NI 9203 accuracy problem

[hrwang]

I am using NI9203 to measure a current. According to the manual, I can have 330 nA accuracy. However, I am far away from that accuracy when I set it up in a 9178 chasis, used a DAQ assistant. For example, I am trying to read a current of 6.268mA, the multimeter that connected to 9203 in serial shows a stable current out put of this value, meaning the module 9203 is fine. However, the reading from DAQ assistant shows only 3 stable digits, 6.26mA. It appears that the instrument control (wherever it is), or the chasis, or something else only give me a 8bit accuracy, although the module 9203 can have 16bit accuracy.

Any suggestions to solve this problem?

Thanks,

[Henrik_Volkers]

Try to sample the channel as high as possible (you can use the MAX test panel) and have a look at the waveform.

Multimeters usually avarage to present a stable (mean) value.

As a coarse guess: sample 300ms with at least 1k SPS and have a look at the mean value....

How stable is your current source?

[hrwang]

Thanks for reply. I am trying to read a balance of 5 digits from an analog port, it supposed to be stable. The multimeter I was using was a professional one that is supposed to be stable too. 

[Henrik_Volkers]

Again,

Open MAX (Measurement&Automation eXplorer)

Move to your card

open the testpanel

choose 100kSPS and 30000 samples (300ms)

Have a look at your incomming signal 🙂  (and hit Alt-Print , open paint, Ctrl-V , save as PNG and share your findings)  😉 

[hrwang]

Just contacted with NI tech support. It is ironic. The card has 16bits resolution, meaning you can read 1.2345 mA in from INDICATOR. However, the card only has the precision of 0.05mA. However, I couldn't find where 0.05mA comes from. I only saw from the data sheet of 0.04% unipolar accuracy. Anyway, the tech support has no idea how to improve this for now.

[hrwang]

test picture attached

[macncheese]

It seems that you have checked the specifications of the NI-9203, so it would be interesting if we can also take a look at the specifications of the multimeter. Do you know the sampling speed of that device? How many bits does the ADC have? So in that way we can check if the NI-9203 is working properly or not. Maybe we are dealing with noise here.

[Henrik_Volkers]

Your signal is quite noisy, the picture only show 10ms so I couldn't guess the line freq. but again:

Read 300ms (or 1s) of data and calculate the mean of that array.

I don't know your DAQ card. maybe it's possible to activate an internal low pass filter ... or you do it in software....

[hrwang]

Turned out to be a noise from the balance.

Thanks.

[paperduck]

I am having a similar problem reading current from balance into 9203 card.  Looks really noisy and kind of lost.  hrwang, how did you fix your problem?

Also, can someone elaborate on how to use the low pass filter?