There is no better way to go about finding troubles than to use plain common sense. One should ask if system performance seems to be reasonable for the conditions?
Do changes in the controls produce a logical result? What about the product? Does its condition correspond with what the instruments are saying?
How to Test a Used Thermocouple
First of all, one can not ordinarily remove a suspect thermocouple from service and ‘test’ it in another place. The very fact that the device is used means that it may no longer be homogeneous. Subjecting an inhomogeneous thermocouple to a different set of temperature gradients, even if only subtly different, can result in a different output and reading. So, ‘recalibrating’ a used thermocouple will certainly yield a ‘number’, but that number will probably be meaningless in the thermocouple’s place of use.
Perhaps the best way to evaluate a used thermocouple is to ‘probe’ the location by placing a new thermocouple that has a known output alongside the suspect one in an operating process and compare the readings. If it is not practical to have two sensors in place at the same time, the next best thing is to remove the suspect probe and replace it with another one known to be good. Then, as long as the good probe is located in the same place as the removed one had been, and the process has not changed during the exchange, the readings from the two probes can be compared.
Note that it is not necessary to keep and use an unlimited supply of new probes for these tests. A few suitable replacement devices can be kept on hand and one of them selected for testing use. Under normal circumstances, thermocouple drift, or degradation, is a gradual, and very slow process. A single replacement probe can thus be used a number of times to probe a process before it accumulates very much time at operating temperature, and can usually be considered reliable for a number of such repeated tests. And, when a drifted probe has been found, the test probe may simply be left in place as a working sensor, while the next replacement becomes the test device.
One very useful instrument for troubleshooting thermocouple systems is a portable temperature indicator. A number of these devices are capable of operating with two or more different thermocouple types, and some offer an ‘output’ function that will produce an electrical output to simulate a thermocouple operating at any temperature of choice.
In use, the instrument is normally attached to the wires of a circuit being tested at some convenient access point, such as in a connecting head. Care should be taken to ensure that the correct polarity is maintained. In Australia we use the ANSI colour code where the negative is always RED. There, the output of an operating sensor may be monitored and evaluated. Or, using and instrument’s ‘output’ function, a simulated thermocouple signal may be sent back to the circuit’s permanent indicator or controller to verify proper operation of the rest of the circuit. When driving a signal back towards an instrument, it is usually necessary to break one side of the circuit to avoid ‘loading’ the portable tester by the low resistance of the thermocouple itself.
Sections of extension wiring in thermocouple circuits may be checked for proper connections with a portable tester, as well. The section being tested should be electrically isolated from the rest of the loop, and one end of an extension wire pair should be shorted together. If a tester is connected to the opposite end of the shorted pair, the tester should indicate the approximate temperature of the shorted end. Note that if both ends of the extension pair happen to be at the same temperature, it may be necessary to warm the shorted end a little and verify that the tester ‘sees’ the temperature change correctly. Note that is the possibility of an incorrect, reversed connection that is being checked in this test.