Thermocouples vs RTDs is usually presented as a choice between two ways of measuring temperature, but they do not measure the same quantity at all. A thermocouple generates a small voltage that depends on the temperature difference between its measuring junction and wherever its wires terminate. An RTD changes resistance according to its own absolute temperature. That difference explains most of the practical behavior of each, including the errors people run into.
Why a thermocouple needs a second temperature
Join two dissimilar metals and a voltage appears across the pair that varies with temperature. Connect those two wires to an instrument and you have created more junctions, at the instrument’s terminals, which generate voltages of their own. The measured signal is therefore the difference between the sensing junction and the terminal junctions. To turn that into an absolute temperature, the instrument must independently know how warm its own terminals are and add that in. This is cold junction compensation, and it is done with a separate sensor built into the connector block. When the compensation is wrong, every reading shifts by the same amount, which is why a meter carried in from a cold vehicle can read several degrees off until it equalizes.
What follows from that mechanism
Several familiar thermocouple quirks are direct consequences. Extension wire must be made of the same alloys as the thermocouple, because substituting ordinary copper creates an unintended junction pair at the splice and inserts an error. Connectors are color-coded and keyed for the same reason. The output voltage is small, typically tens of microvolts per degree, so wiring picks up noise easily and long runs benefit from shielding and twisted pairs. On the positive side, the junction can be tiny, it needs no excitation current, it costs very little, and different alloy combinations cover an enormous temperature span reaching well beyond what most other sensors tolerate. That range is the main reason thermocouples dominate high-temperature work.
How a resistance sensor behaves differently
An RTD is a length of metal, usually platinum, whose resistance rises in a well-characterized way with temperature. The instrument passes a small excitation current and measures the resulting voltage. Because the relationship is smooth, repeatable, and closely modeled, resistance sensors generally offer better accuracy, better long-term stability, and much better interchangeability than thermocouples over the moderate temperature ranges where they work. No cold junction compensation is needed. The costs are a narrower usable range, a larger and more fragile element, a slower response because of that larger mass, and a higher price. Thermistors occupy a related niche with far higher sensitivity over a narrow band and a strongly nonlinear response.
The error sources that catch people out
For resistance sensors, lead resistance is the classic problem: the wires connecting the element are in series with it, and their resistance adds directly to the reading. That is why three-wire and four-wire connections exist, and why a two-wire RTD on a long run reads high. Self-heating is the second: the excitation current warms the element slightly, and an instrument using too much current will report a temperature the element created itself. For thermocouples, the equivalent traps are compensation errors, wrong extension wire, and degradation of the junction alloys after long exposure to heat, which shifts the output gradually and silently. Neither sensor announces any of these failures.
Choosing between them for a real job
Decide by range, accuracy, response and environment rather than by habit. Very high temperatures, tight spaces, fast response, or a need for cheap disposable sensors point to thermocouples. Moderate temperatures where accuracy and repeatability matter, such as process control and reference measurements, point to resistance sensors. Check what the instrument you already own supports, because a meter with a thermocouple input cannot read an RTD and vice versa. Whatever you choose, look up the stated accuracy of the sensor and the instrument separately and add them, since the combined uncertainty is what your reading actually carries. Where the measurement is made on or near energized equipment, treat probe placement as electrical work and use sensors rated for that contact.
What to check before you act
- Let the instrument reach room temperature before trusting a thermocouple reading.
- Use matching extension wire and connectors, never plain copper splices.
- Use three or four wires for any resistance sensor on a long run.
- Add the sensor and instrument accuracy figures rather than quoting one.
Common questions
Why does my thermocouple reading shift when the meter warms up?
The cold junction compensation sensor is inside the instrument. Until its terminals stabilize, the compensation is based on a temperature the terminals are not at yet.
Can I extend a thermocouple with ordinary wire?
No. A copper splice creates new junctions at a different temperature and introduces an error that no calibration can remove.
The practical takeaway
A thermocouple reports a difference and needs the instrument to know its own terminal temperature; a resistance sensor reports an absolute value and needs a wiring scheme that cancels lead resistance. Pick by range and accuracy, use the correct wire, and combine both accuracy specifications when you quote a result.
References and further reading
- CDC/NIOSH: hazards of working near energized equipment while instrumenting it
- NFPA: electrical safety research and education
- Department of Energy: interpreting equipment performance specifications
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
