Sampling frequency for a analogue pulse voltage signal

[colorizedice]

Hi, friends:

I have an analog voltage signal with the period of 4.3 ms, pulse width of 0.4 ms, and amplitude of 5.2 V. I tried to catch the signal by BNC 2110 & PCIe 6321, using it as a trigger signal.

By my design, I can see the waveform from the graph, however, the waveform got by Labview is different with the waveform from oscilloscope, just part of it. I think it means labview just caught part of the pulse, maybe the pulse width is too narrow for the labview to catch.

Do you have any advice? Thanks a lot.

[JensG69]

Can you explain in detail the difference between LabVIEW and your oscilloscope? Perhaps by providing screenshots from both signals?

Regards, Jens

[colorizedice]

Yeath. Here are the pictures.

[JensG69]

First step: Adjust the time scale in LabVIEW & Zoom-in:

The oscilloscope shows 5 pulses, ie. about 25 ms. The LabVIEW graph currently shows 1 s, i.e. about 230 pulses.

Or measure just for 25 ms.

Regards, Jens

[colorizedice]

Thanks, Jens.

The period of the pulse signal is around 4.3 ms, which is not hard to catch. What I'm worried about is the pulse width, which is just 0.4 ms. To the picture, we can see the pulse is even in the scope, but randomly in the labview. So, I am thinking the labview can not catch every pulse signal.

[Henrik_Volkers]

You sample with 100kSPS (means one sample every 0.01ms) so for the rising and the falling edge you get an uncertainty of 0.01ms .. so you can expect a 0.02ms variance for a 0.4ms pulse.

(your scope has what...1GSPS? ), Your DAQ has a bandwidth of 700kHz, the scope XX0MHz 😄

Small improvement: Increase the samplerate to maximum 250kSPS.

But to get better measurements of the pulse periode and length you should use the timer/counter inputs. Have a look at the examples on how to measure periode and pulse length with counter.

[colorizedice]

Thanks for your advice, friend. I tried to use the counter port, and solved the question.

Thanks.