Heating Problem

[jamr]

Hello,

I have a raspberry pi that boots up just fine and connects to the wifi as well. All function is normal except that within minutes of booting up it's thermal warning shows up and I checked it's cpu temperature and that was at 79.5 C i.e. near the max. Shortly afterwards the pi shut itself down. I am unsure what is going on here.

Thank you

[LTolledo]

Its hard to decipher your problem due to lack of pertinent information...

Please provide:
a. a complete component list of your setup (include all accessories attached)
b. OS version
c. program running when the problem arose.
d. additional info that you think is related to the problem.

I suggest you do it in list form/ small paragraphs, not jumble everything in one big/long paragraph

[jamr]

a. This problem occurs whether there is anything attached to the Pi or not. I noticed the restart when there was the GPIO ribbon attached and I was trying to verify the PWM waveform from the Pi. Later I disconnected the ribbon and just used the hdmi cable to verify if the pi was booting up properly and it did only to flash a heating earning in 3 minutes and shutdown. the problem persists even if nothing is attached to the pi.
b.I am running raspbian stretch with desktop kernel version 4.14
c. It is a simple python script that I am using to communicate over BLE
d. it is a raspberry pi 3b+. I was using its wifi + ble when these problems arose. Other than the temp heating and pi shutdown I haven't specifically noticed any peculiar behaviour. I was unable to see the PWM on the oscilloscope when this problem arose.

[rpdom]

What hardware did you have connected to the GPIO?

[jamr]

The first time I noticed this problem, I was just verifying the 5 V output of the GPIO pin through a DMM

[gordon77]

Could you have shorted the 5v pin to any others?

[jamr]

The 5v pins still give the correct voltage. Would that not be affected if a short had occurred?

[jamr]

I was doing rigorous testing with the pi. It was on continuously for around 2 to 3 hours while I set things up. Could that have affected it?

[rpdom]

The 5v pins still give the correct voltage. Would that not be affected if a short had occurred?

Probably not. If you shorted the 5V to one of the GPIO pins it may have damaged the SoC internally, but the small current needed for that wouldn't affect the 5V connection.

[gordon77]

The 5v pins still give the correct voltage. Would that not be affected if a short had occurred?

Probably not. If you shorted the 5V to one of the GPIO pins it may have damaged the SoC internally, but the small current needed for that wouldn't affect the 5V connection.

Maybe pin 4 to pin 3?

[jamr]

Hmm is there another way to verify this? There is practically no voltage drop at the gpio.
If this is the case is there a way to fix it as well besides tossing it and getting a new board?

[rpdom]

Some of the gpio libraries have a pintest command (I can't remember which). They will tell you if the GPIOs have been damaged. If they have, then you may as well bin that Pi as it has been damaged internally and cannot be repaired.

[mahjongg]

Try powering the PI up with nothing connected to it, and NO SD_Card inserted, so that it cannot boot up.
if the SoC (metal covered chip) still heats up, then you have damaged the SoC.

My thinking is that if you think that a GPIO can output 5V,

I was just verifying the 5 V output of the GPIO pin through a DMM

then you are probably not aware that GPIO's are 3.3Volt ONLY and are NOT 5V tolerant, if you attach a greater than 4 Volt signal to any GPIO at any one time you can cause a latchup effect that will burn out internal silicon structures in the chip, which causes fatal damage to the RPI, causing it to heat up rapidly, even when no code is running.