LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Measuring voltage with the NI USB-6211 at MHz

Hi all,

I am not an expert on data acquisition however I would like to have a VI to measure the voltage of a signal which is about 1-3 MHz and 0-5V by using the NI USB-6211. Now I am aware of the fact that the 6211 has a rate of 250kS/s which is not enough. Now, I don’t really care about the frequency of my signal. I just want to get the AC voltage like peak to peak. Doses somebody has an idea how such a VI would look like?

Thank you in advance 

0 Kudos
Message 1 of 7
(2,697 Views)

Hi Nigel,

 

Doses somebody has an idea how such a VI would look like?

Yes.

 

Did you try to search for suitable example VIs in the ExampleFinder?

Something like "Continuous Voltage Measurement"?

 

want to get the AC voltage like peak to peak.

Try to find the min/max of the voltage signal to calculate peak-to-peak…

Best regards,
GerdW


using LV2016/2019/2021 on Win10/11+cRIO, TestStand2016/2019
0 Kudos
Message 2 of 7
(2,691 Views)

Hi,

thank you for your reply. I already check some examples but in the matter of measuring they are more or less the same like I have DAQ AI Voltage->Sample Clock->Start Task-> aso.. But here I always end up with some anti-aliasing signals. I test it with a frequency generator and the measured voltage was also lower.

0 Kudos
Message 3 of 7
(2,670 Views)

Back in the old days a RF-Probe was used for things like this.

http://www.qrpkits.com/files/RF_Probe_How_To.pdf

 

 There are a lot of DIY tutorials on the web. Keep in mind that the type of diode is important..

I would do some sort of amplitude calibration at 100 kHz ... or use a scope ..

 

Or you build a RF RMS converter .. linear has a IC for that ...

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


Message 4 of 7
(2,664 Views)

Normally, you'll get recommendations that your sample rate should be at least 2x the signal frequency when you need frequency-domain information and more like 10x the signal frequency when you need time-domain information.  Your board's max sample rate is in the order of 0.1x the signal frequency.

 

Normal isn't gonna work.  Abnormal might.  Here's an abnormal idea that's a little dubious but just might work good enough.   First a little trip down "background info" lane...

 

Let's suppose the signal of interest is exactly 2 MHz, making the period 500 nanosec.  Let's further suppose that the signal is pretty regular, like a pure sine wave.  Normal 10x sampling would sample every 50 nanosec so you'd get a 10-pt approximation of a sine wave.

 

Here comes the tricky part.  The signal of interest repeats every 500 nanosec.  You'd like samples at every 50 nanosecond fraction.  But you can't sample fast enough to do that in a literal way.  So instead, we're going to try to sample every 5050 nanosec.  That'll give you a sample at 0.0 cycles, 10.1 cycles, 20.2 cycles, 30.3 cycles, etc.  Even though you've missed a lot of stuff in between, the data you get will be exactly the same as if you sampled every 50 nanosec (still assuming a pure sine wave on the signal of interest).

 

The corresponding sample rate for this 5050 nanosec interval is about 198 kHz, putting it within your board's capabilities.

 

The next further refinement is that for your kind of measurement, you'd like to sample *all* parts of the repeating signal, not just the exact 1/10 cycle parts.  Well, a small tweak to the sample rate will make those "landing points" not line up quite so perfectly.  And so 1000 samples will probably catch 1000 different phase points, and you've got a good shot of actually catching a real pk-pk measurement, again assuming the signal is regular and repeatable.

 

I'd probably repeat the measurement at a few different sample rates to make sure I didn't accidentally pick one that happened to line up as an integer divisor of the signal of interest.  Then I'd believe the one that saw the largest pk-pk result over the same # of samples.

 

WARNING!  Always keep in mind that your results are kinda dubious.  The theory I presented is sound, but it's built on assumptions about signal purity and regularity that probably don't pan out in the real world very often. I would venture that this approach would usually tend to do a pretty good job of catching pk-pk if you take enough samples and the signal is reasonably regular.  But you've gotta always remember that you get no samples at all from the vast majority of signal cycles.  Who knows what it's doing when you aren't looking?

 

 

-Kevin P

CAUTION! New LabVIEW adopters -- it's too late for me, but you *can* save yourself. The new subscription policy for LabVIEW puts NI's hand in your wallet for the rest of your working life. Are you sure you're *that* dedicated to LabVIEW? (Summary of my reasons in this post, part of a voluminous thread of mostly complaints starting here).
Message 5 of 7
(2,663 Views)

Hi all,

thanks for the interesting ideas. Well I was hoping it could be done with some just some special DAQ settings in a nice VI, but it seem to be more difficult. The RF-Probe or the RF RMS converter sounds like a good plan B.

0 Kudos
Message 6 of 7
(2,629 Views)

http://www.analog.com/en/design-center/evaluation-hardware-and-software/evaluation-boards-kits/dc427...

 

4MHz AC to RMS DC converter demoboard for 50$...

Greetings from Germany
Henrik

LV since v3.1

“ground” is a convenient fantasy

'˙˙˙˙uıɐƃɐ lɐıp puɐ °06 ǝuoɥd ɹnoʎ uɹnʇ ǝsɐǝld 'ʎɹɐuıƃɐɯı sı pǝlɐıp ǝʌɐɥ noʎ ɹǝqɯnu ǝɥʇ'


0 Kudos
Message 7 of 7
(2,618 Views)