Hi I was wondering how you made out with this? I am looking to do the same thing. I have a 2 zone vinotemp, lower is running to cold, looking to replace the existing control panel with an arduino reusing the driver board. The built in controls maxes out at 66deg I would prefer 70 to use as humidor. In the event of a power off, the microcontroller also loses the temp setting ans defaults back to 54 which plays havoc with my humidity levels. hence desire to control myself. I am also planning on cutting out the middle divider to make one large compartment instead of two.
On mine the thermoresistor is on the inside, behind the fan cover, little black epoxied stub sticking out of the plastic, I have taken most of the cooler apart to trace wires, its wires do not goto the back driver board. there are 4 wires going into the pleitler, 2 power wires for element itself, and then two black ones that are for a defrost element.
The microcontroller for the temp control has a 3 wire plug connected to the driver board, red, black, and white. driver board will not run if disconnected, i think red/black is the power leads. reads 8v. When actively cooling, the white lead is at 7.5v, and when "idle" its at 4.5 and the fans run much slower. Connecting the power pin to white control pin does not trigger board to run.
Looking at the driver board, it has only 2 ICs. One is an dual opamp, and the other is a PWM power control chip. I am assuming at this point that the while lead is a PWM signal. I think it should work if I wire in a resistor between it and my USB logic analyzer which maxes out at 5v.
you can measure idle state voltages by putting a bag of ice on the thermoresistor. I am not really an electronics guy, just learning in the trenches for fun. Vinotemp also has some good manuals available in their support section. Edgestar has some very indepth service vids on youtube which are quite similar design.
If i can gain control of the driver board, I will add the temp control functionality to this project:
https://github.com/dzzie/humidor.net
does the assumption on the white lead being pwm sound reasonable? I am guessing the voltage drop is the analog appearance of a slower signal being sent?