A couple of weeks ago I powered my Uno from a dodgy wall wart. The 12V wall-wart is ok once turned on but on initial power-on it spikes the DC output to god-knows-what...but this killed my Uno via the barrel connector. What happens now is that I get a few seconds of power on (either barrel connector or USB same), flashing LED's and PC connection confirmation but then it shuts down. When connecting via USB I'm pretty sure the polyfuse is kicking in.
Can anyone have a stab at what components might have fried and any method to check these to confirm...this is just a small 'project' now as not really expecting to bring back to life!
Are you sure it's dead.
12volt on the DC jack and extra load on the 5volt rail (sensors, relays, etc.) could overheat the regulator and make it shut down within seconds.
500mA will also trip the polyfuse.
So what's hanging on the 5volt (or 3.3volt) line.
Leo..
The polyfuse coming on suggests one of the chips is either melted internally (regulator perhaps?) or
going into CMOS-latch-up (one of the logic chips). Whatever it is will probably be getting warm before
the fuse kicks in.
@Wawa: It's certainly not happy! When I say it shuts down I mean with nothing (at all) hooked to it..bare board.
@Mark T & DrAzzy: Agree, something is quickly drawing current (and possibly overheating) but it all happens a bit too quickly to be able to feel temps or even get a meter onto components. I've ordered some 1117's and will look into that lm358 - cheers DrAzzy.
In case you were not aware of it, some wall warts (the analog , non-switching type, are load dependent output voltages. This means that if it is rated for 12V @ 1A, the loaded output voltage is 12V (at 1A). If the load is a 500 mA load, the output would be higher, like 14 or 15 V. An UNO only draws 55 mA. If you connected a non-switching wall wart to the barreljack, you could damage the 5V regulator and subject the entire board to 14 V.
EVEN if you HAD a switching wall wart, it would NOT be a good idea to use a 12V wall wart for the input to a 5V regulator because you force the regulator to disipate the difference between what it needs (7V) and what you connected (12V), so 12-7=5V that must be dissipated as heat, causing it to run hot. You can do it for a short time but not a good idea as a regular thing.
There is absolutely a polyfuse on the Arduino board. It's the green square resistor shaped thing on clones (brown on real ones), bigger than a resistor - near the USB port. Some clones don't have one.
There is also current limiting of some sort on the USB port - I don't know if they use a PTC (if you look at most USB host and hub chips, they have a way to turn off the ports if there is an overcurrent event, and the app notes don't show PTCs). Different PC's seem to have different levels of grace in handling this situation, though.
I'm glad I never knew about "magic smoke" when I was teaching myself electronics 35 years ago. I probably wouldn't have tried anything with all that anxiety. It would have tied my hands.
I've seen components smoke but there was no magic... What is scarry is a TO92 transistor going ballistic with a loud BANG ! It blew a little crator out of the flat face and you could see the metal junction at the bottom of the crator. No smoke though. You could hear the little tiny piece bounce off two walls. It sounded like a BB. Never did find the piece but I'm pretty sure you would have felt it if it had hit your face. It was someone else's project and I was too busy to investigate so I never found out why it blew. I did read the "2N3906" on the package though. Maybe it was supposed to be a "2N3904".
When I say it shuts down I mean with nothing (at all) hooked to it..bare board.