Cap heating up

I have (had) a 12 VDC 500 mA wall wart from goodwill.

if I connect a 1000uF/35V cap to it, it heats up.

So I got a new cap, and it heats up, too. I measured the voltage of the wall wart at about 15 volts, with nothing connected to it. But just connecting the cap to the wall wart makes the new one heat, so would that mean the wall wart was letting AC current thru? The wall wart has been tossed, so I can't reproduce the problem, but I'd like to understand the problem. Connecting a cap to a DC source shouldn't do anything but charge it, as long as it's not under rated...
Thanks

As long as you connect it the right way round ... :roll_eyes:

As Paul__B hinted, putting an electrolytic backwards can quickly cause it to overheat and blow up.

It wasn't backwards. Neither of them.

If it was heating up with no load that's very suspicious - you wouldn't expect much ripple
current if that were the case.

Two things heat an electrolytic, reverse connection and high ripple-current.

Thanks. So likely the wall wart just went 'bad'.

Well then it too would have heated up too I think. Sure it wasn't backwards? Or under high load?

SouthernAtHeart:
It wasn’t backwards. Neither of them.

Frankly, it’s not worth the argument. :roll_eyes:

We simply do not believe you insofar as the original description.

The only way you can get a high ripple current is if something is actually discharging the capacitor at that same current. You simply refer to connecting the capacitor; if nothing else is connected, the capacitor charges to the peak and then no current at all passes.

If it indeed is a DC supply, then it cannot be passing AC unless the bridge rectifier fails in a peculiar fashion.

So some part of the description is clearly wrong. The fact that you disposed of it suggests that you knew that. In any case, the original question becomes nonsensical.

Sorry. No, the reason I disposed of it is because the first one was soldered on a board, which worked for a year. I know it wasn't backwards. Then it blew. Then another new one I put on heats up. Then I remove the board as being a variable and test just the cap on the wall wart.
I believe the wall wart my have diodes or something in it that went bad. I don't know.

One of the rectifier diodes must have been shorted.
That puts AC from the transformer directly on the cap.
Good that you have thrown it out.
Most of those (heavy/old) supplies are unregulated and can output close to 17volt unloaded.
Modern (universal mains voltage) switchmode plugpacks are stabilised and overcurrent protected.
No heat and regulated. Much better for electronics gear.
Leo..

Wawa:
One of the rectifier diodes must have been shorted.

I was careful to avoid that simplistic suggestion.

It is not really possible for one of the rectifier diodes to be shorted.

SouthernAtHeart:
I have (had) a 12 VDC 500 mA wall wart from goodwill.

if I connect a 1000uF/35V cap to it, it heats up.

Peak charging current for that combination could possibly have blown the rectifier

Interesting. The purpose of that this cap: my circuit is just an atmega328 chip controlling a relay, controlling a small 12 V DC motor. When the motor kicks on via the relay, it would blackout my 5VDC power circuit and my Atmega chip would reset. With a 1000uF Cap it fixed that problem. The motor only draws about 200mA, but the initial startup would reset my chip.

SouthernAtHeart:
Interesting. The purpose of that this cap: my circuit is just an atmega328 chip controlling a relay, controlling a small 12 V DC motor. When the motor kicks on via the relay, it would blackout my 5VDC power circuit and my Atmega chip would reset. With a 1000uF Cap it fixed that problem. The motor only draws about 200mA, but the initial startup would reset my chip.

Not surprising , the stall current of that motor is probably an amp or more which would easily depress your voltage rails.