thermistor calibration
ok I understand that this topic is as ancient as 3d printing itself, in fact a thread surfaced years back
Now here is the motivations:
If you simply go to the "online flea markets" e.g. a simple search for a thermistor returns all of them claiming '3950' thermistors with *no specs*
https://www.aliexpress.com/w/wholesale-thermistor-3950.html
no doubt they are cheap, but my own experience suggest that I've once got a (hotend) thermistor that measures 50k at room temperature (25 deg C) for what is a 100k thermistor, even with all that 'beta' calculations, I can't get a 50k value for a 100k 3950 thermistor at that temperature, it is possibly off by a scale of 50% against freezing to boiling
https://www.ametherm.com/thermistor/ntc-thermistor-beta
Then comes the idea that if I can get an accurate reference, e.g. a semiconductor temperature sensor (some of them are pretty accurate) e.g. but not limited to
https://www.ti.com/lit/ds/symlink/lm35.pdf
I can probably calibrate those unspecified thermistors against the Steinhart–Hart equation
https://en.wikipedia.org/wiki/Steinhart%E2%80%93Hart_equation
Then comes the 'million dollar question':
supposing that calibrating room-100 deg C is feasible, and hence I can try to fit the values against Steinhart–Hart equation or Beta equation up to boiling point.
Now for the hotend thermistor, can we do *without* calibrating *hotend temperatures* which is more like 200 C - 250 C?
i.e. fit the Steinhart–Hart equation between room to boiling and use that equation for *hotend temperatures* that is *double* and higher than boiling point?
RE: thermistor calibration
Short answer is 'it depends', long answer is 'not really'.
If you search for `thermistor graph` you will get different results showing that thermistor response depending on the range is different.
In certain situations you may be lucky if it is linear in such a wide range of temperature.
Otherwise you are not lucky and get completely different readings and you can fry something (worst case towards disaster) or it will not heat enough ( worst case towards not causing harm).
See my GitHub and printables.com for some 3d stuff that you may like.
RE:
hi @_kaszpir_ ,
Thanks much for your response. I've been suspecting that the Steinhart–Hart equation
https://en.wikipedia.org/wiki/Steinhart%E2%80%93Hart_equation
fits differently at different temperatures, i.e. that it is not 'perfect' as in calibrating a thermistor for say 0-100 deg C does not imply the extended range 100-250 deg C works the same with the same parameters.
Unfortunately, calibration at high temperatures is difficult to do, up to boiling, a simple thing is to wrap the thremistor and the semiconductor sensor in aluminium foil possibly with some silicone thermal grease so that the probe and thermistor is practically at the same temperature. And simply heat to boiling and stirring, and record the temperature graph, this can be easily repeated for the cooling graph
But above 100 C, getting the 'right liquid' to do that calibration becomes difficult, maybe things like mineral oil works? And instead of a semiconductor thermal probe, it'd take like a thermocouple to go that distance.
RE: thermistor calibration
You can get multimeters with thermocouple sensors which are able to reach quite high temperatures.
Just make sure to get the compatible sensor to the multimeter so that they are calibrated.
See my GitHub and printables.com for some 3d stuff that you may like.