ELVIS II Reads high mV in AI and DMM

[ptompkins]

Howdy! I am having trouble getting a good reading for a mV signal from my ELVIS II box. I am running Windows 7. Whether I use NIMAX or the DMM for ELVIS, I am getting roughly 52mV when the power supply indicates the voltage should be closer to 23 mV. A hand held DMM reads the 23 mV. I have other items with much higher readings, and they do not seem to be wrong. When the power supply is off, I get a good zero reading.

[Karl-G]

How is the power supply connected to the Elvis II (single ended or differential)? Also, is the Elvis measuring inputs from anything else on the same circuit? If it is accurately reading other inputs and other voltage ranges, it is likely a wiring issue. If you decrease the range of the measurement (the smallest range should give the least error), does the offset decrease?

[ptompkins]

Thanks for the help. I think that the problem was that I was taking a measurement  from a resistance that was so low, 0.01 ohms that I could not make a two wire measurement. this is the voltage across a shunt resistor.  I need a four wire system that would allow me to take into account the lead resistance. I could not figure out how to make a four wire measurement with the ELVIS. Maybe I am completely wrong. I ran the current measurement through the ELVIS current measurement and am currently living with the 2A limit. The measurement works great. I would prefer to run the experiment at 3 amp, but at least it works.

[ptompkins]

Forgot to mention that I had been using the highest gain in both cases.

[Karl-G]

Are you currently measuring the voltage off of the current shunt? As you mentioned, the very small resistance value in the shunt and its accuracy may lead to inaccuracy in the voltage measurement.

[ptompkins]

I was trying to measure the voltage across the shunt. What was weird was that a handheld DMM was perfect, but either the AI or DMM on the Elvis was way off. I am trying a differential measurement with the (-) tied to ground with 100kOhm resistor.

[Karl-G]

That is bizarre that you are still seeing such a large voltage offset. The best resource we have to determine the cause of noise and offsets is our Field Wiring Guide white paper (see the link below). You could try adding another shunt resistor between AI+ and AI Ground, as per Table 1.

http://www.ni.com/white-paper/3344/en/

[ptompkins]

Thanks for the suggestion. I have always used assymetrical tieing of floating signals to ground, since that always has worked in the past. I tried the symmetrical version and did not get different results. I am getting an approximately double signal, 35 mV  (should be 15mv) with a noise signature of +-5mV or constant but random swing from 30-40mV. I am not sure what is happening. I have read the white paper before, and used grounding techniques since I learned LabVIEW in 1992. Both the AI and ELVIS II inputs read incorrectly, though they are probably using the same input. thanks

[Karl-G]

That does seem like a lot of noise, as the resolution of the ADC at the smallest range should be orders of magnitude less. Are other inputs reporting accurate measurements?

[ptompkins]

Yes they are, but their gain is set to 1.