LabWindows/CVI

cancel
Showing results for 
Search instead for 
Did you mean: 

pci 4461 DSA minimum voltage and precision

Hi.

I'm using PCI-4461 on LabWindowsCVI and I'm reading the voltage on an AI channel of a signal.

When calculating the Vrms it seems to work except when the cable is unplugged from my Function Generator.

For instance, when reading 1 Vpp I get 0,7 Vrms. Fine.

When reading 10 mVpp I get 0,07 Vrms. Double fine.

But, when I unplug the cable, Signal Express tells me that the noise signal has 7 mVpp of amplitude and my CVI gives me 16 mVrms. Not good.

Is there a problem with CVI's API? or I'm I doing something wrong? Could it be that Signal Express is using a Sampling Frequency diferent from mine?

I hope I was clear with my question.

I apreciate any help.

thanks in advance.

 

Regards,

Daniel Coelho

Daniel Coelho
VISToolkit - http://www.vistoolkit.com - Your Real Virtual Instrument Solution
Controlar - Electronica Industrial e Sistemas, Lda
0 Kudos
Message 1 of 3
(2,760 Views)

Hi, Daniel.

 

The behavior you're seeing is not entirely unexpected. When your generator produces a signal that's 1 Vpeak (actually 2 Vpeak-to-peak), you should read 0.7071 Vrms pretty much only when the signal is a sine wave. Similarly for 10 mVp - you should have 7.071 mVrms when the signal is a sine wave.

 

For any other waveform, the story is entirely different. When you unplug the cable, all kinds of garp gets picked up, usually some power line noise with lots of harmonics and spikes. The peak-to-rms ratio (the crest factor) will be something besides the 1.414 that you expect with sine waves. Probably something a lot higher.

 

Now as to why your peak reading is lower than the rms reading, I can't explain. The crest factor should never be less than 1.0. Your guess that the sampling rate is different between the two measurements is likely a good one. Since this board has anti-aliasing filters, the bandwidth will be lower at a lower sampling rate, and so the picked-up noise would be lower. You should be able to figure out from the software what those sampling rates are.

 

Hope this helps,

Ed

 

0 Kudos
Message 2 of 3
(2,742 Views)

Thank you for your quick reply.

I looked into Signal Express for the sampling rate and this is my current situation:

- Signal Express returns 2,2 mVrms with samples = 2 kSamples and SampFreq = 200 kHz

and returns 1,8 mVrms with samples = 2 kSamples and SampFreq = 100 kHz.

- My LabWindows DLL returns aprox. 6 mV on both cases.

Wich means that my DLL returns at least the double of the real Vrms.

But with higher amplitures my DLL returns correctly the Vrms value.

I'm using the RMS and Bw_FPB functions after aquiring the signal from the 4461 card.

Any hints on what is wrong?

 

Daniel Coelho

Daniel Coelho
VISToolkit - http://www.vistoolkit.com - Your Real Virtual Instrument Solution
Controlar - Electronica Industrial e Sistemas, Lda
0 Kudos
Message 3 of 3
(2,732 Views)