hmm ... trying to figure out how to measure current with one probe connected to VCC and the other to GND ...It could implement CC/CV and limit the current as well as voltage. This way shorting the power would never damage anything. It could implement this using shunt, amplifier and diode with some resistors, perhaps 40-50 cents more cost.You should never make any connection on/to the GPIO whilst powered.
How should the Pi prevent from short circuits on the external HW - especially some VCC to GND?
A cheaper way would be to have a polyfuse on the 3v3 pin for about 5 cents more, but in any case I will just put the polyfuse on my own board to achieve this. I am not really suggesting the pi zero should be changed, only trying to understand why this happened, and why the 5v got shorted to 3v3 when drawing too much 3v3 for several seconds which just seems strange.Yes I intend to put polyfuses on any wires that use 3v3 power that could ever potentially be shorted.You may want to add your own DC/DC to the external board, enabling the circuit via some GPIO pin (EN signal of your DC/DC chip). This will allow you to add whatever protection level you need and/or are willing to pay for (galvanic isolated I/O i.e.).I took an amp meter on a working pi zero, and put it on 3v3 and ground pins and measured 1.2 amps. I stopped just after getting a reading and it didnt damage the pi. I think it would supply even more power if not going through the meter. In any case, it is able to deliver in excess of an amp at least momentarily, but I would suggest that exceeding 500mA for any duration could damage things. What is unexpected is how it failed, shorting the 5v to 3v3 and then destroying the sd card.Btw..the DC/DC on a Pi Zero is not able to deliver anything close to 1.5A...
I am still wondering if this is normal, but not really interesting in frying lots of pi zero to experiment.
Statistics: Posted by aBUGSworstnightmare — Wed Feb 07, 2024 9:29 am