3phase BLDC motor speed & driver

[dougy83]

Hi all,

I found some 3 phase motors on scrap and was wondering how to work out the rotation speed based on the drive frequency. The motor has 63 coils and 56 magnets..

Also, just showing my ignorance, but what stops an RC ESC being used to drive a larger motor? Is it just the rotation speed? e.g. could a 60A, 25V ESC **broken link removed** be used to drive a 24V BLDC motor?

Thanks.
Doug

 

[Mr RB]

Hi dougy, re question one, to find motor "poles" you run some DC current through one coil, and turn the shaft and count detent positions. That's the easiest way. Then speed:frequency is at one motor pole per full cycle of your driving waveform (excluding slip which is usually only a very small percentage).

Those small ESC devices used for toy motors usually sense the back EMF of the motor coils to determine phaseing and commutation. So with a big motor the back EMF would be very different and they may not work. A 65A 25v RC ESC is meant to drive a motor that has maybe 10 to 15 turns of wire total on each phase! Your big motor likely has hundreds of turns on each phase from multiple coils in series.

 

[dougy83]

Hi Roman, thanks for the reply.

Hi dougy, re question one, to find motor "poles" you run some DC current through one coil, and turn the shaft and count detent positions. That's the easiest way. Then speed:frequency is at one motor pole per full cycle of your driving waveform (excluding slip which is usually only a very small percentage).

Thanks. I thought there might be an equation to determine the number of cycles per revolution. I just put the thing back together, so I'll have to take it apart again to get access to the phases (there's a controller inside), then put it back together to turn it (the case contains the bearings), then repeat to put back to normal. Maybe it'll be easier to just measure the ripple frequency on the supply cable while rotating it, or perhaps shorting the supply leads and feel for the detents as you mentioned (and divide by three).

Those small ESC devices used for toy motors usually sense the back EMF of the motor coils to determine phaseing and commutation. So with a big motor the back EMF would be very different and they may not work. A 65A 25v RC ESC is meant to drive a motor that has maybe 10 to 15 turns of wire total on each phase! Your big motor likely has hundreds of turns on each phase from multiple coils in series.

I think the coils are all in parallel and there are multiple wires going to each phase termination point; there's 63 coils in total, i.e. 21 triplets, so a mixture of series/parallel isn't possible. Phase resistance is <0.1 ohms (my meter can't measure it).

I didn't pay that much attention to the number of wires, but I don't think there would have been 42, so I guess the windings are in star configuration.

The controller described in AVR444 uses a zero crossing detector on the undriven phase to determine commutation. This is done using the ADC to compare to a reference channel that is some fraction of the motor supply voltage.

The tower pro schematic looks like it could use a similar configuration (but no PWM drive) http://i57.photobucket.com/albums/g236/kratchouf/TowerPro25aEscSchematic.png
Dragonsky uses the same (no PWM, identical feedback) **broken link removed**

I believe the actual output voltage and current capabilities (unless the specs are non-repetitive peak values) of the devices should be sufficient. I don't see how the feedback waveform shape will be that different depending on motor size; there might be a need to adjust the divider resistors, I wouldn't know. The thing I would think would be potentially very different is the phase drive cycle frequency (due to number of poles and motor RPM likely very different), and potentially the drive waveform potentially being squarewaves rather than PWM sinewaves.