Help: Thermocouple measurement has nothing to do with the dial gauge
Thread Content
Thermocouple measurements generally involve using a multimeter to measure the mV value at both ends of the compensation wire, and then referring to a calibration table to determine the temperature corresponding to that mV value. . . . . However, at a certain factory, there’s a senior employee who says that the temperature can be calculated based on the measured mV value, without any reference to the calibration table. Moreover, here the temperature calculated by comparing my measured values with the calibration table differs significantly from what is displayed on the DCS. . At one point, I wondered if I had never even come into contact with thermocouples at all. I don’t understand what principle this is based on; when I asked others, they ignored me. The thermocouple here is of the K type with a range of 0–800. For example, I measured a mV value of 16.8, and after checking, the corresponding temperature was 409.5°C (without taking the cold junction temperature into account); however, the DCS displayed a temperature of over 540 degrees. . . . I’ve looked up a lot of information but still can’t figure out what’s going on.
The measured mV value is not subject to cold-junction compensation. Only after adding the mV value corresponding to the temperature at the measurement point can the actual measured temperature be determined by referring to the calibration table. This temperature should then be compared with the temperature displayed on the DCS. If there is a significant discrepancy, it’s necessary to investigate the cause—whether it’s a problem with the DCS display, the thermocouple, or the wiring, etc.
Based on your measurements, if the temperature at the measurement point is around 30°C, then the value displayed by the DCS can be considered accurate
Thank you, I’ve tried what you mentioned as well; even by adding a 30-degree millivolt value, the result is still around 437.8 degrees. The DCS shows around 540 degrees, resulting in a difference of about 102 degrees between the two. I checked the process data and the historical trend of the curves; there’s nothing wrong with the value of 540. As I also mentioned in the post, the mV values measured here can be calculated without the use of a scale; however, I don’t know which method to use, which is why I’m seeking help from everyone.
I’m sorry; I made a mistake and took 540°C for 440°C. A K-type thermocouple generates approximately 0.04 mV per °C, so it’s possible to make an approximate estimate without referring to the calibration table. By adding the mV value corresponding to the temperature at the measurement point, we get around 440°C. If there’s a difference of over 100 degrees, it’s necessary to investigate the cause. If you have standard instruments, using them for calibration will help identify where the problem lies. For accurately calculating the temperature using a thermocouple, it’s necessary to consult a calibration table. I believe what this veteran employee meant is that the calculated temperature has nothing to do with the calibration table; instead, it’s calculated based on 0.04 mV/°C. Generally, experienced maintenance technicians make judgments in this manner when troubleshooting on-site.
Have you mixed up K-type and N-type thermocouples? Nickel-chromium-nickel-silicon (K) and nickel-chromium-silicon-nickel-silicon (N) differ by just one character; if it is nickel-chromium-silicon-nickel-silicon (N), then 16.8 mV plus the cold junction temperature results in a value close to 540 degrees.
Thank you; I think your explanation is great. I’ve done the calculations and it’s pretty close. But I checked that both the thermocouples and the DCS are of the K type. . . .
That’s strange! 1. With a K division of 16.8 mv, 540 degrees can be displayed in DCS. 2. The temperature can be calculated based on the measured mV value; it has nothing to do with the scale. Both of these points deviate from the principle of thermocouples.
Yes. I’m completely confused too. . .
This post was last edited by xxkhc on 2022-8-7 at 10:51. Is the multimeter used by the original poster accurate for measuring thermoelectric potential? It is recommended to use a standard meter for measurement; otherwise, an inaccurate multimeter might confuse you. If there is no standard thermometer, the multimeter can be used to measure the temperature at other points. If the readings remain lower than those displayed on the DCS, it indicates that the multimeter is inaccurate.