sync 2 period measurement tasks

[Eyal_S.]

Hi John,

When I check the problomatic signal on an ocsilloscope the edges look clean. The peak is at 4V for high impedance termination and 2V for 50 Ohm termination. What voltage is the trigger set to (and at what termination)? Maybe my signal is too close to the trigger. Maybe there is an impedance mismatch which causes all the problems? 

What is the ideal signal for counter tasks on the x-series cards? 

Thank you,

Eyal

[Eyal_S.]

The 30ns filter fixed the problem with the problomatic signal.

[John_P1]

 Great news!

The digital inputs are high impedance and require TTL logic levels to activate (2.2V minimum for logic high, 0.8V maximum for logic low).  The protection circuitry on the PFI lines limits external frequencies to 25 MHz.  However, this is assuming clean signals and connections.  Any fast signal coming in through the PFI lines is going to be somewhat altered by front-end of the DAQ board.


Best Regards,

[Eyal_S.]

Everything worked well for a while until I bought a new detector.

The new detector generates a 15ns wide output signal and so the digital filter no longer works.

I'm puzzled why the period measurement task does not have any problems with my input signal (i.e. no spurious edge detections) but the edge counting task does. Any ideas? This could be the key to getting rid of this problem for good.

My input signal looks clean on an oscilloscope but it still seems to create spurious edge detections with the edge couting task. What can be the cause of that? I can try and filter the signal before it goes to the card but since it seems clean to begin with I'm not sure what to do with it. What are the optimal signal conditions in order to eliminate spurious edge detections?

Thank you,

Eyal 

[John_P1]

Hi Eyal,

I also do not know why you don't have any problems when you configure the counter for a period measurement.  Are you using the external signal as the Gate (Input Terminal) or the Source (Timebase) for the period measurement?

If the time between pulses is at least 40 ns you should be able to use a second counter to "filter" the signal by triggering a single pulse to be generated on each edge of your input signal.  Duplicate edges that occur while this counter output is still generating would be ignored, and you could count the internally generated trigger signal instead of the external signal which is giving duplicate edges  Of course, this effectively halves the number of counters you have available at a given time for performing measurements.

If you don't mind external conditioning, you could alternatively look into using an external buffer with hysteresis (schmitt trigger) to clean up the signal.  Something like this perhaps (although you'd need 2 of them to cover all 4 counters).

Best Regards,

[Eyal_S.]

Hi John,

I'm using the input signal as the gate in the period measurement task.

I can't use another counter to filter the signal because I'm actually using all 4 counters.

I'll try to clean the signal externally.

Thank you for the help,

Eyal

[Eyal_S.]

I just realized something...

When I use the period measurement task my input signal goes into the gate input of the counter.

When I use the count edges task to do the same thing I had to reverse the clock and the source. This means that my input signal is now the clock. 

On the cards hardware, does it treat the gate input and a clock input differently? Is the hardware circuitry different for the gate input compared to the clock input? If the hardware has some additional filtering circuitry on the gate input, it could explain the reason why my signal works fine

using the period measurement task but causes problems during the count edges task. 

Thanks,

Eyal

[Frank-L]

This issue is being taken offline to be discussed on the phone, when a resolution comes we will update this Discussion Forums accordingly to account for the resolution. Thanks!

[Diesel]

Hello,

Was a resolution found for this measurement? I have 2 PWM measurements (measured in duty cycle) I would like to synchronize and I am experiencing the same error.

Regards,

Scott

[Aldo_A]

Hi Diesel,

I am currently looking if a resolution of the measurement was found for this specific case.