Regen braking

I've been trying to understand regenerative braking of a PM DC motor. I've Googled and Ya-hood but can't get my mind around it. I think I could get it if someone could explain to me what switches in the H have to be on to get the voltage from the motor higher than the battery voltage. (I think I have a current block)
Here is a schematic with the motor shown as a battery (generator) that has a back emf of 6volts because M1&M3 have been driving it at 50% pwm. Am I trying to make it to simple or to hard?

Put the H bridge in "coast mode" and put a DC-DC converter/battery charger on the output of the motor. I don't think you have another choice.

If the battery is charged, dump the result to a "load". To maintain efficiency, you might have to dump to the load and battery.

You don't need a hbridge for regenerative braking, you only need a half bridge (2 FETs).

It's also easier to understand in a half bridge if you need a starting point.

The thing to remember is the motor acts as an inductor and (depending on switching) can produce a higher voltage than its "generator" voltage.

Got it. Top FET on when bottom FET off.
Thanks.

By adding a third set of mosfets to a H-bridge (the drains connected to the motor leads, the sources connected to battery) you can then run the motor forward and reverse, and when the H-bridge is in the off mode, you can turn on the regen mosfets to return the back EMF to the battery. This is the way a EV regen brake works.

But regen only works if theres enough mass in the machine being powered to create enough force.

The H-bridge itself will provide regen braking if switched properly. It does not require any more mosfets.

Regen braking can be done with a 2 FET half bridge or a 4 FET h-bridge. The main difference is the h-bridge is much more complex, slightly less efficient and provides only one advantage; the ability to reverse the motor without needing a reversing relay.

This is distracting: "The thing to remember is the motor acts as an inductor and (depending on switching) can produce a higher voltage than its "generator" voltage."
Most electric cars don't have a gearbox. The motor is direct drive via a reduction and differential. Most motors are very high RPM. This is to get a high HP motor into a small space (diameter).
This means that when you want regenerative breaking, the voltage from the motor will be less than the voltage you would supply to the motor for the same RPM. So you have to put the voltage generated by the generator into a circuit that accepts the voltage and increases it sufficiently to charge the batteries. This way it puts a load on the motor and creates breaking.

Sorry Colin but that's a common misconception.

Under regeneration the motor acts like a voltage source (generator) with a series inductance. If you short the output of the motor (generator) then release the short it generates back EMF higher than it's "generator" voltage. Exactly like a boost converter. So the motor which may only be generating 5v can still supply current back into a 12v battery.

So the halfbridge has 2 FETs, and when the duty cycle is >X the motor runs as a motor with current supplied by the battery and the overall effect is one of a buck converter (large Vin, small Vout to motor). If the motor is turning at the same speed and you reduce the duty cycle <X it temporarily "shorts" the motor, causing current increase in the inductive winding and when the short is released each cycle that current is sent back to the battery, acting as a boost converter (small Vin from motor, large Vout to battery).

It's a property inherent to synchronous rectification. It's a little counter-intuitive but the only real difference between buck and boost (and the direction of current flow) is the duty cycle, and it's relationship to Vin and Vout.

I hope that explains it!

I can see it but I'm not sure I understand it yet.

I'm going for the ugly schematic award at the same time.

Another thing an H-bridge can do MR RB aside from allowing directional control of the motor is to allow for electronic braking (not regenerative braking) Where the H-bridge is simple set to short circuit the motor terminals, that state is the PWM'd to allow for electronic braking.

Yea, I think that's the way my golf cart works. As a matter of fact it has no mechanical brakes. I'm thinking of building a small 3 wheel car (my wife doesn't know this yet.) and would like to use the regen and dynamic brake.

Ahh, RB I think I got it. It is when the back emf is shorted then released that the voltage goes above the battery voltage while the lower fet is on?

Correct. As the FETs can conduct current in either direction when they are on, it either operates like a buck or a boost, so in one mode the top FET is a switch and the bottom FET acts like the diode (buck, motor driving) and the other mode the bottom FET acts like the switch and the top FET like a diode (motor generating, boost back to the battery).

As another example if you put a 2 FET synchronous buck converter between a 24v battery and a 12v battery, and run it always at 50% duty, current will flow (in either direction) from the more charged battery to the less charged battery, keeping the 24v and the 12v always in that 2:1 voltage ratio. So you can charge either battery and it will charge both, or load either battery and it will drain both.

Some of the high-end solar products do bi-directional 24v to 12v battery bank equalising using that basic technique.

Scaedwian, I'm curious how that's going to work? The h-bridge FETs will conduct using their integral body diodes and will deliver current back to the battery in boost mode, which again will be regen braking. Unless you have a big load resistor or something there's nowhere for the bulk of the braking energy to go than back into the battery?

If both of the high side FET's or both of the low side FET's are triggered on at the same time while their opposing pair are off there is no path for current to flow except through the motor, it effectively dead shorts the motor leads, the only voltage seen is what the bulk resistance of the FET's creates? You'd need one hell of a stored magnetic field to create a diode drops worth of voltage to cause the opposite sides FETS to conduct, the voltage the two terminal should be effectively zero, as I understand it at least. I'll need a spice schematic of your exact meaning though as I can't follow this in text. I thought the only time the body diodes would conduct in an h-bridge is if all four FETS were off?

Reply to RB #8 Your concept of short-circuiting the motor is totally impractical as the motor (acting as a generator) can deliver up to its full rated HP when braking and you trying to say that this energy is extacted at TWICE THE HP for 50% of the time.
Not ony would this put enormous pressure on the components but it would require circuitry of twice the normal capacity, just to cater for braking. I don't know of any EV that uses your concept.

Colin, are you replying to Scaedwian re his braking suggestion in post #14?