6255 sample rate causes dc shift

[Tobin]

I'm seeing a dc shift in the analog input measurements as I select different sample rates.

I have the USB-6255 (mass termination) multifunction DAQ and I'm using Measurement and Automation Explorer to set up my inputs.

My raw analog input is -0.625 volts dc, I set up a task that uses 4 differential input channels with no custom scaling.

I've set the signal input range to +/-0.8v to make sure that I get good resolution.

Acquisition mode is continuous, samples to read is 1k, and I'm playing with the Rate in the range of 10kS/s to 50kS/s.

While this task is running in MAX, I can place my cursor in the Rate field and use the up and down keyboard arrows to alter the sample rate. As I do this, I can see the reported signal shift as much as 150mV from one rate to another.

This is a significant shift when the total input span is less than +/- 1v.

The direction of the shift is independent of the increase or decrease of sampling rate.

For example,

at 23kS/s, the reported value is -0.540v,

at 24kS/s it shifts to -0.620v,

at 25kS/s it shifts to -0.690v,

at 26kS/s it shifts back some to -0665v,

at 27kS/s it shifts back some to -0.625v,

and at 28kS/s it shifts back to -0.535v.

At first, I thought the sample rate change was altering the input impedance and changing the loading on my source, however, the entire time, my source dc signal remained at -0.625v (as measured with a fluke meter at the point of connection to the DAQ).

Why is this happening and what can I do to correct it? I want to give my users the ability to choose their desired sample rate.

My guess is that I will need to add a fixed gain amplifier with a gain of 5 to 10 in order to make the signal input use the max scale of the analog input (+/10v).

I'm using MAX version 5.0.0f1

Thanks for any help,

Tobin

[CertifiNATEtion · Accepted solution]

Hi Tobin, 

What are you using to generate the -.625 volt signal? If you are using a switching power supply, you may be experiencing aliasing from where the power supply is switching on and off. 

Additionally, are you seeing the same voltages at the same sampling rates? Do you always see -.540v at 23kS/s or does it vary over time?

Lastly, do you have a second 6255 that you can try to duplicate this on? It could be that the device is malfunctioning. 

-N

[Tobin]

Nathanael,

You hit the nail on the head, Thank you.

I am using an op-amp output of a 4-20mA signal conditioner as my signal. However, the circuit apparently has poor power supply decoupling. I'm seeing a synchronous high frequency ringing in the signal. When I set the signal to zero volt output, I see the dc shift as the sample rate is adjusted. However, if I disconnect my signal and short the differential inputs together, the shift vanishes and I get a stable zero volt report regardless of sample rate.

The values that I was seeing did not noticeably vary over time. They were repeatable at the specific sample rates.

The circuit uses a charge pump to generate the minus 5v supply. I'm sure this is where I need to put my efforts.

Thanks again,

Tobin