Paul and Sabrina's Cheap 3 Phase Inverter (AC Controller) with Field Oriented Control

[Automcdonough]

I think you should put a load on that motor to truly test it. Even a flywheel would do. Something to let you tinker with rate of acceleration and regen.
This should also involve mounting to something to keep it from rolling around like a metal wrecking ball.

[Automcdonough]

silly question here.. with that sort of motor is the regen something you can control/adjust? It's not like an alternator where you have a field strength control.

[MPaulHolmes]

With field oriented control, you have i_field and i_torque. For an ACIM, you can just have i_field = i_torque. You have i_torque proportional to throttle. Then if you want regen, just have throttle be negative. Then you are basically regenning at whatever negative torque you want, as long as it's available. I don't know how to have that type of control without FOC though.

[dcb]

I *think* you can drive the gates at a slower speed than the rotor and it will push the current back through the controller. Maybe.

[MPaulHolmes]

That makes sense. I'll try it, once I add much smaller gate resistors. hahaha. That friggen thing became a toaster oven.

[Automcdonough]

I dunno about that, it could be an active brake, using more juice to stop until it finally switches direction. If that's how it works out it's not a total loss, would be a nice feature to have if the brakes fail!

This made me think of something though.. if it's passive only then it's being rectified through the FET diodes. You'll be in for a bad surprise as those diodes suck!! high drop and terrible reverse recovery. Back them up with something substantial, one across each FET (a 3 phase rectifier).


Well, I ought to tell you why I went there.

I'm starting to think that the ideal motor would be more like a giant repurposed alternator. The control code would be based on brushless DC instead of induction, but the rotor field would be adjustable for the purpose of controlling torque (throttle or brake). So I'd need another PWM source there for the field. It should also be possible to replace the slip rings with a rotating transformer of sorts but really they aren't nearly as troublesome as brushes on a DC motor to begin with.

At first I figured it would be nice to get a stronger field than magnets or induction will do, DC's main advantage.. but the other benefits are nice. If doing a hybrid field+magnet I guess only modest magnets would do well to lower the static field current without causing too much EMF at light load.

someone at the diy board posted this:
Induction Versus DC Brushless Motors | Blog | Tesla Motors

Your controller should be able to drive brushless DC with only software modification (assuming the position encoder that comes with the motor is compatible).

So.. if you start tinkering in that area could you save an extra PWM out?

[thingstodo]

Originally Posted by Automcdonough

silly question here.. with that sort of motor is the regen something you can control/adjust? It's not like an alternator where you have a field strength control.

Yes, you can control regen back to the batteries.

Slow the output of the VFD down to the speed of the motor minus your slip and it starts to generate back to the DC bus.

If your four pole motor is turning at 1750 rpm, your vfd output is 1800 rpm and your slip is 50. To regen, set the vfd output to 1700 rpm and the energy in the motor is dumped back into the DC bus, and the batteries.

As the speed of the motor drops, you need to keep dropping the output of the VFD if you want to continue to regen.

My VFD isn't a vector model. It only does Volts/Hertz. I can only regen down to about 300 rpm.

thingstodo

[MPaulHolmes]

Hi Louman! I'm sorry, I just got busy and haven't checked the thread lately. I haven't been to sleep since. The super annoying thing is, the baseplate got really hot because my gate resistors were way to big, and I'm STILL waiting on mouser and digikey to send me smaller gate resistors as some schottky diodes. I should have used 1.1 Ohm resistors to turn the gates off, and 3 Ohm to turn them on, but I have been using 10 Ohm on and 10 Ohm off. If they don't get here tomorrow, I'm going to friggen wig out.

[MPaulHolmes]

Well, I was experiencing "shoot-thru" as they say. My turn off times were so dang long that the built-in dead times that I had weren't long enough, so both the high and low of each of the 3 phases were on at the same time for short instances, which creates an unholy amount of heat. I feel very sure that the reduced gate resistor sizes will take care of that.

[NiHaoMike]

Originally Posted by MPaulHolmes

Well, I was experiencing "shoot-thru" as they say. My turn off times were so dang long that the built-in dead times that I had weren't long enough, so both the high and low of each of the 3 phases were on at the same time for short instances, which creates an unholy amount of heat. I feel very sure that the reduced gate resistor sizes will take care of that.

An old Philips digital amplifier I have had a similar design problem. The driver chips have programmable dead times, but it turns out the resistors that set the dead times were actually way out of spec! I replaced them with the suggested values in the datasheet and the amplifier now uses only 20w in idle instead of 44w.