Need a clarification in Counter Measurement

[kumch]

Dear all,

Overview: 

I am using NI PXI 6722(Multifunction DAQ card) to count the no of pulses generated from source.

it has two counter I/P channel. I configure "Count edge" to count the no of pulses generated.

http://www.ni.com/pdf/manuals/370735f.pdf

My source signal Frequency ranges from 1HZ to 10 HZ. signal waveform is Square waveform. Low state is 0V and High State is 5V

Issue : Getting extra no of pulses than expected. (Ex :For 10HZ, Expected 600 pulse/min but actual 800+ pulse/min)

Observation/workaround : 

NI PXI 6722 has two Onboard clocks(100KHZ and 20MHZ). I am not using any external clock so far for this requirement.

I assume, By default the card is taking 100KHZ clock as ref clock. So it is looking for the change for every 10us. 

I tried with three different sources with the same frequency(10Hz). Two sources out of three sources rising time is less than 10us which counts(600/min) are perfect as per the frequency. The third source (Actual source)raise time is 50us so its giving extra no than expected.

I assume, this issue will get over if I use 50KHZ or less than 50KHZ as a clock ref.

Please suggest me whether my observation is making sense or not. 

Thanks,

Kumaran

[GerdW]

Hi kumch,

The third source (Actual source)raise time is 50us so its giving extra no than expected.

Such slow rise time is a problem…

The TTL specifications are mentioning rise times of as low as 50ns. With that "3rd source" device you should use a signal amplifier like a Schmitt trigger (TTL 7414) to improve the rise time of your signal!

[kumch]

Hi Gerd,

Thanks for your reply.

I have generated Square waveform from a AO channel (one of thr source)with same frequency with 0 to 5V which rising time is less 10us. I routed AO channel output to Counter Input PFI8(Source). its giving perfect edge counts 

In oscilloscope, we have a standard source of Square waveform with 0 to 5V with frequency 1KHZ  which rising time also less than 10us. this is also giving exact expected edge counts value.

we have done some tuning in our actual source so rising time is less than 10us now but edges counts are giving additional 30%. than expected.

Thanks,

Kumaran