Input glitches on USB-6211

[MichaelWebster]

Greetings,

I'm having a bit of an issue on a USB-6211 with intermittent, regularly-spaced data glitches on a single-sampled input.
A little background: this is only happening on one of three machines that I've tested it on, a laptop running Vista32.  The machine that I developed the code on is a laptop running Vista64 and I've also tested it on an XP32 desktop, neither of which give me the aforementioned trouble. The two non-development machines I tested were running off just the installer that I created and had no previous traces of NI code on them.
I can replicate the problem using the 'test-panels' in MAX using a sampling rate as low as 100Hz.  At relatively regular intervals, the input spikes down to 0.00000 when measuring an input signal that is generally varying between 1 and 5 Volts.  I'm using the differential input on the daq and the input source is a floating (battery-powered) pulse-signal amplifier.

Has anyone seen data glitches like this before and/or have a solution to this problem?

Thanks for your time,
Mike Webster

[MichaelWebster]

I replicated the problem on a second XP laptop.  A little further exploration shows this to be 60Hz noise.  Changing the mode from differential to rse to nrse changed the noise characteristic but did not eliminate it.
I wound up hooking the input "ground" to the AI GND terminal as well as AI1- and the problem pretty much disappeared.  I'm still not real sure what the problem is/was since the noise is entirely absent on 2 of the 4 machines I've tested it on...
A little more testing shows the "problem" shows up even on the 2 "good" systems if using RSE or NRSE but goes away (or at least falls off to mV level vs +/- 10V swings) in differential mode.  With the ground wire connecting AI1- to AI GND, differential still looks okay and so does RSE, but NRSE has a small 60Hz rider (~200mV or so).

Anyway, any clues as to why differential seems to "work" on 2 systems and not the other 2 are welcome...

Mike

[johnsold]

Mike,

On some devices a bias current path to ground is required in differential mode. Perhaps the impedance to ground is very near the threshold and so works on some devices but not others. You mentioned that the signal source is floating so this may be a factor.

Can you check with a scope or voltmeter to see how large the common mode power line frequency signal is? It may be much larger on some systems than on others which could also account for the differences in behavior. Check to make sure that the common mode voltage rating of your devices is not exceeded.

Lynn