Pumukli - 21-6-2024 at 07:10
Dear Mads,
I'm pondering about a possible molecular weight determination method, strictly home-made. :-)
Two cheap K-type thermocouples, both dipped in a thin layer of polymer resin and cured. One thermocouple is put into a small plastic centrifuge tube,
filled with distilled water. (Reference temperature.) Other one is dipped into a solution of unknown molecular weight substance, dissolved in
distilled water.
Both reference and unknown tubes are sitting on the cold side of a peltier cooler block. Cooling starts, reference freezes at some point (reference
point), unknown freezes a few tenths of a celsius "later". Molecular weight can be calculated from the temperature difference.
Did anyone tried something similar?
Has anyone have any experience with dealing with small signals from thermocouples and extracting meaningful data from them?
Would it be possible to achieve 1/100th of a Celsius (or better) resolution in such an arrangement?
Sulaiman - 21-6-2024 at 07:30
I think that a plastic tube is a poor choice due to it's low thermal conductivity,
any thickness difference between tubes will result in differing thermal time constants.
Rainwater - 21-6-2024 at 07:54
Freezing will be an issue aswell. As super cooling will alter the control runs. Working the other way around would be better(Melting)
Next is the thermocouple. For low temperatures and high resolution a thermocouple is not the easiest technology to use. Non-linearity can be corrected
but emi will kill your accuracy.
A ptc/ntc device would be better and much easier to design and get the level of performance needed without the problems created by radio rich
environments.(pt100/pt1000)
[Edited on 21-6-2024 by Rainwater]