Technical question about "blown" hardware that starts working again

TL;DR: some dead hardware started working again a few weeks later, think I have elves in my closet repairing things.

Hoping someone can help me understand something. Over the past year I’ve had a few pieces of hardware I thought I’d killed, that later started working again. Having a hard time understanding how this happens.

Recently I was changing out CPU/mobo in my system. Ryzen 5 2600X and an ASUS ROG B450i ITX board. Doing custom cables, I accidentally reversed two of the wires on the CPU power cable (switched power and ground.) Tried to boot, I think I heard a small pop (could be my imagination), and it wouldn’t boot. Realized my mistake and fixed the cable. Tried to boot again and it was dead - different PSUs, cleared CMOS, everything I could think of, just dead.

About a week ago, on a whim I put the CPU and mobo back together to see if it would do anything, and it booted up just fine! There was one RGB dead, and the WiFi adapter isn’t working, but otherwise is fine.

I had a similar experience with another mobo several months ago. Is there an explanation for this?


Could be an electrolytic capacitor that blew shortening in the process. Over time, the electrolyte dries up becoming unconductive.
But just a theory

1 Like

So would that make it “dead,” but then once it become nonconductive it would work again? That would explain the delay and return to function, I guess.

1 Like

Its elves.

1 Like

Hobby electronics guy with physics background here. I would also point to that as a likely possibility.
Another strange thing that can happen with electrolytic capacitors is, that they sometimes can selfheal from reverse polarity (depending on electrolyte chemistry, type of defect, dipol charge present, alignment of Mars, Jupiter and Neptune, …)
There are also some protection circuits that are kind of slow to reset but this would be in the order of minutes at a maximum. (I think this is unlikely)
I would neet to poke around the board with a multimeter to be sure, and ideally have a schematic.

Yes, exactly.

Some semicunductor components can also fail in unpredictable ways but generally those fail in a thermal runaway kind of way and destroy themselves.
Another possible candidate is a solder pad/joint that just barely hangs on and produces intermittent working/ non working conditions.

@OldManGrimm Just a word of caution in case you intend to keep using the now working again motherboard. In case it was a filter cap that was the problem it is possible that increased ripple voltage will have a decremental effect on components that are connected to it. I would recommend to repair the damage if you can find it or only use components where you are okay that they will have a decreased lifespan.

How many times can something “die” but then come back to life?

I hate to be that guy, but … it depends.
When you reverse polarity on an elko the thin oxide layer gets destroyed by electrochemical reduction. There are several possible failure modes

  • The voltage breaks through but is removed fast (i.e. some kind of slow protection circuit) and there was no time for dendrites to grow or some other mechanism to permanently short the capacitor. In that case the chemical capacitance left in the dipol layer creates a forward voltage that can regrow the oxide layer. This mechanism can work several times but generally it degrades the capacity and voltage rating and at some point will fail in a short circuit mode.
  • The voltage breakes through and current increases until the cap vents or explodes. After this happens the short is generally removed by nature of the capacitor no longer beeing there. Depending on the function of the cap, the circuit can appear to work but ripple voltage gets really bad or it stops working outright.

  • From my experience the two points above cover most cases where electronics still “work” afterwards.
1 Like

Thank you!

1 Like

another riddle summed it up perfectly
these caps can be replaced using the same value and voltage or you can use a slightly larger voltage value as long as the capacitance is the same.
given that the repairs will work you should seek further into the cause of the failure in the first place.
its one of these cases where i don’t like loose ends that may bite you in the butt when your back is turned!