Four quadrant drive
Thanks for your reply. The statement about feeding energy from the lower output voltage to a higher output voltage in quadrants 2 and 4 I agree with. I guess my question is, is there more than one way of doing this that is termed a "four quadrant drive"?
The method you suggest by switching the H bridge diagonally opposite is one technique that I have come across (although I have not implemented it this way). I have always known this method to be called locked antiphase.
The second technique (which I use at the moment) is to switch two diagonally opposite fets (top A and bottom B for example) and then, rather than switch the opposite fets (top B and bottom A) for the remainder of the duty cycle, I open all the switches and allow the current to flow back to the supply through the flyback diodes. This method is what I always understood to be four quadrant drive, and in fairness it does allow smooth operation in all four quadrants. The downside of this approach is, at low current level, the current in the flyback diodes can reach zero before the switches are activated again leading to a discontinuous current flow and non-linearity in the current control loop.
I guess really i'm asking is, in motor control terminology, if someone said four quadrant drive would you immediately assume a configuration such that the switches are always on and diagonally opposed as you mentioned in your reply?
I have also heard the method I described being referred to as a pseudo four quadrant drive, although this may have just been the terminology used by the author of the particular paper that I read.
Added after 23 minutes:
I just noticed I didn't answer your question. By pwm off period I mean the time during which I have opened all the switches and the motor current is flowing through the flyback diodes back to the supply