Inaccuracy Question

[RBPLC]

If I'm measuring a temperature at 600 deg C using a type J TC, my inaccuracy is +/- 18.2 deg C. Am I interpreting this correctly?

[BobO]

If I'm measuring a temperature at 600 deg C using a type J TC, my inaccuracy is +/- 18.2 deg C. Am I interpreting this correctly?

Worst case, yes I think that's correct.

[RBPLC]

This seems really large.

[Garyhlucas]

In nearly every measurement repeatability is far more important than accuracy.  When I was an industrial electrician I would see people test a 1 hp motor with a clamp on digital ammeter and tell me the motor load was just fine.  The meter could measure 200 amps with +-1% accuracy so an accuracy range of 4 amps on a motor that draws 1.6 amps!  I would wind a ten turn coil and power the motor through that with the clamp on meter on the coil. So now the motor current was multiplied 10 times and that 1.6 amps should show up as 16 amps, and often be much more indicating that the motor was overloaded.

[johnr]

Specifications notwithstanding, we're observing much better accuracies in the ambient temperature range using type K thermocouples with the BX-04UT as compared to the BX-04THM. 

Our initial measurements with the BX-04UT are accurate to within a couple degrees near ambient, which is about the same as we get with our old Direcct Logic F0-04THM modules.  The BX-04THMs were reading about 10-15 degrees higher than expected in two separate setups, both with Type J and K TCs.  Our working range for these machines is 90-200 F, and accuracy within 3-4 degrees matters to us.  Maybe we needed to do something different with our wiring on the BX-04THM, but rather than dig any deeper in that regard we just decided yesterday to switch over to the UT instead.  Price is about the same and I can get rid of my filtering in ladder logic.