PXI-4462 Gain Accuracy and Measurement Validity

[ral112]

I am using a PXI-4462 to perform different dynamic measurements. The PXI was coming due for calibration and before the unit is sent out, a calibrated function generator is used to do a calibration precheck of the accuracy of the different channels and gain levels in case the unit was damaged or lost in shipping. During the check, all channels met accuracy requirements except for DC coupled measurement at the +30dB gain level. Since the last calibration, all measurements have been done at the 0dB level in both DC and AC coupling.

Does the fact that accuracy requirement was not met at the +30dB level call into question the validity of the measurements done at the 0dB gain? (Or any other gain, for that matter?)

Looking at the analog input block diagram for the PXI-4462, the signal at the different gain levels passes through the same differential amplifier, but it seems logical that the accuracy at a specific gain level only has bearing on the accuracy of measurements performed at that gain level.

Thanks for any insight you can provide.

[santo_13]

What does your calibration certificate state? did the module perform as expected across all ranges?

[ral112]

The calibration certificate from last September stated that the module passed all verification checks. No adjustment was needed. The calibration was performed with Calibration Executive Version 3.6, Procedure Version 3.0.2.1, and NI-DAQmx 18.5.0.

I seems logical that if the module were to pass at all gain levels except for one, that the measurements performed at those passing gain levels would be valid.

[santo_13]

Yes I would assume the same. It would be good to get a detailed calibration test report of the ranges and values tested.

[ral112]

Thanks for your reply. The calibration report, if it is like the last one, is very detailed and has the acceptable range and the "as found" measurement for each level. So I will know after the fact if the 30dB gain level is not meeting accuracy requirements.

When an instrument is first calibrated, the assumption is that the measurements done with it are meeting the accuracy requirements. When the instrument is recalibrated, if one gain level is found to be out of spec, then any of the measurements done at that gain level may not be accurate. I guess the measurements could be re-evaluated with the accuracy shift in mind to see if having a little less accuracy may have impacted the results. Measurements where the inaccuracy may have affected the validity of the results could then be redone.

I guess there really is no way to totally guard against this situation even with a yearly calibration schedule. Fortunately, I think that for my current situation that since my critical measurements were done at a gain level that meets accuracy requirements, that the fact that a different gain level is not meeting them does not invalidate the original measurements. Hopefully the module will ship well and the calibration report will state the same.