C245 Thermocouple calibration (not an issue, the forum doesn't work)

[alpemwarrior]

Hi there, I'm working on a controller of my own (nothing wrong with unisolder, just too complex for me :) ) and wanted to find what exact thermocouple type is on the C245 series cartridges by JBC. I've done quite a bit of research I finally think it's a so-called type M thermocouple (Ni/Mo - Ni/Co), because at the temperature these cartridges were calibrated at (I understand that would be 183ºC for 63/37 solder) this type of thermocouple has the exact same EMF output as unisolder would expect (using 43,5uV/ºC). It would also fit all the descriptions of the materials I've been able to find. Attached is a graph I made for reference.

What I wanted to ask, seeing that you worked with both pure lead and 63/37, is whether the TC output was observed to be strictly linear or if when calibrated at 183 °C the station displayed a temperature higher than real at 327,5 °C (melting lead).

I'm happy to close this issue if deemed inappropriate, thanks in advance.

[labjr61]

Hi there, I'm working on a controller of my own (nothing wrong with unisolder, just too complex for me :) ) and wanted to find what exact thermocouple type is on the C245 series cartridges by JBC. I've done quite a bit of research I finally think it's a so-called type M thermocouple (Ni/Mo - Ni/Co), because at the temperature these cartridges were calibrated at (I understand that would be 183ºC for 63/37 solder) this type of thermocouple has the exact same EMF output as unisolder would expect (using 43,5uV/ºC). It would also fit all the descriptions of the materials I've been able to find. Attached is a graph I made for reference.

What I wanted to ask, seeing that you worked with both pure lead and 63/37, is whether the TC output was observed to be strictly linear or if when calibrated at 183 °C the station displayed a temperature higher than real at 327,5 °C (melting lead).

I'm happy to close this issue if deemed inappropriate, thanks in advance.

Interesting. Let us know if you start a new thread elsewhere.

[sparkybg]

I work mostly with 63/37. Pure lead is just for confirmation that everything is OK. And yes, it appears to be reasonably linear, and when calibrated to 183C with 63/37 it melts pure lead, displaying exactly the temperature expected to melt pure lead.

But... I think you are wrong about the coefficient. 42.5 in the firmware means 42.5C/mV (degrees per millivolts), not microvolts per degree, so your assumption about the thermocouple is wrong. This would be a 1/42.5 = 23.5uV/C.

The thermocouple is a non-standart one. You won't find coefficients for it. What I did was to calculate the coefficient myself after many hous of experimenting and refining both the hardware and the software in order to exclude any offsets and other parasitics. This is where the guy (all my respects for the time spent on this), who made the "small" and "one channel" variants, failed - the amplifier has a gain of 750 and we are dealing with microvolts and microamps, and the layout is VERY important. So important that I had to make 5 major redesigns and maybe 10 minor adjustments to it in order to make parasitics as low as possible. So yes, it can be made smaller, but not by just playng a Sudoku game with the elements, but by thinking where the currents are flowing on the PCB it and what noises and offsets do these create in the signal path of the amplifier and current sources.

[alpemwarrior]

Dear lord am I dumb. Thanks a lot for the answer.