Probably stupid question about regen braking

[Pete9008]

Could someone explain regen braking for permanent magnet motors? I just can't seem to get my head around it.

My current understanding is that during regen the motor works as a generator and pushes power back into the battery. The issues I have with this are:

1. To be able to push power back into the battery the back emf generated by the motor needs to be higher than the battery voltage. Isn't this only the case in the the field weakening region?

2. If the motor is working in the field weakening region then it will naturally push power back into the battery through the IGBT reverse diodes even if all the IGBTs are off. This suggest that regen braking in this region is essentially done by backing off the field weakening to allow this to happen. Is this correct?

3. If the motor speed is below the field weakening region then I can't see how regenerative braking is possible. I can see how turning on both lower or both upper IGBTs (to use a single phase analogy) could generate braking by allowing the motor current to circulate around the windings but this will generate braking due to resistive losses in the windings not by pushing it back into the battery. This will also ask more of the motor cooling than normal operation would.

4. The only way I can see it might work is by doing 3. via PWM. The motor inductance might then generate a flyback voltage in the same way boost converters do which would be rectified by the IGBT diodes and so push power back into the battery. The problem I have with this is that the inverter would need to change its switching mode and I've not heard of this being the case.

5. Is the above why the Prius has a boost/buck converter rather than designing the motor to work at battery voltage. It runs the DC bus at whatever voltage is needed to allow full regen to occur irrespective of motor speed?

I'm sure I'm missing something fundamental - could someone explain what it is please?

[mjc506]

4 is essentially correct In normal (motoring) operation, and inverter basically acts just a switching buck converter, so when regening, becomes a switching boost converter back into the battery.

Also, witchcraft. Lots of witchcraft

[Pete9008]

Also, witchcraft. Lots of witchcraft

So does the inverter actually start switching both the low and high side IGBTs on at the same time or does the boost effect somehow happen with the normal switching mode?

[johu]

never ever are high and low side on at the same time.
It's just the normal switching mode. I didn't even think too much about the physics, just commanding negative current does the trick

your assumption #1 is also correct and very disturbing! Keeping the motor from regenning above base speed is proving somewhat hard to do.

[Pete9008]

Sorry I didn't explain very well. The flyback but was having two low side IGBTs on together to allow current recirculation within the motor. As you say high and low of the same phase on at once lets the magic smoke out!

I still don't get how regen below base speed works but if you see current flowing back into the battery below base speed it must work?

The only way I can think of preventing regen above base frequency is to keep FW on and slowly back it off as the regen requirement increases.

[mjc506]

Does this help? https://electronics.stackexchange.com/q ... a-dc-motor

If you can get your head around that, a 3 phase motor isn't much more complicated really (the concept is the same)

[Pete9008]

Yes it does, thanks

It had pretty much clicked overnight (honest!) and I woke up this morning realising that I was thinking about it wrong, once you start thinking about in terms of inductances, in particular back emfs generated by the rotating machine, it makes sense and that article confirms what I had been thinking (and explains it a lot better than I could!). #4 above was close but I hadn't understood that it fitted in with, and could be modulated by, the standard PWM switching scheme.

As you say witchcraft. Lots of witchcraft

[Pete9008]

your assumption #1 is also correct and very disturbing! Keeping the motor from regenning above base speed is proving somewhat hard to do.

I'm now curious about the field weakening, what algorithms are used to set Id and MPTA? (I've had a look in the code but can't find them)

[catphish]

1. To be able to push power back into the battery the back emf generated by the motor needs to be higher than the battery voltage. Isn't this only the case in the the field weakening region?

This puzzled me too, so I simulated it - this shows a 50V motor pushing current back into a 100V battery. It's difficult to explain in words how this is working, but essentially the inverter creates a forward voltage of 150V to create a current, and then reverses the polarity, creating a voltage of -50V, which the energy (inductance) in the motor coils can overcome. I hope the simulation is somewhat helpful!

https://tinyurl.com/y7cxrn72

[Pete9008]

Yes thanks that does help to visualise it.

The trick appears to be thinking about the back emf due to motor rotation and the flyback voltage due to energy stored in the inductor as two separate things (which becomes obvious as soon as you draw the equivalent circuit). I was originally trying to think of them as one and the same thing - which doesn't work.

Happy now!

[slociviccoupe]

Ive had this question for few days now. So what happens when the pack is full and the bms commands to shut off regen. But you still have to stop or are going down hill. Or in my case the honda hybrid ima motor is alwsys soinning with the crankshaft. What does the inverter do or where does the electricity go if you cant regen ?
Is the inverter smart enough to feed current from one phase and conne t it to another to cancel the output so that no current flows? Trying to wrap my head around this.

[remy_martian]

Friction brakes. Look at the size of the friction brakes on a Tesla...it has to stop the car from high speed, going downhill both ways on the way to/from work when the battery is full and there''s no place for the energy to go.

I suppose you could put a 50kW resistor on a roof rack, though...

[mjc506]

^^^That.

I guess, it depends how things are programmed... If the BMS tells the inverter/controller that no regen is allowed, and the inverter/controller obeys, you'll simply get no regen and have to rely on the friction brakes. If the inverter/controller ignores the BMS, the regen will work fine, and continue pumping energy into the battery... which may not cause an immediate problem if the pack has the capacity to absorb that energy without damage!! If the BMS notices the current flowing into the pack, and the pack voltage rising further past it's allowable maximum, and has control over the pack contactors, it could open the contacts to protect the pack. This would destroy the inverter. (The energy from the regen-ing motor would go into the bus capacitor raising the voltage across it until it fails, or until the voltage across the IGBTs reaches damaging levels)

Probably the best option is to have the inverter/controller obey the BMS charge/discharge limits

[slociviccoupe]

I know the friction brakes. Thats no brainer. But say on an ev pack is full and you can't regen. But you are coming up to a stop or coasting downhill. The electric motor is dtill turning thus producing current. Its not like it can be disengaged from spinning with the wheels its geared but direct drive. Where does that current go?
I guess from reading is pack should be higher voltage than the motor can ever make at its max rpm. That way pack will never go over voltage?

[celeron55]

The motor won't produce current if the inverter isn't taking current from the motor. The inverter does this by producing the same exact voltage (as measured by averaging the produced PWM voltage) at the same exact phase that the motor generates, by using a feedback loop based on phase current sensor outputs. This essentially open circuits the motor.

And yes, the battery voltage should be higher than the motor back EMF, unless you have very reliable field weakening implemented which can continue when coasting. Openinverter isn't really capable of that at the moment, but OEM inverters seem to be able to do it. Then the voltage that the motor generates can be surpressed to a level lower than the battery pack, enabling the inverter to zero out the phase currents like I described.

[slociviccoupe]

Celeron55 thank you for that explanation. I was concerned as i am putting together a performance hybrid and the honda ima motor is directly coupled to the crankshaft and spins with the engine. Im using a prius gen 3 inverter with a dual motor board. And the regen had me concerned. I did not know how the inverter handled that situation.