noise in frequency measurement using a counter

[LTong]

Hi John,

This solved the digital filter problme! I used ctr 2 to generate a pulse train (retriggerable, by the signal to be measured) and wired the output of ctr2 to source of ctr3, then measured frequency from ctr3 where the digital filter works well. The generated pulse has a suffeciently high time so that it can also display on the oscilloscope.

One more question: when you say "duplicate edges", do you mean that a second pulse comes in while the generated pulse is still at high? I think I can also use this to reduce the noise level (which is from the APD itself, now sure, after the struggling :-)) if the appropriate frequency of the signal can be estimated.

Again, thanks a lot!

Best regards,

LT

[John_P1]

Hi LTong,

Great to hear the suggestion worked, I think the issue was that your input signal is not stable and is quickly switching back and forth between High and Low values. This would have to be happening very quickly since even the lowest filter setting does not allow the signal through. It would also explain the difficulty seeing the signal on an oscilloscope.

The trigger is not re-armed until the output is complete, so we can adjust the pulse width to be long enough to remain high during the time that the input signal is unstable. Thus, we would only trigger off of the first digital edge, and would not trigger off of any subsequent edges until the pulse has finished generating.

You will probably have to scope out the signal to get a good idea of when it is safe to re-arm the counter and begin looking for another input signal. The cool thing is that the pulse width of the counter output is completely configurable, so we have quite a bit of flexibility once you determine how long to hold off before re-arming the counter.

Best Regards,

John

Message Edited by John P on 11-06-2009 10:28 PM