yes, running this motor to 6500rpm unloaded would use 1kw as would producing about 40Nm (that it can do), so loaded to that level is fully maxed out and realistically making about 22kw output and around 90% efficient. -not a bad result for a 4kg motor despite the nay-saying..
Click to expand...

What controller are you using or recommending that can supply the 611Amps DC required to generate 22kw at 90% eff driven at 40v? That is 2.4kw of heat, how long is a 4kg motor going to take before it turns into a demagnetized paper weight.
 
kiwifiat said:
I wouldn't, i recommended half the kv or half the volts and far less current than this, the point im making is that for a torque oriented motor at this high an rpm level (although the motor is evidently capable) total motor losses are 2kw at peak efficiency and happens at over 4 times the power he needs..

However, on the subject you have now raised with light motors, high power and heat dissipation here's (link below) a 2.7kg motor that's set up for similar max power levels that also has to dissipate a peak of over 2kw of heat in use. zero cooling added, and usually stays below 100deg even in long races. -have a little think about it and see if you can work out how its possible for this little motor to be leading the pack whilst a 15kg hub motored version of the same bike sits in last position and overheats mid race?

With the type motors the OP is seeking help with, they can be very effectively fan cooled or ordered with water cooling as std and i can from my experience answer your question about how long it can last before becoming a paper weight: -over 6 years so far.


https://youtu.be/_Y-_i24yV44
Click to expand...
 
Punx0r said:
Does your 120/90 motor have sufficient torque to maybe gear it down and run it slower? Eddy current losses are proportional to the square of the frequency.
Click to expand...

Exactly. This would be the perfect motor to gimp out and run at like 4,000RPM and extract maybe 3-4kw continuous from. One could run a single stage reduction to the rear wheel easily with it. Even if it has the crappy 0.5mm lams, the motor is useful for *something*.

At 4kg or 8.8lbs, it would be less than half the weight of a proper 3-4kw continuous rated hub ( the boat anchor QS 50mm wide motors are good examples of what a proper 3-4kw hub motor weighs. IE 24lbs or over. )

These crappy, cheap, thick lam, big RC motors actually have a use case. The problem is as usual, that they overpromise and underdeliver.
 
toolman2 said:
kiwifiat said:
What controller are you using or recommending that can supply the 611Amps DC required to generate 22kw at 90% eff driven at 40v? That is 2.4kw of heat, how long is a 4kg motor going to take before it turns into a demagnetized paper weight.
Click to expand...

I wouldn't, i recommended half the kv or half the volts and far less current than this, the point im making is that for a torque oriented motor at this high an rpm level (although the motor is evidently capable) total motor losses are 2kw at peak efficiency and happens at over 4 times the power he needs..

However, on the subject you have now raised with light motors, high power and heat dissipation here's (link below) a 2.7kg motor that's set up for similar max power levels that also has to dissipate a peak of over 2kw of heat in use. zero cooling added, and usually stays below 100deg even in long races. -have a little think about it and see if you can work out how its possible for this little motor to be leading the pack whilst a 15kg hub motored version of the same bike sits in last position and overheats mid race?

With the type motors the OP is seeking help with, they can be very effectively fan cooled or ordered with water cooling as std and i can from my experience answer your question about how long it can last before becoming a paper weight: -over 6 years so far.


https://youtu.be/_Y-_i24yV44
Click to expand...

Agreed you would be hard to find anyone to argue a mid drive would be inferior to a 15kg hub motor on the same bike where vehicle dynamics are important. Despite being at the back of the field I bet the hub motor driver is having almost as much fun as you. Your bike looks very quick and a load of fun. My question regarding longevity was at 22kw. A back of the napkin calculation indicates that with 2400J/s of heat your motor would exceed 200C in less than 2 minutes so clearly when you say peaks of 22kw they must be short and the average output of the motor over the coarse of the race is a long long way short of 22kw. Certainly at an output of say 5kw, 1kw of losses does not look so great but that is not to say the motor is of no use only that it could be better.

I am curious to know what controller you are using on your bike and how your motor is cooled.
 
You must log in or register to reply here.
Back
Top