MAX31855 temperature differences
TC
Posts: 1,019
Hello all,
I am having some problems with temperature differences between 2 MAX31855. I bought 2 MAX31855PMB1 Peripheral Modules from DigiKey about a week ago, The Modules are nice because I know they would work first time. But I am having temperature differences between them. I am seeing a 1/2 deg C difference from one cold junction to another. And because of that I am getting about 1/4deg to 1/2deg difference on the hot end (thermocouple). Now maybe I am being to picky (been known to be), but I would like to have the MAX31855's as close to a same temperature as I can.
The modules are side to side, in open air, with no heat source around. And I am still seeing a difference. and just for experimentation, I placed a computer fan(12V) 4" away from the chips on the modules, just to make sure there was no stale air creating the difference.
My question to everyone is. What do you think would be the best option to try and make sure the modules are the same temperature?
And what ideas do you have about trying to keep the cold junction temperature at a content temperature for a DIY reflow oven?
One idea I had, was to have a PCB made up where its main purpose is to conduct heat from one module to another and try to equalize the temperatures. I was thinking taking a 2 layer board, having vias placed every .05" in a checkerboard pattern, and attaching the modules to the board using some 3M thermally conductive tape.
Then taking that setup, and placing it in a plastic project box that is lined with insulation, to try and keep the modules in a stable temperature environment.
I do not have the engineering background that some do, and I have no understanding of thermodynamics. But I am assuming someone on here does, and could help me.
Thanks
TC
I am having some problems with temperature differences between 2 MAX31855. I bought 2 MAX31855PMB1 Peripheral Modules from DigiKey about a week ago, The Modules are nice because I know they would work first time. But I am having temperature differences between them. I am seeing a 1/2 deg C difference from one cold junction to another. And because of that I am getting about 1/4deg to 1/2deg difference on the hot end (thermocouple). Now maybe I am being to picky (been known to be), but I would like to have the MAX31855's as close to a same temperature as I can.
The modules are side to side, in open air, with no heat source around. And I am still seeing a difference. and just for experimentation, I placed a computer fan(12V) 4" away from the chips on the modules, just to make sure there was no stale air creating the difference.
My question to everyone is. What do you think would be the best option to try and make sure the modules are the same temperature?
And what ideas do you have about trying to keep the cold junction temperature at a content temperature for a DIY reflow oven?
One idea I had, was to have a PCB made up where its main purpose is to conduct heat from one module to another and try to equalize the temperatures. I was thinking taking a 2 layer board, having vias placed every .05" in a checkerboard pattern, and attaching the modules to the board using some 3M thermally conductive tape.
Then taking that setup, and placing it in a plastic project box that is lined with insulation, to try and keep the modules in a stable temperature environment.
I do not have the engineering background that some do, and I have no understanding of thermodynamics. But I am assuming someone on here does, and could help me.
Thanks
TC
Comments
As Mike G said, the cold junction corrections should be done in software.
There are 2 calibrations:
First is for the cold junction.
Second is for the range.
Remember, the thermocouples also have an error value which shows up in the range calibration.
As to keeping the 2 chips at the same temperature:
Cut a piece of copper sheet and glue it across the tops of the chips to thermally bridge between them.
I have seen similar techniques done in commercial equipment.
Lastly, I should have chimed in earlier.
While thermocouples are very handy to use the use of RTDs, generally 100Ω for Pt100 types, are much more stable and accurate. Besides the sensed voltages are not in th uV range but in the mV to V range. Easier to measure and don't need cold junctions as they read the temperature directly. Often the system is cheaper to use. There are other resistance types available.
I have made crude wound copper wire sensors wound with very fine copper wire salvaged from automotive ignition coils. As I recall they were about 1000Ωs or so.
Homemade copper RTDs make nice thermocouple cold junction compensation sensors.
Geek mode on.
The electrical resistance of materials is closer to the fundamental physics than thermocouples which operate on secondary principals.
In general, resistance is closely linear with absolute temperature in °K, 0Ω at 0°K and on up to 100Ω at 0°C or 175.84Ω at 200°C for the Pt100 types.
Geek mode off.
Duane J
I like the idea of taking copper sheets to thermally bridge the chips together. I am assuming a copper bar stock (1/8" thick) would work too?
Duane J