Pulse Width Counting

Hello All, I am using a Micrologix 1100 with a HSC counter. The HSC counter is working the way it should be. I am trying to detect the Duty Cycle of a PWM input. I have been trying to figure out the time between pulses, but can't seem to figure out the proper way. The frequency of the PWM is 140hertz. Anybody have any examples or suggestions? Thanks

Normally I just setup a timer to take a sample every X seconds. So if the accumulated value increases by 280 every 2 seconds your would be running 140 hz

The poster refers to the Duty Cycle. I believe they are trying to measure the individual ON time and OFF time of each cycle.

As the cycle lasts for a little over 7 ms you would need a clock of much higher accuracy (say to the micro second) to give an accurate measurement. I think that the Free Running Clock is the fastest clock available to the program and that increments every 100 microseconds giving approximately counts in one cycle.

If you connect the PWM signal to two inputs and use two interrupt (EII) routines (one for rising edge, one for trailing edge) and sample the Free Running Clock at each one you may have some usable information but not of high accuracy. You would be able to tell if the signal changes by a large amount, maybe 10% but not much closer.

Good luck.

Bernie, I was thinking the same thing...would the resolution be 70 counts then? And half of that could be on or off so you end up diving integers like 30/40 to find the duty cycle. But we must add in any input signal latency, variability, and that of the code being executed in the interrupt routine.

I would just but a din rail mount wilkerson converter...I think they have a model with a comparator for frequency pulse width...

Surely there's a dedicated device, or a way to reconfigure the signal at the sending device.

http://www.electronumerics.com/micro/micro_duty_cycle_meter.htm
http://www.laurels.com/duty-cycle.htm