unwanted analog input DC offset in NI USB-6251

[Jamieg]

Dear Forum,

I am using the differential analog input channel of my DAQ to acquire a signal output from an optical demodulator.
I have placed a 100kohm resistor between the DAQ's -ve and ground pins.

Usually, the acquired signal has no offset.  However, I have noticed that sometimes there appears to be a charge buildup and a DC offset appears on the analog input measurements.  The offset has gotten so bad on occasions that the analog input (when set on limits of +/-0.5V) has become saturated.  I have been able to discharge the charge by connecting the negative pin to a ground, or using a different analog input channel.

Can anyone provide advise about why this DC offset (charge buildup?) is occurring and how I may be able to prevent it.

Regards,
Jamie

[Mark_E]

Hi Jamieg,

Thanks for posting. I've got some questions so I can help troubleshoot the issue. Which channels/pins are showing this issue and which are not? How much of charge buildup and/or DC offset are you seeing? How quickly is this changing? Once I have your same setup I can try to simulate the problem here. Thanks.

[Jamieg]

Hi Mark,

I experienced the problem using the channel 0 analog input i.e. pin 1 to +ve and pin 2 to -ve.  I have placed a 100kohm resistor between pins 2 and pin 3 (ground).  Let me say however, that this problem occured after not using the DAQ for maybe 2 weeks.

When I shifted my wiring to the corresponding channel 1 analog inputs, then the offset went away.  I'm not sure the DC offset, but it must have been greater than +0.5V but less than +2.5V.

I would say that the DC offset happens between using the DAQ for a for weeks, as when I stop using it, it is fine.  But sometimes when I use it weeks later, the offset has occured.

Not sure if that's sufficient info for you?

Jamie

[Mark_E]

Hi ATMA,

Maybe you have more things in your setup that need to be considered before fully answering your question. Will you outline your entire setup so I can take all of that into account?

To successfully detect a short circuit on a switch you need two things:

1.     A known supply on one end of the switch

2.     An input line (analog or digital) to measure the voltage

If there’s a reason why you can’t use the supply on the DAQ board itself, use any external supply available. Again, if there’s more to consider in your application, please let me know.

Have a great weekend!

[Jamieg]

Was this an answer to someone's thread?
ATMA?
Jamie

[Mark_E]

Hi Jaime,

I apologize for that. I had written an answer for you, but apparently posted this answer, not yours.

When you don’t use the DAQ for a few weeks, are you leaving it plugged in and channels connected, or is it completely unplugged?

“I would say that the DC offset happens between using the DAQ for a for weeks, as when I stop using it, it is fine.” How is it fine after you ‘stop using it’? Does the DC offset happen only when you first use it again after letting it sit for a few weeks? Does the problem go away after using it for a specified amount of time?

What signal are you measuring? What frequency, amplitude, etc.? I’d like to duplicate this problem, but it sounds pretty sporadic. If you could also get this behavior to repeat and send me a screenshot and/or specific measurements, I could gain more insight as to what exactly is happening.

[Jamieg]

Hi Mark,

>> When you don’t use the DAQ for a few weeks, are you leaving it plugged in and channels connected, or is it completely unplugged?

The power is switched off and usually unplugged.  The analog input channels may or may not be left connected to the amplifier (which is off), but the wire connecting to the amplifier remains screwed in to the channels regardless.

>> “I would say that the DC offset happens between using the DAQ for a for weeks, as when I stop using it, it is fine.” How is it fine after you ‘stop using it’? Does the DC offset happen only when you first use it again after letting it sit for a few weeks? Does the problem go away after using it for a specified amount of time?

When acquiring my signal there is no DC offset.  Then I don't use it for several weeks, use it again, but can be a DC offset.
The problem does not seem to go away by itself.  It has gone away in the past after I have touched the channel pins to ground or said before, I rearrange the wiring.


>> What signal are you measuring? What frequency, amplitude, etc

Amplified EEG signals; acquisition rate ~ 2.5kHz; frequencies of interest < 200Hz; amplitude < 0.3V.


>> If you could also get this behavior to repeat and send me a screenshot and/or specific measurements, I could gain more insight as to what exactly is happening.

What occured last time was the signal was saturated at +0.5V (my max setting for the AI), however artifacts in the signal which are large in magnitude could be seen just below the 0.5V.  Upon increasing the max and min AI range to +/- 2.5V, the entire signal was present.  Attached is a time domain sample, mostly corrupted by 50Hz noise, with a clear offset.  The amplitude has been divided by 2100 to compensate for the amplifier gain.

Message Edited by Jamieg on 10-01-2007 11:00 PM

[Mark_E]

I hate to sound like a broken record, but I’m still not 100% clear on your setup. Let me explain what I understand is happening and you can correct me.

1.       You have your signal on pins 1 and 2 and are measuring away, happy as can be.

2.       Job done, hardware turned off and (usually) unplugged.

3.       Left for 2 weeks.

4.       Plugged in, turned on, ready to go.

5.       Measure signal and it has the DC offset. You touch the signal to ground (by putting a resistor between pins 2 and 3) to clear out the offset. You also change channels and there’s no offset.

Does this same offset happen when you look at a test panel in Measurement & Automation Explorer (MAX)? Do you leave the pull-down resistor in or is it only used to discharge the offset once? When you’ve tested on another channel do you leave it for a few weeks like you do with channel 0? I’m trying to determine if the problem lies in channel 0 or if something you’ve got hooked up to the USB-6251 is causing a charge buildup.

I would suggest looking at the DAQ Getting Started Guide for any other considerations neither of us has brought to the surface.

Let me know what the results are for some of these steps.