Analog input measures low

[Sirmutt]

I'm using a PXI-6251 to measure several voltages.  3.3V x2, 5V x2, 15V x1, and 24V x1.  I divide the 15V and 24V by three to get it into a range that can be handled by the 6251.  I used a simple voltage divider of a 200K resistor in series with a 100K resistor.  My issues are with the 15V and 24V measurements.  They measure lower than they are.  I have low tolerance resistors to make sure there isn't any variation.  I measured the voltage on the input to the 6251 with a multimeter and got 4.9996 and 7.96 respectively.   I put an o'scope on them as well and the voltage was solid at 5 and 8 with less than 0.1V variation.  But the 6251 measures 4.85 and 7.75 respectively.  I take 200 samples at 100Hz.  The variation in the measured values is very small as I'd expect, but the values are 0.1 and 0.3V lower than they should be. 

I tried swapping the inputs on the 6251 and the problem followed the 15V and 24V lines.   Is it possible the 300K is loading it down somehow?  I saw the input impedance was >10G so I don't think this should be a problem, but it  appears like it might be???

Thanks.

[PBear]

Hi SirMutt,

This can occur if you are set up taking RSE measurements and your external circuit is at a different ground than your DAQ device. This can create a ground loop and thus, the measured difference. I would recommend opening a test panel in MAX, setting the input configuration to Differential and connecting the positive voltage to ai0 and the other end of the resistor to ai8. If this is not an option, you can also try an input configuration of NRSE and connect the negative terminal to AISENSE.

The 6251 does have a large input impedance and should not be loading your circuit. Are the other instruments connected at the time or are the connections across the resistor made independently? I hope these ideas help and I recommend looking at the "Field Wiring and Noise Considerations for Analog Signals" article in our developer’s zone for some reference. In other news, your icon and username crack me up! Is that the dog from Duck Hunt?

Field Wiring and Noise Considerations for Analog Signals
http://zone.ni.com/devzone/cda/tut/p/id/3344

[ChrisR.]

It could also be a "ghosting" problem. What happens if you just take measurements from one channel at a time? If the problem goes away when you don't scan, then it's ghosting. See http://digital.ni.com/public.nsf/websearch/73CB0FB296814E2286256FFD00028DDF?OpenDocument

Chris

[Sirmutt]

Ghosting was the issue.  Thanks!!!