Bi-directional PWM Motor Control

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
I have a device which runs a small DC motor and can reverse the direction. I want add a speed control without completely starting again.
In the circuit below, if the 'PWM' box (perhaps a '555) generates a waveform which is positive for 75% of the time, will the motor be powered for 75% of the time regardless of the polarity of the input voltage?
upload_2016-8-10_19-21-29.png
 

crutschow

Joined Mar 14, 2008
34,283
No.
That circuit would work only if the transistors were driven by opto isolators.
As shown, the transistors will not be biased properly when the voltage polarity is reversed (Input 1 negative with respect to Input 2).
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
That was the bit I found confusing.
My reasoning was, when input 1 is negative, Q1 body diode is forward biased, the source of Q2 is a diode drop positive of that negative input and its drain is positive of the source. The negative power to the PWM box is also one diode drop positive of Input 1 (equal to Q2 source). So when the PWM output becomes positive (close to input 2 voltage) it will switch on Q2, and when the PWM output goes negative it will apply roughly 0V Vgs to Q2, switching it off.
No?
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
I am familiar with an H-bridge for controlling a motor, but in this case that would mean reverse engineering how the directional control is achieved - that would be difficult. So I already have the polarity switch doing what I want, but I want to be able to control the motor speed.
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
As shown, the transistors will not be biased properly when the voltage polarity is reversed (Input 1 negative with respect to Input 2).
I have realised that it won't work when input 1 is positive. It is trying to use an N-channel as a high side driver so the gate drive is missing.
Here is the MKII version:
upload_2016-8-10_21-7-55.png
 

crutschow

Joined Mar 14, 2008
34,283
You circuit in Post #7 will work but it does exhibit a one-diode drop in the motor voltage due to the conduction of the current in the OFF transistor through the substrate diode when the other transistor is ON.
Is that drop acceptable?
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
You circuit in Post #7 will work but it does exhibit a one-diode drop in the motor voltage due to the conduction of the current in the OFF transistor through the substrate diode when the other transistor is ON.
Is that drop acceptable?
Yes, I don't need it to run as fast as it does now.
 

crutschow

Joined Mar 14, 2008
34,283
Yes, I don't need it to run as fast as it does now.
Then you need to be sure that the MOSFET can safely conduct the motor current in the reverse direction and that the MOSFET can dissipated the power due to the voltage drop and motor current.

You could add a Schottky diode in parallel with the source-drain diode of each MOSFET to minimize the MOSFET current and reduce the voltage drop.
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
Not so. here's what I had in mind :-
View attachment 110296
Ah, I see. That would work. The PWM would have to run from the voltage on C1, but that's OK. This circuit loses an extra diode drop from the maximum motor voltage, and it won't work at 100% (as C1 would discharge) so that loses a bit more from the maximum speed. I can't see any advantage against my MKII version so I think I will stick to that.
 
Top