Controller for (re)generator function only?

[RRR]

Hi all,
After building batteries for a while, I finally start with the actual motor stuff. In the past I have been putting batteries into something that already had its own motor/controller/throttle and so on.

However, I want to start with using an electric motor for generator purposes only. Lets say a 10kW 48v BLDC motor.
Application is like a mild hybrid system, and I will most likely be using a mild hybrid generator, without the inverter.
I know it would be nice to reuse the whole system, usually CAN enabled systems, but I haven't seen to much succes on that, and the ones without the motor controller are easier to work with and cheap.

I think it would be nice to use a controller for that, because then I can just tell it how much to regen from software.
For example a Kelly controller can also have a 0-5v regen power signal.
I guess it would also be possible to work with some big diodes and a controllable DC-DC converter?
Using a motor controller would let me, for later use, have the option to power the motor i.s.o. generating only. But it is not so important right now.

However, I do have some questions.

1. What happens when the controller is idle or off, for long periods of time? Just like EVs aren't allowed to be towed typically, because the motor is always connected and is therefor always generating something which is unloaded. I don't think that is any different in the case of a 48v BLDC with something like a Kelly controller.
Are the controllers in some way built different, for example by being able to withstand those high unloaded voltages or by disconnecting some of the phases when not used?

2. A controller is for example suitable for 48v and 200A continuous for running the motor. Would the generator power be any less than that? I couldn't find any other data on this, but I can imagine that continuous or 1 minute power ratings are different for generation.

Thanks in advance

 

[mxlemming]

Most controllers should be fine with starting up after a long time off, I'm unsure of the exact reasons for ev towing but it might well be due to extensive field weakening use in them... Tldr towing would result in large regeneration or over voltage.

A controller will nearly always be more efficient than diodes because you don't lose the diode voltage.

Generation and motoring power capacity should be basically the same. It's just current through the coils and iron losses causing heat, they work out the same for both quadrants.

 

[amberwolf]

For the motor itself, traction and braking (generator) are electrically the same. For the controllers, they may not be.

There are different methods of regenerative braking in different controllers, some of which are more efficient than others (creating less waste heat within the controller, and getting more current back to the battery from the motor). I can't recall the names of the various methods, but some effectively just use the FETs as a 3-phase rectifier, which isn't typically variable--it always presents the same load for the same motor RPM, and the load is based on motor RPM. Some actively switch the FETs during regen to create higher battery currents (and better braking), and some of these allow control of that current by the throttle input or a separate analog braking input. Some have RPM or battery voltage limitations, not allowing regen when above or below those limits.

So the "rated current" of the controller for driving a motor might be more than the current it can regenerate back to a battery. Also, the current in some ocntrollers (like at least some Kellys) is not the battery current, but rather the driven motor phase current. For those, the current back to the battery will be significantly less than that.

Some of the regen methods actually use power to help brake the wheel, so you don't want those.

I suspect the best type for your usage would be an FOC controller that you can tune for the best regeneration current at the RPM you will run the generator at under load.


For EV towing, there can be multiple reasons to not have the powered wheels on the ground. It could be a gearbox or other physical design limitation, not allowing backdriving safely, or it could be an electrical limitation, where the unpowered controller or some other part of the electrical system isn't designed to handle the generated voltages, etc.

 

[Jrbe]

For question 1:
There will be some system losses turning the generator at no regeneration.
The generator will make different voltages at different rpms when it's not connected to a battery.
The generator might not regen at low rpm if it's not spinning fast enough to make the voltage required to charge the battery.
1 & 2 A failure mode could be the battery being disconnected from a blown fuse or opened contactor. At the generators max driven rpm, the unloaded voltage would ideally be under the mosfet rating in the controller. If it's above and it disconnects the mosfets could blow. There's also battery voltage / charge to watch, if you can't charge the battery any more because it's full / too cold, etc. or it can't use the generated power the system voltage can go up with rpm and cause the same blown mosfet issue.

 

[RRR]

Thanks for the answers.
For question 2: I agree that it is just current through mosfets and coils. So the direction doesn't matter too much. But since controllers are mostly built for driving motors and a bit of regen, you might be fooled by some specs, when the manufacturer did not specify everything. Probably not the case, and full regen power hopefully available.

For question 1: I think the description of @Jrbe is probably comes the closest to what EVs have to endure when they are towed. Or going downhill when full.

My application will be 48v but I guess it would be better to choose a controller that goes up to 96v. This way I should end up with mosfets inside that can handle higher voltages.

I can actively disconnect the motor as well with a contactor on 2 of the 3 phases