I think there are three possibilities:
- the channel is faulty, as you are asking
- the injected current signals (0mA, 4mA, 20mA), when actually measured, are not as assumed (e.g. 0ma "injected" is actually 5.3mA at the module; 4mA "injected" is actually 9.06mA).
- Is crosstalk with another channel possible?
- Is the wiring correct?
- the values (2600, 10127, 32017) being read are not the engineering values from channel 13 e.g.
- they are from another channel
- they have been re-scaled
- etc.
Are you able to verify the injected signals i.e. with a DMM or similar?
Does anything change if the [Enable Hart] box is unchecked?
Where are you reading the 2600, 10127, and 32000 engineering values (you call them raw, but they are engineering values even if the represent the raw counts).