Fault:

The readings on a 2100-M were showing incorrect readings, but have been working some time before going faulty.
No other recent site wiring changes have been made.

Cause:

The cause was tracked to a 0-10V input signal going into one of the inputs on the A16 that had an isolation or earthing issue.
Inputs on an A16 are not isolated from each other, thus if one has a problem it is possible it will affect all other inputs on the A16 to a greater or lesser degree.

Test Method:

Remove the 4-20 mA signal from the input on the A16 and replace with a 4-20 mA generator. This should now give a known good signal for all 16 inputs on the expander. If it does not, then the fault is related to one of the other inputs on the A16 interfering with the expander input. Remove the other inputs on the A16 one by one and see if the fault clears. More than likely to be 4-20 mA or 0-10V DC etc signals coming from other gear causing the issue.

The particular expander was -50/+50 so, 20 mA in would read 50 Deg C on the lines.

Additional Notes:

The stations Test dialog box could also be used to check the readings of the station, check also that the stations ambient sensor is reading a reasonable value as this can indicate a fault in the station.

Open circuit RTDs on 2100-A16 always read full scale (slightly over the range selected 0-100 Deg C will read approx 133 Deg C).
Open circuit RTDs on 2100-M normally read full scale (according to the range scaled for the line i.e 50 for -50 to +50 range) or zero scale reading depending on S4-3 dip switch position.

This fault has also been seen on a A16 populated with 15 RTDs and one 4-20 mA signal. When the 4-20mA signal was connected, all RTD input went to strange values. When the 4-20mA signal was removed, the RTDs read correct values. The solution is to add an 4-20 to 4-20 isolator to the mA signal to isolate any earth or grounding effects from the input.