How do you cyclicly trigger data acquisition after n pulses counted

[Kevin_Price]

Have only had time to skim through the thread.  A few additional comments:

1. Your experience of needing to request a period of 73 "ticks" in order to get an actual period of 57 ticks is definitely not normal.  I'd be *very* reluctant to depend on any solution that continues to depend on such an odd discrepancy.

2. Just looking by eye, the incoming pulses look a little cleaner on their falling edges than their rising edges.  I would recommend an additional configuration step that would make your pulses depend on the falling edges rather than the rising edges.  It should at least be helpful for troubleshooting, and is probably a good idea in the long run too.

   You'll need to use one of the DAQmx property nodes.  I'm not at an LV box to check now, but I think you might find it in a DAQmx Channel property node.  I'd look in the region of Counter Output-->General Properties-->Timebase-->Active Edge, or somewhere roughly sorta around there.  There's a property in there somewhere where you can specify whether to react to rising or falling edges of the signal that is acting as the timebase, in this case an external signal wired to a PFI.

3. The latching behavior is also not normal.  I don't have any real good ideas.  Are you checking your task's error out cluster for any clues?

-Kevin P. 

[fezline]

Yesterday I decided to clean up the pulses by puting a buffer (two inverters in series) inline with the incoming pulses. This appears to have fixed the problem, I am not sure why the traditional daq functions were not quite as sensitive to the not so square pulses. It appears that once I cleaned up the pulses with the buffer everything works as expected. I must say I feel bad thinking this was a software issue and wasting your collective time and effort. I will post some screen shots of what the results are and also what the Traditional DAQ code was that I was using.  I thank you all for your help, I really appreciate it.

~ Randy Brown

[fezline]

Ok, to start out with the traditional DAQ block diagram that worked with the horrid pulses is shown below. Now I think the reason it worked is becuase I programed that one to work on the falling edges which as indicated by the previous Applications Engineer could have fixed the problem with the DAQ mx function (with a property node).

Here is the front end

This is the resulting scope shot I was getting with the poorly formed pulses

The following scope shot will be the pulse train with the buffered pulses going into the PFI line.. It now works stable and the pulse spacing actually relates to the numbers I input into the high and low ticks.

   I belive this would have also worked if I just changed that property node such that the active edge would be the falling edge as suggested by the applications engineer. I thank you all for your help I really appreciate it once again.

~ Randy