Consider the h bridge circuit from here:
http://www.allaboutcircuits.com/textbook/semiconductors/chpt-11/pulse-width-modulation/

I refer to the transistors as a matrix:
11 12
21 22


When PWM is 1, 11 and 22 are ON and the motor goes in one direction. When PWM is 0, 12 and 21 are ON and the motor goes in the opposite direction.

But I do not understand. For xT time, the motor goes into one direction and for (1-x)T time the motor goes in the opposite direction. So in one period of the PWM signal, the motor switches directions.
I thought that we use the PWM signal in order to determine the speed of the rotation in some direction by using the duty cycle, and not the direction itself.

What am I saying wrong?


Or is the fact that because it rotates for just a little bit in the opposite direction for (1-x)T, because of inertia and what not, it still rotates in the right direction but just slower?
So for duty cycle of 100%, it rotates at full speed in one direction. And for 0%, it rotates at full speed in the opposite direction.
For 75% it rotates in one direction but slower. And for 25% it rotates in the opposite direction but slower.

Is this the case? If yes, what happens at 50%?

 

The way it works is by activating diagonally opposed transistors only. So for one direction you use 11 and 22, and PWM one of them (most of the time the one at the top, depending on your driver design), while keeping the other one on. And to switch to the other direction you turn 11 and 22 off, and activate 12 and 21.
In more advanced designs, to go in the first direction, you PWM 11 and keep 22 on, while the PWM is alternating between 11 (on a high input signal) and 21 (when the same input signal goes low). This takes care of the motor's back EMF by dissipating its generated current in the motor's winding itself.

 

You may want to check out this thread, to help you see more or less what I'm talking about.

 

Ok, thanks.