Inaccurate current measurement using Shunt resistor

[Jake809]

Hello,

I am attempting to measure the output of a 4-20 mA pressure transducer using the PCI-6251 DAQ card and a 249 Ohm shunt resistor.  The measurements made by the DAQ card are slightly lower than expected.  Measurements made by a digital multimeter seem to be more accurate.  I have tried both configurations in the link provided below.  I have been reading about common-mode voltages and ground loops to attempt to solve this problem to no avail.  Any help would be much appreciated.

http://zone.ni.com/devzone/cda/tut/p/id/7077

[AnalogKid2DigitalMan]

Is the resistor a precison unit (=<1% tolerance)? Can you measure the resistor with the DMM to check its value?

[Jake809]

The resistor is not a precision unit, it is just what I had on hand for preliminary testing purposes.  However, I measured it's value to be about 245 Ohms, less than 2% off.  This accounts for some of the offset is not enough error to fully account for the innaccurate readings I am getting from the DAQ card.  I am expecting a current of 4 mA with a 0.996 V across the resistor and have achieved these values using a DMM.  I am seeing about 0.9 V with the card, though.  The transducer and DMM are both calibrated.  Thank you for your time and input.

-Jake

[Jake809]

I should add that I only achieved the proper values using the DMM when the DAQ card was not connected.  With the DAQ card attached, the DMM agrees with it.

[AnalogKid2DigitalMan]

It sounds like you may have a ground loop created when connecting to the DAQ card. Have you configured the DAQ for a differential measurement, and have you set the range to +/-5V for optimal resolution?

If that does not help, you may need full electrical isolation, companies like Dataforth and Analog Devices make 4-20mA isolators and converters with isolation (current to voltage conversion built in).

-AK2DM

[Jake809]

I have tried both of your suggestions, and am still having trouble.  I am doing my testing with the pressure transducer at atmospheric pressure, and have verified that the output is 4 mA.  After some more investigation I've found that only about 3.6 mA is flowing through the shunt resistor, other 0.4 mA is flowing into the DAQ card.  This is a significant amout of current lost, and seems to be what is skewing the measurements.  Thank you again for your help.

-Jake

[AnalogKid2DigitalMan]

Jake:

I am at a loss, hopefully NI or someone with more experience with such can assist. The spec for that card is 10Gohm input impedance with 100pa input bias current. I can't figure where that current would be flowing. Have you tried another analog input channel?

[Jake809]

I'm with you, it is acting unexpectedly.  I have tried other channels, hoping there was a problem with the one I was using, but the results are similar.  The hardware is not brand new, so I'm hoping that there is not a problem beyond repair.  Thanks again for your advise and punctuality.

[Wayne.C]

Jake809,

Can you post transducer make, model and detailed description of existing connections to DAQ?

[Jake809]

Sure, the transducer is a Viatran 349BSG with a range of 0 to 30,000psig.  It is brand new and properly calibrated.  However, I've made an attempt to simplify the problem.  Knowing that the DAQ card is only capable of reading voltage and simply measures current by using Ohm's Law, I have tried to take a voltage measurement from a battery I had lying around.  A 3.6V battery reading 3.67V with a DMM will show about 2.8 V in MAX using the DAQ card.  If measured by a DMM while connected to the DAQ card, it the DMM will show 2.8V, in agreement with the value obtained in MAX.  It seems like the DAQ card is altering the signal it recieves before measuring it.  Sorry, I've sort of changed the subject of this thread.  Hope you can still help me.

I am making meaurements with AI0 through an SCXI-1302 terminal block and feedthrough panel and have been trying both differential and ground referenced modes.  I've removed all other connections to the DAQ while doing this troubleshooting.

-Jake