analog input offset USB 6009

[dpak27]

Hi,

I am using several USB 6009 units and with some of them I seem to have some offset in AI differential mode, I am using 0 to 1V range. Grounding both terminals with  resistors as discussed in NI tutorial "Field Wiring and Noise Consideration ...." seems to solve the problem for some of them, but for the ones that I do not have any offset to start with, the grounding gives me offset.  Can somebody please suggest?

Thanks,

DS

[Aaron_W. · Accepted solution]

DS,

Thank you for submitting your question on the NI Forums. Are you trying to take in a differential measurement? What devices are you trying to measure and which channels does the pull down resistors work with and which channels do they not work with? 

[dpak27]

Aaron,

Thank you for your quick reponose. I ams using uisng AI 0 for 0-1V Differntial mode, AI 1 and 2 ,for 0-5V RSE mode, and AI 3 and 5 for 0-1V RSE mdoe. The one I am having some troble is with the AI0.

Thanks,

DS

[Aaron_W.]

DS -

If you are using differential on AI 0 it would be necessary to use a resistor such as indicated in the "Field Wiring and Noise..." document. This is because with a differential signal there is no reference to ground and by using the resistor connected to ground you give your signal that reference to ground. Hope this helps.

[Sagarmatha]

When I use the reistors I get  offset of 4 mV compared to offset of 0.3 mV without any grounding?  Do I use the grounding or or no? Any idea why I am getting offset? How much offset is normal? Thx.  - DS

[Aaron_W.]

DS -

What size of a resistor are you using between ground and your AI(-) signal? Ideally you should not see an offset at all. 

Aaron

[dpak27]

Aaron,

Is it better to use higher or lower reistors in general for grounding? In this case I was using 50K. Becuase of the higer offset, currently I am not using grounding.

Thanks,

DS

[Aaron_W.]

I would recommend using a resistor in the neighborhood of 10 Mega-Ohms, that is the size of resistor that we use on our differential cards. That should provide a smaller offset than what you were previously seeing.

[dpak27]

Aaron,

The tutorial suggests 10 Ohms  to 100K Ohms., why 10M? Still would like to know if it is better to use lower or higher Ohms for grounding.

Thanks,

DS

[Aaron_W.]

DS -

Looking at the Field Wiring and Noise Consideration article recommends values between 10 kΩ and 100 kΩ saying that work well with low-impedance sources such as thermocouples and signal conditioning module outputs. It really depends on the source you are using to take your measurements. In general I would do what you did and split the difference and use a 50K resistor. You might try and vary the resistor values and see if that gives you a different offset. I only suggested a 10 mega-ohm resistor because that is the size of resistor that is used in some of our differential Thermocouple modules. 

You mentioned earlier that without the grounding resistor your measurement was off approximately 0.3 mV. I would say that the accuracy at that point is pretty good. If this offset that you are seeing is good enough for your program then I would say leave it at that. The amount of offset that you can expect depends on how long you have had your device without it being calibrated and a couple other factors. Depending on what software you are using to view the data (LabVIEW or Measurement and Automation Explore) you can set the setting for each input to differential, reference single ended, or non-reference single ended. Are you using LabVIEW for the measurements? 

If you have the input configured for differential then you should need no reference resistor. If you tell me a little more about your exact software setup it would shine some light on the situation.