PWM fail detect?

[chico]

Is there a simple way to tell if there is a PWM failure? without using software.
I only really care if the signal stops all together.

This isnt a huge issue, its mostly that i couldnt think of anything that would work. Measureing voltage across some cap to me seems like it would "suck up" the PWM signal entirely.

Maybe in the future i shouldnt try to avoid software, maybe everything would be easier

 

[OlPhart]

Total frequency failure can be a 74123 retriggerable 1-shot. Set the timeout to be just longer than the base PWM frequency & use whichever edge is the constant. Anything dealing with over/under width problems makes using software Way easier. <<<)))

 

[KMoffett]

Google: "missing pulse detector" circuit
Google: "watchdog timer" circuit

This is assuming the PWM never goes to 0% or 100% during normal operation.

Ken

 

[misterT]

Is there a simple way to tell if there is a PWM failure? without using software.
Maybe in the future i shouldnt try to avoid software, maybe everything would be easier

If this is some sort of safety device then doing it in hardware is much better than doing it in software. Much more reliable.

 

[crutschow]

You can couple the PWM signal through a small cap and then rectify the signal. Monitor the rectified signal with a comparator (or just a transistor). When the PWM signal stops, the rectified signal will go to zero.

Proper choice of the capacitor and rectifier circuit values will cause it to have an insignificant affect the PWM signal.