Noise on 9215

[NICO_MTS]

Hello,

I have a NI9215 but the first channel is very noisy compared to the 3 other inputs :

On a 5 V tension generator, i have a noise of about 50 mV for the first Analog input and less than 5 mV on the 3 others.

I don't understand why.

I wired only 1 channel a time and only AI+ and Ai- are pluged.

What can i do to fix that?

Thanks.

[Roohcifer]

Hi NICO_MTS,

To see if the noise is inherent to the card or a result of the sensor/wiring, try the following:

If you have the screw terminal version of the 9215, jumper AI0+ to AI0- and then jumper AI0- to COM.  If you have the BNC version, install a BNC terminator on the BNC connector, and tie the shield to the ground lug of your chassis.  If you're in a cDAQ, run a MAX test panel in DIFFERENTIAL mode and see if the noise in this configuration is any more prevalent than if you wire the other channels the same way and run the same test panel on those channels.  If you're in a cRIO, configure the module to run in DIFFERENTIAL mode from the project.  Either way, if you see that AI0 is still noisier than the other channels, please provide screenshots with a sufficiently long record (say, 1000 samples captured at 100 S/s) such that we can see the noise as a function of time.

As another note, just because one channel has slightly more noise than the others doesn't mean the device is operating out of spec.  Each channel has its own unique front end circuitry and ADC, so each channel could be subject to slightly different errors and thermal states, leading to different amounts of noise.  If one channel is not just noisier, but actually out of spec in the configuration described above, then it may be able to be sent in for repair.  If you post the screenshots and a picture of how you've wired the module to capture the data, we can help you determine if the device is or isn't in spec.

Regards,

[NICO_MTS]

Hello,

See the screenshot of what I have fron NIMAX after shunt between AI0+/AI0- and COM

What do you think about that?

[Roohcifer]

Hi NICO_MTS,

That looks to be a peak to peak noise of about [.0015 V - (-.0008 V)] = ~2.3 mV

The spec sheet claims an input noise of 7 LSB, or, taking the maximum input range on p. 21 of that same document, (10.6 V * 2)/(2^16)*7 = ~2.3 mV.

Your device looks to be operating within specifications.  You could expect to see up to this much noise on any of the channels, and the device would still be within specifications.

Regards,