# 4 Quadrant Controller, Regenerative Mode Timing

Discussion in 'The Projects Forum' started by DubP, Jan 2, 2015.

1. ### DubP Thread Starter New Member

Sep 9, 2013
16
0
Hi All,

I am interested in building a 4 quadrant controller for a golf cart. I am modeling the H-bridge in LTSpice and I have gotten to the point of working out the switching strategy for the regenerative operation. I read THIS fine article and THIS one, which got me up to speed with the idea that the existing H bridge circuit can work with the back EMF and inductance of the motor to form a boost converter that pushes current back into the battery.

My question is, what are the calculations typically used to calculate the PWM frequency and duty cycle of a boost converter? It seems like I will have the following information available to my control system:

· Vbatt
· VbackEMF
· Motor inductance (constant value)
· Current in/out of the motor

So, I’m thinking the most straightforward way to do this is to shunt the motor until I see “X” current coming out, then switch to the battery, wait until the current drops to “Y”, and go back to shunt. But how do I select values for X and Y? The battery voltage will be changing a little, and the back EMF voltage will be changing a LOT, and I would like to develop a strategy that keeps the regen process as efficient as possible under all conditions. Or is my way of thinking flawed and there is a different way this should be handled?

Last thing; is there a classy way of controlling the system without sensing current?

2. ### ronv AAC Fanatic!

Nov 12, 2008
3,398
1,455
I wouldn't worry about all the boost converter stuff. You probably can't change any of it anyway. Just set the frequency high enough say 20 KHz. It's pretty hard to build an H without regen. You just need to make sure the top FET shorts out the clamp diode when the motor is not being driven.
It's not classy, but you can size the FETs so you don't need current sensing.
Q4D has a pretty good write up on regen.