Yep, it looks like I'm that retarded. I was in the process of wiring up a USB plug up to my quad to power my GoPro. Well, I mistakenly hooked it up to 12V direct from the battery instead of regulated 5V and POP it went. A bit of smoke even came out the USB port.

Of course, after that it wouldn't turn on and got very hot if I plugged it into USB power.

Anyway, at this point I figure I've got nothing to lose so I've stripped down the camera to exactly what fried. I could only find one charred spot, on the middle PCB with the USB connector.

Looking closely at this however it looks like all the components are still there. The circle below C1 is just a locating hole for the lens assembly.

Closer inspection of all the boards reveal that this charred spot was likely caused by an exploding capacitor on the board mounted in front of it.

So I may try replacing this component and see how I go. Like I said, nothing to loose at this point. All the capacitors on this footprint seem to be the same value (100 6K, whatever that means), so I'll try that first. Anyone got a dead GoPro I can scavenge from?

Not doing too well this week, smashed my iPhone screen two days ago and now this. At least I know I can repair my iPhone. Anyone else done something this stupid?



Views: 13913

Reply to This

Replies to This Discussion

Google the SMD chip. I would solder the burn part out. Digi key carries a lot of chips and maybe the one you burnt out. Use alcohol to clean the burnt area and check traces. Everything else looks fine. 2 in SMD chips will save you couple of hundred.

Of course the cap blew, that's what happened when you overvolted it. Being that it's a filter cap, the circuit will likely work without it so take a soldering iron and remove what's left of the cap and power it up. Likely, you find the circuit works a normal without it. No need to worry about replacing it first for testing, clearly from the scenario, we want see if something else got damaged. This is just a bump in the road of that testing and in theory, the higher voltage and current would have burned what was left of the cap out until it was open circuit.

I've looked at your pictures and followed the traces, that is for sure just an input filter cap across the USB power plug (it's basically in parallel) so removing it would only affect the circuit from an noisy power supply and probably not do a lot anyway. Have you tried powering it via USB? In theory, the blown cap shorts the USB port shutting it down, but if it's open circuit, then the camera might work, and removing it will certainly let you test that theory. I'm pretty sure this thing might just fire up again.

Thanks for the suggestion Vernon. I gave that a go but no joy, nothing happened when I plugged it into power. I think the two capacitors at the top of this board may have also failed (C4 and C5, also marked with 100 6K).

They certainly got very hot, as the ribbon cable passing over them got a bit scorched. I've tested it and it still works.

I'm not sure what these capacitors are. "100 6K" makes no sense to me and doesn't seem to follow normal conventions.

Should be a 100pf rated for 6 volts.

That really looks like the cap blew, but IME the reason a part burns out like that is often because another part failed first and the resulting circuit fried the obviously damaged part.

It's worth replacing though.  You might try taking one of the others off and using it for a test.  Find one that is just a filter between the ground and a positive trace, those are non-essential.

just a filter between the ground and a positive trace, those are non-essential.

That's what this cap is!!!

I think you missed the point, he applied 12 volt input on the USB, right next to where that capacitor is being fed on the opposite side of the board with a through hole VIA. Careful tracing shows that cap is in parallel, thus the obvious reason why it was exposed to more than 6volts and blew. Thus he can take that shorted cap out of the circuit and plug it in USB to a correct 5 volts and it will likely work if no other voltage sensitive ICs blew or the overcurrent that happened when the cap blew and shorted could have burned a trace or often a small SMT indcutor that is in series with the power input but we don't see any evidence of that. The bottom line as a first step is to remove the bad device and power the circuit with a safe source such as USB which is current limited (most laptops/desktops shut power at 1 amp or less on a USB. Right now, with that bad cap which is most likely shorted in place, the USB port is shutting down on overcurrent. If he was lucky, sometimes windows tells you when a port is overloaded confirming my suspicion.


I didn't try too hard to trace what that cap was doing, but I see that he already tried the simple solution and it DIDN'T work.

So he can start tracing the path and replacing components or give up.  Otherwise, the only other trick I can think of is to remove the component and run some amps through the board and hope that the faulty component actually burns up so it can be identified.  If it was in series I would solder a piece of wire across it and hope the next thing that burned out was the bad component.  If it's in parallel then all you can do is remove it and apply amps and volts until something else burns up.  Obviously, that's a last ditch attempt.

He might also make sure that everything on the positive side of that cap is connected together and no traces burned up.

Thanks for your interest guys. It's a new day and I've done a bit more poking around.

I did remove the blown cap (C18), a continuity tester shows that this failed as a short circuit. Furthermore, the two at the top (C5 and C4) are also shorts, so I'll try removing them too.

Vernon, when I plugged this in OSX did throw up an error that it was drawing too much current.

Not sure if these are all filter caps, but their negative terminals all connect to ground. C18 looks like it's in series with a diode, C4 is connected to some kind of transistor, and C5 also runs from a diode.

To hunt down a shorted component, there is a technique somewhat less brutal and less at risk of causing further damage.  Trace the track and where it divides, put a small cut across it to isolate the sections.  Find the shorted section with a multimeter and keep doing that until you've found the shorted component.  To repair the cut track(s), scrape back a little of the green coating and put a blob of solder across the break.

You might be lucky and find that there's say a 5V to 3.3V regulator that has sacrificed itself to the greater cause.  However, if the 5V is applied to IC's along the way, then it might be curtains for the GoPro.

BTW, that blown component is a 10uF 6V tantalum electrolytic capacitor.  The 100 means 10 with zero noughts following.

Thanks Dave! I think I might actually have some spares of that value. Might have to shoehorn some through hole one on. Is there a difference between orange and black tantalum caps?

I went ahead a removed C4 and C5 (they looked a bit crispy). C5 had indeed failed as a short, it vented sideways which probably saved the ribbon cable above it.

Orange/Black - no difference just different manufacturers.  The orange ones on the board are different values, though.

As Vernon said, as a test, the GoPro should run to some extent without the caps, before you go to too much trouble replacing them.

It's definitely worth trying to save. Your only hope is if all those filter caps shorted out quick enough to shunt the evil 12 V. Don't power it up until you know what the damage is. Even if you can't save it, you will have learned valuable skills. Check teaces point to point with tour DVM in beep mode. You need to test diodes and transistors in diode test mode. In ohms mode a DVM will purposely test with voltage below the the threshold voltage of a semiconductor device(< ~ 0.6 V). When forward biased it will read the threshold voltage of the device. I would remove all shorted caps and then start poking around with the meter. Make sketches as you go. Start with D6. It looks like its after C 18. These things labeled Q# are transistors. Those things on the right side of the board with 5 pins are the voltage regulators. They can also be regulator/switches for tuning thins on/off. These will all be most likely MOSFET devices for the Q's. Try to get a part # and google it. Then you'll know what you're dealing with.

Holy crap! It powered on!

Might try and make a trip out to Jaycar today, looks like they stock 10uF tantalum caps with the correct footprint.

Reply to Discussion


© 2018   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service