[bart_dood]

Hi Folks,

I am trying to develop my own controller (brushless DC) based off a microchip application. Its proving to be tricky but part of the problem is I don't know how a regular vehicle brushless controller works, I understand exactly how it works in a steady state way but here are some of the things I don't have a grasp on:

-- From stopped to accelerating. Does the controller start in an open loop by forcing an increasing rotation until its completely a number of revolutions before switching to torque control (IE throttle) based? Does it use hall sensors on a normal brushless motor to estimate position before using something like back EMF for positioning? Or does it work some other way?

-- No load and no throttle (IE push clutch and put into neutral, no regen) does the controller merely switch off the fets/igbts and leave the whole rotor gradually spinning down? or does it go to some very mimimal current/torque value?

-- Regen, does the controller generate a controlled negative torque and therefore negative current (IE feeding back to the battery pack)?

These are some high level questions I don't know the answer to, until I find out I can't start working on my code as I don't know what to work towards.
If anyone has a simple outline on how regular commercial controllers work it would be very useful.

Thanks!

 

[peggus]

-- From stopped to accelerating. Does the controller start in an open loop by forcing an increasing rotation until its completely a number of revolutions before switching to torque control (IE throttle) based? Does it use hall sensors on a normal brushless motor to estimate position before using something like back EMF for positioning? Or does it work some other way?

For traction application a position sensor is practically required. Airplane applications can do an open loop start to get the rotor spinning fast enough that the position can be sensed from the back EMF. I suppose if you have a clutch or a torque converter you cold do an open loop start in a car as well. But if you try to do open loop start with a load on the motor you will just loose sync.

-- No load and no throttle (IE push clutch and put into neutral, no regen) does the controller merely switch off the fets/igbts and leave the whole rotor gradually spinning down? or does it go to some very mimimal current/torque value?

Depends on your controllers programming, could do either.

-- Regen, does the controller generate a controlled negative torque and therefore negative current (IE feeding back to the battery pack)?

That is correct, negative torque = negative current, regen!

Good luck

 

[bart_dood]

For traction application a position sensor is practically required. Airplane applications can do an open loop start to get the rotor spinning fast enough that the position can be sensed from the back EMF. I suppose if you have a clutch or a torque converter you cold do an open loop start in a car as well. But if you try to do open loop start with a load on the motor you will just loose sync.




Depends on your controllers programming, could do either.




That is correct, negative torque = negative current, regen!

Good luck

Sounds like from standstill the motor uses hall sensors as a commutation feedback and just switches phase for a few cycles based off the halls (rather than just trying to drive a ramping speed), then it switches to closed loop field orientated vector control? (which my controller does)

 

[aeroscott]

the prius uses a resolver

 

[bart_dood]

the prius uses a resolver

Is that the same thing as a rotary encoder?

 

[peggus]

Sounds like from standstill the motor uses hall sensors as a commutation feedback and just switches phase for a few cycles based off the halls (rather than just trying to drive a ramping speed), then it switches to closed loop field orientated vector control? (which my controller does)

Once you have position sensors in the machine it doesn't make much sense to switch to sensor less. It sounds like you're trying to start under six step control and then switch to FOC? I guess that might work but I think you'll have less problems using a resolver or similar absolute position sensor and sticking to one control mode. Less headaches that way.