there is no prove the WDT will work when the MCU hangs for unexpected reason
The unexpected hangs are the reason why they invented the WDT!
--> So be sure the WDT will work "especially" on unexpected hangs.
& yes i have my experience with this earlier
For me personally: This is a "one man´s opinion without proof".
It is way more likely that what you "experienced" was due to a different mistake. Either schematic, PCB layout, wiring, soldering, WDT setup, timing (even in SW), measurement.. or simply wrong expectation.
I personally have never seen any evidence for WDT malfunction, nor have I heard of an according silicon malfunction. If there was ... you surely would see some press releases about it since in worst case this may cause a risk of life. WDT is a security feature ... it simply HAS to work.
lets say i have my first Pic12c508A working since 2008 with WDT enabled coded in assembly & it works until now
maybe the WDT reset it when something wrong happen or not but i trust it
in case of what it does not do (the WDT)i want an external solution to solve this
It is way more likely that the power supply voltage dropped ... and the microcontroller did not work at all. In this case one needs to add "default mechanisms" to activate a safe status.
Let´s say now in this case you add your external power down RESET circuitry. This may cause the circuit to power down.... but it never can auusre to go out ouf power down ... and make the microcontroller do usueful things. A microcontroller needs a specified minimum voltage to run.
How a microcontroller behaves on low voltage, on RESET, on WDT reset is defined. Which pins and registers default to a known state, which one default to an unknown state. Pin´s often become high impedance... resulting in an unpredictable state. It may change with time, with temperature from device to device.
And it´s the job of the designer to add techniques to get a predicatble state (often simply by using pull up / down resistors).
If the designer misses this ... no external circuitry will improve the situation.
Or let the reason be a bad solder junction in the XTAL circuitry of a microcontroller. No external circuitry can make the microcontroller to RUN safely. BUT the WDT will cause a RESET ... and do what the dataheet describes in that situation.
Or let there be an overvoltage (maybe ESD, maybe on the power supply, maybe on any pin) problem causing a destruction of the silicon in a way that the microcontroller can not work as specified. Then no external circuitry can cure the microcontroller. It will stay defective. Maybe with unpredictable result. But then the root cause is not the WDT malfunction.. it is a missing (ESD) protection .. or a defective power supply, that may oscillate at a random temperature.
--> the same behaviour would be if you have an external power down circuitry.
If I had the same problem ... I would find a proof, an evidence .. and not rely on my opinion. I´d search for the root cause.
I can not say what leads you to this opinion. But I´m very sure:
--> If there was/is a WDT problem on a PIC12C508A ... then you will find plenty of discussions about it in the internet. (I did not do a search)
Maybe affected batch numbers, maybe workarounds ...
Klaus