If you work with bead thermistors, you know that simply putting a bead thermistor in contact with a hot surface is not enough to get an accurate temperature measurement of the surface temperature. However, I have found through experiment that the relationship between thermistor temperature and surface temperature is linear. That is, given a thermistor in contact with a surface (whether taped down, pressured with help from the wires, etc) the thermistor temperature (calculated from resistance) appears to be a fixed fraction of the actual surface temperature, relative to room temperature.
Further experiment is necessary, namely with different geometries and further contact methods. But this means that measuring the temperature of a surface does not require special hardware, just a consistent mounting scheme of a cheap bead thermistor, and a bit of measurement and calibration. I have not found this sort of information from any sources on thermistors, yet, and am interested in hearing if this "contact fraction" is already known.
Anyone interested in the experimental methods is free to ask, I'll give more information, but for now I'll keep this brief. Given a thermistor with known Steinhart-Hart coefficients, held in contact with a surface at temperature T_S, and ambient temperature T_a, the thermistor temperature T_K will measure:
(T_K - T_a) = κ (T_S - T_a)
Where all temperatures are in Kelvin or Celsius, and κ is a constant coefficient unique to the method of making contact, surface and bead materials, and possibly ambient temperature, bead geometry, surface geometry, contact pressure, and thermistor leads length and wire gauge. The repeatability of κ and the variation of κ with the listed parameters is still to be determined, but high repeatability and low variation are expected, making this type of approximation a useful and computationally inexpensive alternative to specialized mounting techniques.
These are preliminary findings, from an unrelated project, so I'm interested to hear thoughts from others.
Further experiment is necessary, namely with different geometries and further contact methods. But this means that measuring the temperature of a surface does not require special hardware, just a consistent mounting scheme of a cheap bead thermistor, and a bit of measurement and calibration. I have not found this sort of information from any sources on thermistors, yet, and am interested in hearing if this "contact fraction" is already known.
Anyone interested in the experimental methods is free to ask, I'll give more information, but for now I'll keep this brief. Given a thermistor with known Steinhart-Hart coefficients, held in contact with a surface at temperature T_S, and ambient temperature T_a, the thermistor temperature T_K will measure:
(T_K - T_a) = κ (T_S - T_a)
Where all temperatures are in Kelvin or Celsius, and κ is a constant coefficient unique to the method of making contact, surface and bead materials, and possibly ambient temperature, bead geometry, surface geometry, contact pressure, and thermistor leads length and wire gauge. The repeatability of κ and the variation of κ with the listed parameters is still to be determined, but high repeatability and low variation are expected, making this type of approximation a useful and computationally inexpensive alternative to specialized mounting techniques.
These are preliminary findings, from an unrelated project, so I'm interested to hear thoughts from others.
