Frequency Measurement with NI-9361 not stable

[Luisa-lv]

Hello,

I am playing around with a cDAQ-9185 system with a NI-9361 to measure a 60kHz frequency signal.

But the read data abnormal occur periodic. I have tried to changed every parameters even the channel but not work.

As the attached picture show, the normal situation keep 13 sec and the  abnormal keep 18 sec.The waveforms changed periodically.

I have used the Gantner Instruments measured the signal with same sampling rate, the waveform were stable.

Can anybody tell me how to measure frequency with NI-9361 ?

Thanks!

[Kevin_Price]

I can't read the language of some parts of the code screenshot, but this looks like an issue with electrical noise getting into your signal path.

I'd start by being extra careful with signal cable routing and paying extra close attention to other equipment in the area that may be cycling in a way that only *sometimes* induces this noise in your measurement.

The "normal" data looks fine to me.  The quantization down to 5 distinct possible values makes sense as they represent consecutive integer divisors of the 100 MHz timebase used for counter timing measurements.  If both the external signal and the cDAQ timebase were absolutely 100% steady, you'd only get 2 distinct possible values.

The "abnormal" data looks suspiciously regular.  Here "suspicious" is good b/c it probably means there's one specific thing causing the many little glitches at a pretty constant frequency near 3.4 MHz.

-Kevin P

[Luisa-lv]

Thanks for your reply.

Now I just know the "specific thing" is when I powered the SIEMENS servo modules,the "abnormal" appeared. And cut off the servo modules the waveform would be normal. Do you know what's wrong with it? How the electrical noises disturb the 9361 counter？

I think maybe the reason is the hardware channels of 9361 is not isolated enough. Because it's ok when I test with Gantner Instruments, the difference between them is that Gantner's hardware channels are isolated.

I don't know the real reason.Hope forward to your reply! Thanks a lot!

[Kevin_Price · Accepted solution]

It can be tricky trying to eliminate the effects of electrical noise.  Among several things that might be worth trying, if you haven't done them already:

1. Power your cDAQ system from a different circuit than the Siemans motor drives.

2. Route your freq signal wiring as far as possible from the motors and drives

3. Use a shielded cable for your freq measurement, ground only 1 end of the shield.

4. After doing everything you can to prevent or reduce the noise from getting into your signal path, *THEN* try configuring your task to include digital filtering.  It's always better to first try to *eliminate* noise before merely *compensating* for it.

    You should check your device's manual to learn more about the specific behavior of the digital filter - the details are a litttle different across various device families.

-Kevin P

[Luisa-lv]

Hello Kevin,

Thank you very very much! I have learned a lot from your reply!

Finally, I found the 24V power cable of 9361 was not shield. The high frequency interference disappeared when the cable shield.