My next thermocouple problem that I'm hoping to solve with an Arduino-shaped hammer...
I need to create an air gap in a thermocouple wire.
To be more particular, I have a thermocouple on a rotating shaft, and a recording instrument that isn't. Rather than using slip rings, I want to try and do it with an RF link between two Arduinos. One attached to the shaft (battery powered), one on the bedplate.
Ideally, I want the (existing) recording instrument to not know that there's anything abnormal - it wants K-type thermocouple voltage coming in on that channel, so I want to give it K-type voltages to record.
The rotating unit should be easy enough. Measure the thermocouple signal with a MAX31855 module (which takes care of cold junction correction etc.), and transmit a digital representation of it through a generic 2.4 GHz transceiver.
The stationary unit will receive the signal through a corresponding transceiver... but then I need to convert it back to a thermocouple voltage. That would involve a DAC of some sort... but how to accurately get my analog signal down to the low voltages (ballpark 0-30 mV) that my thermocouple recorder can understand and measure?
It seems what I need on the output side is an attenuator. Potentially a very simple circuit with just a couple of resistors, to step the 0-5V output of the DAC down to the 0-30 mV voltage my thermocouple recorder needs.
Practical considerations - how stable is a basic L-pad attentuator? If I solder the circuit together and calibrate the digital output in my Arduino sketch, is it likely to drift? I know standard resistors tend to be a bit unstable with temperature. This thing is going to be installed in a machinery space, near a big electric motor (we can talk RF shielding later), a dividing wall away from the hot things I'm measuring at up to 600°C. I expect ambient temperature in there to get up to about 70°C, which isn't crazy hot, but still pretty warm. And pretty cool too... cold starts in winter we'll be down below 10°C in the lab.
Do I need a more sophisticated attenuator circuit to help stabilise things? Just use precision resistors? Or are simple attenuators pretty robust anyway, and not worth worrying about at the level of accuracy I'm going to get out of a cheap DIY instrument like this?
The other thing I've come to realise is that I DON'T want my rotating transmitter unit to have any cold-junction compensation at all. The existing thermocouple recorder that I'm trying to fool wants to measure the voltage coming in, compare against it's OWN cold-junction temperature, and so calculate the hot junction temperature. Nobody cares about the temperature at the transceivers in this machinery space, so I don't want to use a proper purpose-designed thermocouple module there.
I just want a pure instrumentation amp and ADC to digitise the raw thermocouple voltage ready for transmission across the air gap, without correcting or interpreting that voltage in any way. I want to apply exactly the same voltage to the non-rotating thermocouple wire leading to the recording system. The fact that it represents a temperature is irrelevant at this point... it's just an analog signal.
Any suggestions as to a suitable instrumentation amp to use, to measure a 0-30 mV DC signal?