What is the order of roll-off on PXI-4071

[MarkD-B]

I am using 3 PXI-4071s in digitizer mode. At sample rates of 50, 100, and 250, kHz and observing a 1 kHz 540 V pk-pk sinusoid, the peak of the digitized signal on all three cards reads +/- 262 V. If I use the same setup to read 270 V DC then the result is 270 V.

The documentation states that the bandwidth is 245 kHz for 1000V range. Does someone know what the order of the roll-of is please, so I can determine whether the error I am seeing is due to the front-end filter and can be compensated for in LabVIEW?

[James_McN]

Hi Mark,

I have escalated this to get hold of the additional information and will get back to you as soon as I hear back.

Regards

[ChrisR.]

You should not be seeing a -3% gain error at 1kHz.  What are you using for your source?  Are you sure it's really giving you a 270V peak signal?

Chris 

[MarkD-B]

Hi Chris,

  The source is an electric machine. Back EMF is being measured at a given shaft speed. When the voltage is measured with a (expensive!) Yokogawa digital chart recorder this yields +/- 270 V. I have been unable to independently confirm the Yok, but the Yok agrees with the predicted voltage at this speed and the machine designer is convinced that 262 V is not real.

[ChrisR.]

To sort of answer your original question, the rolloff on the 4071 is pretty high order, and at your sample rates, the frequency response will be dominated by your choice of DC Noise Rejection.  In other words, if you use the default DCNR setting, you'll see a sinx/x response with the first zero at your sample rate.  This would introduce an error of only -0.07% at 1kHz, even at your lowest sample rate of 50kHz.  At higher sample rates, or with the other DCNR options, the error would be smaller.

I just tested a 4071 I have here, and it was right on.  Admittedly I didn't have any high voltage sources handy, but the error was tiny with a +/- 10V signal on the 1000V range.

What's the model number of the Yokogawa instrument?

Chris 

[James_McN]

Hi Mark,

Just wanted to see how you had got on with your PXI issue as there has been no activity for a while now.  Have you managed to get your application working?

Regards,

[MarkD-B]

Hi James, Still the same problem at the moment. I've been on holiday for a few days and have been side-tracked by something else but should be able to continue investigating next week.

[James_McN]

Hi Mark,

Just a thought for some troubleshooting.  You say you are using 3 of the 4071s.  Is this to measure a three phase machine or are they linked in some way?  Have you tried this using just one of the devices connected to see if that has any impact on your readings?

Regards,

[MarkD-B]

Hi James,

  One card normally measures each of red-blue, blue-yellow, yellow-red. If I connect all three cards to the same phase, I get the same reading on all the cards but still not the same as the Yok.

  I hope to be able to do be able to do some more testing tomorrow with a third high-speed acquisition system to prove whether the 4071s or the Yok is in error.

All the best,

Mark