We commonly hear, “What’s the maximum thermocouple wire length I can use with your instrument?”
That’s a good question. There are two factors that affect TC cable length:
The longer the cable the more susceptible it is to radiated noise. The TC represents a low impedance source, so noise is mitigated to some extent. But very long lengths may require a shield to bring noise down to an acceptable level. Even so, many if not most thermocouple measurement instruments incorporate low-pass filters with very low cut-off frequencies (3 Hz) on their front ends that will eliminate all but the most insidious noise. However, such interference can produce an accuracy-killing offset that may be difficult to detect, and some detective work may be in order. Make ambient temperature readings using a short length of thermocouple wire, and a second with a longer one. Any difference you see may be caused by noise, the increased resistance of the longer cable length, or both.
Inaccuracy due to cable length is caused by the voltage divider effect produced by a combination of the input impedance of the amplifier and the source resistance of the TC. A first order approximation for inaccuracy uses this equation:
% inaccuracy = 1-[Ra/(Ra+Rt)] x 100
Ra = Amplifier input resistance
Rt = Thermocouple wire resistance (total of both wires)
As the equation implies, inaccuracy is inversely proportional to amplifier input resistance. This is why you should always strive for high input resistance in the amplifier you choose to make a thermocouple measurement. DATAQ Instruments model DI-1000TC thermocouple data acquisition system has 1M Ohm input impedance, while 8B47 thermocouple amplifiers have 50M Ohm impedance. For example, using 1000 ft of 20 gauge, J-type TC wire would produce the following errors for each instrument:
1-[1M/(1M+357)] x 100 = -0.036%
1-[50M/(50M+357)] x 100 = -0.00071%
In power plants frequently there is mismatch in thermocouples reading located at same place.May be it is an effect of noise.