Shorting GPIO pins - harmful?

[kreisler]

What happens to the Pi3 (OS and or hardware), if one short-circuits pin 35 and 36, i.e. GPIO 19 and 16?

[rpdom]

It would depend on how they were configured at the time.

As long as at least one of them are set to Input mode, nothing bad will happen.

If both are set to Output (or one of the other modes where it may output a high value at some point) and one goes high and the other low, then you could kill the output circuit on those pins and maybe even a whole bank of GPIOS. The Pi may crash or restart due to the sudden short circuit between 3.3V and Ground, although that will be limited by the output circuitry to some extent. Worst case it forms an internal short circuit in the chip and it will rapidly overheat and die. RIP Pi.

[kreisler]

Thanks for the detailed answer. I am beginner, so let's assume that the Pi3 has a fresh installation of Raspbian Stretch with nothing further. We do a sudo update&upgrade and reboot.

Once the Pi3 has booted, i don't know which state those two GPIO pins have. What would your guess be? Any harm when shorting those two pins after the fresh installation+fresh boot?

[hippy]

It would be good practice to assume that any GPIO shorted together will at some point be in the worst possible state and will cause damage.

The solution is to ensure no damage can be caused by using a suitable current limiting resistor between the two GPIO. Using 1K should be appropriate ( and you could probably go higher ) and should not cause problems to however you are planning to use those connected GPIO.

[mahjongg]

after boot all GPIO's will act as inputs, except GPIO's used for I2C, and the TXD pin of the UART. but yes, deliberately shorting two GPIO's together isn't a good idea, simply use an 1K resistor in between them if you for some reason MUST connect them together

[Imperf3kt]

I'm thinking this is more a question of "I've accidentally done this, is there a hidden problem I cannot see" instead of "I'm planning to do this, is it a bad idea"

In short, have you caused an issue? Maybe. Likely? No
Should you be more careful and assume a worst case scenario next time: yes

[rpdom]

In short, have you caused an issue?

Was that a deliberate pun?

[Imperf3kt]

In short, have you caused an issue?

Was that a deliberate pun?

No lol. I didn't even notice until you mentioned it.

[Burngate]

Just as a matter of interest, https://www.scribd.com/doc/101830961/GPIO-Pads-Control2 gives the gory details.

On page 2 is the statement

If you drive the pad high and it is shorted to ground in due time it will blow up!
The same holds true if you connect it to 3V3 and drive it low.

Connecting two GPIOs together, driving one high and the other low, is (somewhat*) equivalent to both of those together.

*See below.
With one pad connected directly to ground or 3v3, only its impedance limits the current.
Connecting two pads together doubles the potential impedance, and so reduces the likely overload.
As always, drgeoff is more right than I.

[drgeoff]

Just as a matter of interest, https://www.scribd.com/doc/101830961/GPIO-Pads-Control2 gives the gory details.

On page 2 is the statement

If you drive the pad high and it is shorted to ground in due time it will blow up!
The same holds true if you connect it to 3V3 and drive it low.

Connecting two GPIOs together, driving one high and the other low, is equivalent to both of those together.

It is not equivalent.

A high output cannot supply the same current as the 3.3 volt rail and a low output cannot sink as much current as the ground rail.

Nevertheless there is still the risk of damage and it would be an unwise thing to do.

[boyoh]

Thanks for the detailed answer. I am beginner, so let's assume that the Pi3 has a fresh installation of Raspbian Stretch with nothing further. We do a sudo update&upgrade and reboot.

Once the Pi3 has booted, i don't know which state those two GPIO pins have. What would your guess be? Any harm when shorting those two pins after the fresh installation+fresh boot?

I think you will benefit from a short course in basic electronics, So the breadboard is your next step
with emphasis on using the multi meter ,when testing assume the circuit is alive and not dead ,do not
take things for granted, I worked on testing factory 11kv Ring main systems , if I took a cavalier
attitude to testing I would not be writing this post

[W. H. Heydt]

Thanks for the detailed answer. I am beginner, so let's assume that the Pi3 has a fresh installation of Raspbian Stretch with nothing further. We do a sudo update&upgrade and reboot.

Once the Pi3 has booted, i don't know which state those two GPIO pins have. What would your guess be? Any harm when shorting those two pins after the fresh installation+fresh boot?

I think you will benefit from a short course in basic electronics, So the breadboard is your next step
with emphasis on using the multi meter ,when testing assume the circuit is alive and not dead ,do not
take things for granted, I worked on testing factory 11kv Ring main systems , if I took a cavalier
attitude to testing I would not be writing this post

Yeah...11Kv would really make you light up. When I was a student, the basement of the EE building had a bunch of motor-generator (MG) sets. The biggest had a DC output side that was 1Kv at 980A. I think it was used to charge the 250KJ capacitor bank in the plasma experiment lab. The capacitor bank was charged in parallel and discharged in serial...

[Burngate]

Yeah...11Kv would really make you light up. When I was a student, the basement of the EE building had a bunch of motor-generator (MG) sets. The biggest had a DC output side that was 1Kv at 980A. I think it was used to charge the 250KJ capacitor bank in the plasma experiment lab. The capacitor bank was charged in parallel and discharged in serial...

When I was a kid* I was taken to the Science Museum, where they had a demonstration of a large spark. They may still have it, for all I know.
Some sort of high voltage DC source was fed through some resistances to charge some capacitors in parallel, and with spark-gaps arranged so that they discharged in series, through a rather large gap.
I hung around for ages waiting for it, but finally had to give up. Just as I reached the next floor there was a ginormous bang - I'd missed it!
Never saw it go, most disappointed.

*two hundred miles each way, so we didn't go often ... now it's only forty miles, we haven't been at all ...

[rpdom]

When I was a kid* I was taken to the Science Museum, where they had a demonstration of a large spark. They may still have it, for all I know.

I remember seeing it in action once. I looked for it last time I was there (about 10 years ago), but it seemed to have gone. Probably deemed too dangerous now.