NI PCI 6229 voltage problems

[FEIR]

Hello,

I have the NI 6229 card installed and working but I am recieving some very stange values on the different analog inputs.  Just to make sure Im doing this right I'll explain what I do and then the issue: Open Measurements and Automation program, go to Devices, then select NI-DAQmx, then the device NI PCI-6229 dev1.   Here I open the test panel and use the analog input tab.  At this point it doesnt matter which AI# I select, all will give me the same result.  I believe that I need to be using the RSE for the input configuration but again at this time it doesnt seem to matter because DIFF, RSE, or NRSE all produce the same result.  The mode I have it on which I need is On Demand.  Now at this time I have the card in the PC and the two cable attached to the back side of the card.  From here the other ends of those cables are just lying on the desk not connected to anything.  When I push start on the Test Panel, I will get a reading on the screen of between -5 to -7 volts and it ramps downward to exactly -10.6V and flattens out.  Sometimes I get a straight reading of exactly -10.6V without any ramping but its always that value.  At this time if I stop and change the AI# and or the Input Configuration I will still get the exact same result of -10.6V flat line on the screen.  Maybe I am mistaken but if nothing is connected to the other ends of those cables, shouldnt the reading on the screen be a flat line at 0V?  I have two SCB 68 boxes to hook up the wires to and even hooked in to those I get the same result.  I am really at a loss and thought that it should be at 0V when there is nothing plugged in to give it any feedback.  Any help would be much appreciated.

Thanks,

Ryan

[AnalogKid2DigitalMan]

Ryan:

What you are observing is normal behaviour. The very high input impedance of the analog inputs, along with the front end circuitry, allow them to build up voltages on the inputs when disconnected. They can build up enough charge to saturate to rails (neat +/-10V) I once built a 'person detector' by using an input with nothing connected to it other than a length of wire. When someone approached the wire, the way their body affected the induced 60Hz powerline field would trigger an alarm.

Try shorting the analog inputs to analog ground and you should recieve a zero reading. Or connect the channels to a battery for testing.

Happy sampling!

-AK2DM

[FrankMyAliasDoesExist]

Hi Everyone,

I have exactly the same problem. I tried connecting the ground directly, but it had no effect. Are there any other suggestions?

thank you, Frank

[aNIta B]

Hi Frank-

        How are you grounding out your channels?  Are you directly wiring your AI channels to COM?  As well, how are things set up in software?  Are you multiplexing, or reading one channel at a time?  Are you able to get a good reading from a known voltage, or is it only open channels that are railing?  Have you taken a look at the Field Wiring and Noise Considerations Guide, that explains how you should have your channels and signals grounded and configured to minimize noise and offset?  I would highly recommend taking a look at that document, and then let me know what other questions you have.  Have a good one!

[FrankMyAliasDoesExist]

Dear Anita,

Thank you for your reply. Indeed my wiring was not correct. I used differential input settings, assuming that I could use the Gain channel as a third wire. 

I'm a software engineer, an electronics engineer has helped me out with the right wiring, and my setup works fine now! 

Reason for me to believe it was a problem with the NI card, was that exactly the same setup worked perfecty with an older NI card (which apparently was configured differently by default)

Anyway, thanks for your help!

Frank