>>> I also applied fresh thermal paste on the voltage regulator but that did nothing.
it wouldn't. If the output voltage started off at 12V and then decayed over time as the device heated up - then that could be a solution. However, the thermal paste shouldn't degrade unless it is disturbed.
I would disconnect the OUTPUT pin of the voltage regulator from the monitor circuit. See this website for details of the voltage regulator pinout:
https://www.componentsinfo.com/lm781...ut-equivalent/.
You may need to connect a resistor (say 1 kOhm) between the OUTPUT pin of the voltage regulator and 0V/GND/COMMON to act as a dummy load.
If you power the PET up in this configuration, measure the DC voltage on the OUTPUT pin of the voltage regulator relative to 0V/GND. If you observe 12V - then the regulator is working OK (well, at least it is regulating the voltage under no load conditions) and I would suspect a high current consumption fault within the monitor circuitry that we have to track down next. If the voltage is less than 12V, then the voltage regulator is most likely to be faulty.
You only need to desolder the OUTPUT pin of the voltage regulator and gently pull it back through the hole to release it from the PCB. However, you may find it easier to remove the voltage regulator IC completely from the PCB and then temporarily wire the INPUT and 0V/GND pins back in.
Of course, if you are going to remove the voltage regulator completely, you could test it on the bench by feeding a DC supply of between 15V and 22V between the INPUT and 0V/GND/COMMON pin and measuring the OUTPUT voltage.
If the voltage regulator is OK - then a replacement won't help and we can fault find from there.
Do not re-solder the OUTPUT pin of the voltage regulator back into the monitor circuit when you have finished this step if the voltage regulator is OK. I want you to take a current reading next if it is OK...
Dave