Problem with excitation of NI9237

[spalinowy]

Hello everyone.
I have an unusual problem with my cRIO 9022 and NI9237 card (strain gauge card). Well, some time ago the test bench that was operated using the cRIO underwent an upgrade. The last item was to connect the power supply to the control cabinet, which housed various amplifiers and controllers and the cRIO. Unfortunately, the electrician connected the power cable incorrectly, as he swapped the ground and phase wires in order. This caused me to have 230V at various points on the cRIO. Since then, I don't know why, but the NI 9237 card doesn't give me the correct values at the output.

So far I have performed:
- reinstalling the software on the cRIO
- NI9237 diagnostics via NI (no damage found)
- recompilation of the VI FPGA
- checking the operation of the other modules (all work properly: NI 9263, NI 9205, NI 9403, NI 9411, NI 9401)
- settings of NI 9237 card (external bridge power supply)
- I checked that the force sensors are not damaged

I should also add that the application was working very well before the whole incident.

I have no idea anymore what else to check.

[Henrik_Volkers]

I would go for another 9237 check.

I don't know how NI tested it, if they sourced the voltage input with a low impedance source they don't see leaking protection diodes.

Or in other words, by your description it seems reasonable for me, that the input protection circuit (diodes) had to carry some overload current and might be damaged. A typical damage is a higher leakage current, resulting in a usually non polarity symmetric  'lower input impedance'.

A simple test would be a bridge simulator. take four (10k or 100k) resistors build a bridge. connect it to your 9237. the resistors vary a little (Edit: maybe add a potentiometer  initial 1% resistors can add up with bad luck to a to high bridge voltage 😉 )  and you should read a bridge voltage. Now swap the bridge supply lines, you should read the same bridge volage with swapped sign.

[spalinowy]

1. Just for clarification still, can you tell me what would be the information that the card is working well vs. what to expect if the card is damaged?

2. As a test, can I use a strain gauge sensor just change its power supply polarity?

[Henrik_Volkers · Accepted solution]

Yes, using the gauge sensor is possible assuming a constant output.

If you change the supply polarity the reading should change polarity too but keep the same value.

A decent input sanity check would involve some more tests .. a half fried input protection circuit can cause a lot of 'funny' results (nonlinearities!)

Best check would be to try another 9237 😉  (In the past I would just call my lokal NI guy .. don't know if that would still work....)

Another point could be lost cable isolations ,

and finally it will be a loose contact 😄

[spalinowy]

However, I think the card is defective and you were right with the diagnostic method.
The value measured at power, as before, is 0.01 V, but as I changed the polarity -16.6 V appeared. For me, the value is quite high, because the bridge power supply is only 5 V. Nevertheless, it's time for a new NI 9237 card.

Unfortunately, to use the help of someone from NI locally is not possible, because NI has withdrawn from Poland from the business as it used to be.

If I could, I would have given you 3 KUDOs.