drifting DC in acquisition

[CMX]

I have nasty problem with drifting DC. I'm running NI PCI-6229 with desktop PC and NIDAQmx drivers. When I test the card and input channels with Measurement and Automation programs test panels, the signals look as they should and DC is stable. When I use the same channel and the same source but acquire signal continuously with AO/AI synchronization example VI that comes with labview, the DC is drifting. Also the DC levels of different channels seem to be different. Any ideas what could cause this very weird behavior?

Thanks

[CMX]

Forgot to mention that I'm using BNC-2090 breakout box - all the channels are SE and in NRSE mode.

[johnsold]

Drifitng DC usually means that something is not hooked up correctly. Check to make sure that everything is connected according to the manual. Look especially at grounds/returns on all sources. Terminate any unused inputs. They can float to the power supply level and create wierd effects on other channels.

Lynn

[CMX]

Thanks Lynn,

It should be connected properly but i haven't put BNC caps over the unused inputs. I measured that they should all be well connected to ground and that there's no significant voltage differences between them - do you think that this might still cause the problem? I also have SBC-68 breakout box connected to other terminal of the card with similar ground settings and it's working fine, though the inputs are wired directly under the screw terminals in contrast to BNC in BNC-2090.

The reason why I'm wondering possible problems in the NIDAQmx Labview interface is that these channels look good while using the measurement and automation tool test panels. So I'm wondering that maybe the Labview function that generates tasks from the physical channel input or the later acquisition functions are doing something weird.

[Mcdan]

CMX,

In your LabVIEW application, do you explicitly set the terminal configuration? If not, DAQmx will default to differential terminal configuration. The result of this would appear to be floating DC signals if you are wired for RSE or NRSE mode. The terminal configuration can be set on the DAQmx Create Channel VI, or accessed through the channel property node.

Hope this helps,
Dan

[CMX]

Actually coming to think about that, no. I started with creating all the necessary channels with measurement and automation program configuring them correctly because I planned to use these global channels instead of physical channel variable that I ended up using. I haven't configured the measurement mode separately for the task generating function so it surely is operating in the diff mode. This with the floating voltages of the unconnected inputs Lynn mentioned earlier would certainly explain what I was getting. Sounds like a good explanation - let's hope that does the trick when I test it tomorrow.

Thanks a lot. It's surprising how hard it's sometimes see the obvious problem youself...