PID Calibration ending with 0/5 cycles, Thermal Anomaly with Revo  

PID Calibration ending with 0/5 cycles, Thermal Anomaly with Revo

I'm trying to print on a MK3S+ with a Revo running 3.14.1 firmware. It has been throwing thermal anomalies whenever it exceeds ~100° C. I think something is going wrong during PID Calibration because it is not behaving how I would expect it to. I have 6 MK3S+ printers, 3 have Revo hotends, and this is the only one that is having thermal anomalies.

Here are the steps I took, recommended by E3D:
I reinstalled the firmware, ran reset_revo.gcode, and factory reset my machine, attempting to run PID and Thermal model calibration between each step. Then I changed my heatercore for a brand new one, downloaded the firmware again in case it was corrupted, and did all the steps again. Each time I ran the PID calibration, it displayed the following weird behavior.

It displays "PID cal. 0/5" on the bottom line.
The temperature rapidly increases, and displays the temperature as T°/210°, where T is the current temperature.
When the temperature exceeds ~100 ° C, it switches to displays the temperature as T°/0°, where T is the current temperature, and the temperature starts decreasing slowly.
When the temperature stabilizes around 25° C, the bottom line stops displaying "PID cal. 0/5", briefly displays "PID cal. finished" and changes to display "Prusa i3 MK3S+R OK".

The only time it did anything different during a PID calibration was right after the reset during the setup wizard. Instead, it displayed "TM: initial C est." while heating up, then once it got above ~100° C, it says "STOPPED." and then "Print aborted". Then in the menu I can clear Thermal Anomaly. This is the type of behavior I would expect all the time. When I try to run a thermal model calibration, it runs until it gets just over 100° C, then the text displayed changes to display "STOPPED." and then "Print aborted".

Clearly it isn't completing the PID calibration. It is not heating 5 times and it is not reaching the target temperature the one time it does heat. However, it does not indicate a failure. Is this how the calibration is supposed to respond if there's a thermal anomaly during calibration?

If changing the heatercore, firmware, and factory resetting the machine didn't fix it, what else could be causing this problem?