The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> detect if a pulse train has an irregular pulse intervals
https://designers-guide.org/forum/YaBB.pl?num=1437288804

Message started by jovial on Jul 18^th, 2015, 11:53pm

Title: detect if a pulse train has an irregular pulse intervals
Post by jovial on Jul 18^th, 2015, 11:53pm

i
I have a pulse train. I have to detect whether the pulses are coming at regular intervals or irregular intervals (just yes or no).

Can anyone suggest a simple analog circuit based solution for the same??

Title: Re: detect if a pulse train has an irregular pulse intervals
Post by loose-electron on Jul 19^th, 2015, 1:52pm

without knowing more details:

take pulse train and clock a divide by 2 circuit.

output of divide by 2 should be 50% duty cycle if pulses remain evenly spaced

low pass filter 50% duty cycle wave giving a voltage that is roughly Vpower/2

set up 2 comparators with trip points slightly above and below Vpower/2

digital signals low-middle-high result

when low or high asserts the pulses are not evenly spaced

That is, as requested, an analog method.

A digital method could use the pulses to start-stop counters and compare the difference.

Other solutions are possible as well (PLL and watch the phase detector output and probably others as well)

Title: Re: detect if a pulse train has an irregular pulse intervals
Post by RobG on Jul 21^st, 2015, 12:08am

I can't think of a better "simple analog" solution than loose-electron's, but doing that requires a lot of low pass filtering so it will be large and it will take many clock cycles to settle. I think you want his "digital" solution even if you have to make a simple oscillator to do the counting.

Title: Re: detect if a pulse train has an irregular pulse intervals
Post by sheldon on Jul 21^st, 2015, 1:20am

What about writing a behavioral model that measures the time between pulses and reports 1/0.
0 if the delay is the same
1 if the delay is different
need to include some margin for error to account for finite
numerical resolution.

If post simulation why not plot the frequency. It will be constant if the
interval is constant.

Sheldon