USB-6002 differential mode resulting in measurement drift

[hiphip]

Hi all!
I was trying to measure a voltage signal coming out of an AD8362 chip using USB-6002. Since the signal level changes subtlely I chose differential measurement mode. I used the DAQ assistant (I know it's bad to use express VI but the project just started so I'm just testing the equipment) and settings are continuous sampling, samples to read = 1k, rate(hz) = 1k. 

Here is a picture of the wires I use to connect the signal to usb-6002.

wires for measurement ad8362

the signal comes out of a sma cable. The red and black wires are positive and negative leads. The resistors are 47k ohms and are connected to usb-6002's AI GND to avoid ground loops.

Here is a screen shot of the signal I got:

orignal signal coming out of ad8362

as you can see the signal still contains noise.  So I tried taking the average of the signal every second and passing the signal through an 8 order bessel low pass filter with cut off frequency=0.01hz and then take the average every one second. 

after taking average

voltage signal through filter then taking average

Both plots shows a trend. There are a few big jumps due to me standing in front of the antenna(the ad8362 measures signal coming out of an antenna system). However, the signal level should be steady except for those few jumps since the whole system is placed in a room with air conditioner. Also no temperature jump can be bigger compared to the difference between the temperature of floor and my body temperature.

To make sure that the problem lies with not the antenna system but the measurement  system. I tried measuring the voltage coming out of an ADR431ARZ chip, which is a highly stable voltage reference that generates 2.5v voltage. I used similar wires for this measurement. The positive and negative lead, two 47k resistors to avoid ground loops.

wires for adr431 measurement

This time the original signal looks like this

adr431 measurement

A lot of huge spikes, which are not seen in the voltage signal measurement. 

I also plotted fft of the adr431 measurement

FFT

So these huge spikes come from the 50Hz AC power lines. But one thing to notice is that at 0 hz and signal level is not 2.5v. I haven't figured out the reason for this.

I also take the average and filter+average process.

adr431 taking average

adr431-filter-average

Only taking the average doesn't give much useful information. But the filter-average process also shows a moving trend, which is not supposed to happen.

My questions are:

Why does the signal drift in both measurement?

Why is the 50Hz interference avoided in the voltage signal measurement but not in the adr431 measurement?

[Nochenon233 · Accepted solution]

Hi hiphip,

It would be helpful if you could post the original logging data and the measurement setups of those chips (if possible).

There are some questions for you so we could help you troubleshoot more efficiently:

1. Have you tried plugging 6002 into different USB slot on your PC?

2. Have you tried reading the signal from other AI+/- input channels?

3. When was the last time you calibrated usb-6002?

4. Was the voltage input of AD8362 and ADR431ARZ stable during the measurements?

Best Regards,

Nochenon233

[Nochenon233]

Hi hiphip,

Have you carried out those procedures as I previously suggested? We'd like to hear any updates from you.

If the issues have already been resolved, It would be great if you can share your efforts as it might help others who encounter similar issues.

Best Regards.

[hiphip]

After a really long time I found that the problem was caused by other parts of the circuit but not the NI usb device. Sorry for the late response. Thank you for your time though