Task measurement values not aggreeing with Test Panel measurements

[lmd2]

We are reading accelerometers (with excitation) using PXI 6254 modules. At first we thought we were seeing a DC offset, and we were and we filtered that out. Now if we measure the sensor output at rest with an O'scope we read uV. Similarly if we read the channels using MAX test panels we also read uV. But when we make DAQmx tasks and look at the task voltages in MAX (same sensors at rest) we see readings of up to 25 or 30 mV - orders of magnitude higher. Can anyone suggest what may be causing this?

 (I am attaching a word doc with represetative screen shots)

Thanks

lmd2

[cirrusio]

Imd2,

It's not evident from your attachment that the test panel is showing anything different than MAX.  I assume that ai0 will also be equivalent to Voltage_0 in MAX.  If that is the case, it looks as if that measurement in MAX does indeed measure something near zero (as with Voltage_1); but I am not certain that I am getting my colors correct in MAX and there is no legend so I may be off.  Is this not correct?  Are all measurements on the test panel showing something < 1 mV?

Cheers, Matt

[MarkMoss · Accepted solution]

How many channels are you measuring and at what rate?  What is the output impedance of the accelerometer and/or any filters you have between the accelerometers and the 6254's?

It is possible you're running into settling time error.  The specifications for the PXI-625x cards has a graph showing the expected setting time error vs. channel-to-channel interval for various source immediacies.

One way to test for this would be to slow down your sample rate.  If the error decreases as you slow down the sample rate, it is a good bet you're running into a setting time problem.

Mark Moss

Electrical Validation Engineer
GHSP

[cirrusio]

Mark,

It seems that the fundamental problem is that he says that the Test Panels and MAX are giving different results.  If there is a settling time error, it should be evident in both MAX and the Test Panels, no?

Matt

[lmd2]

the config is evident in the screenshots  we are sampling at 60KHz on 16 channels - pretty much maxing out the card (not quite but close). when I read your response I thought maybe that's it, but then as Matt says why is that not the case when using the test panels  - oh yeah ... the test panels only scan a single channel ...

good catch let me look into this with the client thanx!