eBay gits.

[CJay]

I buy lots of little electronics modules form eBay, very rarely if ever have an issue with them and they're almost always cheaper than I can even buy the parts, let alone spend the time to lay out and have a board made.

It makes it really simple and quick for me to try new chips (MPU-6050 IMU at the moment for a Quad Copter project), use utility parts (nice little TDA2030 amplifier modules or LM2596 SMPSU chips for instance) so I've got boxes of handy little boards I can grab when I need a bubblegum circuit to test something.

Last year I bought one of those little breadboard power supplies, looked like a great idea except it didn't fit my breadboard so it got consigned to the bits box, cheap enough not to worry about but too good to throw away, handy DC socket, two regulators (5V and 3.3V).

Today I needed a 3.3V regulator and just couldn't find one so I desoldered the regulator from that PSU* module, rigged it into the circuit I'm working on and...

SMOKE!!!

The damn thing is either faulty or fake and passes the full input voltage to the output pin (and yes, I got the pinout right), it's short circuit.

Tested the 5V regulator too, that's also short.

Lesson learned, test first and that particular board is in the bin after destroying a nice little Frontier Silicon DAB module.

[Spyke]

The thing about eBay is you really need to know what you're buying and be prepared to break out the soldering iron to rework it if it looks like production slipped up. And make sure to Test Test Test and Test some more! Apart from that the bits/modules are usually reliable once you get past the it works fine phase.

EDIT: Further have to add, a lab PSU with good current limiting would have caught this.

[bitslice]

Test any Chinese regulator at its rated load too, they might look OK just driving a multimeter, but then go ape when passing 1amp

[CJay]

Oh I know, I usually do test but this one had never been tried out because it didn't fit in the first place, just one of those things.

The only reason the lab supply wasn't used was because I was trying to debug a soft power circuit that feeds the regulator.

One of those things, chalked up to experience.

[rrinker]

 I must be the lucky one, at least for now. I have one of those little breadboard power supplies and it's worked fine thus far for its intended purpose, I used it to drive an Arduino Nano project test which had a 1602 LCD and a bunch of LEDs, too much to just power from the Arduino's regulator. This was all at the 5V setting, so maybe the 3.3V is no good, I haven't tried it. It's since been tossed in the drawer now that my bench is up and I can just use my bench supply.

[CJay]

Definitely got the correct pinout, but even if I hadn't, the 5V regulator was still on the original board and just passed whatever input voltage I used, measured short input to output.

[sleemanj]

What was the input voltage?  There are chinese produced AMS1117 which have a 10v input max, and yes, you exceed than and they fail short input to output (experimentally determined, in a similar manner as you found ;-)).

Attached datasheet of the 10v max input variety.

[HwAoRrDk]

Similar thing for me with a cheapo breadboard power supply board. 5V regulator was dead out of the box - input shorted straight to output. But at least the 3.3V worked fine. It was only a bundled freebie, though, so I didn't care much.

[CJay]

What was the input voltage?  There are chinese produced AMS1117 which have a 10v input max, and yes, you exceed than and they fail short input to output (experimentally determined, in a similar manner as you found ;-)).

Attached datasheet of the 10v max input variety.

Ah, that's nasty, of course the 'genuine' AMS parts will tolerate 15V and I would definitely have assumed these should too without checking but the 3.3 was in circuit being fed by 4 AA cells at 6V.

So, I suspect I may have plugged the breadboard module into my always handy 12V fixed supply at some point in the past (don't remember doing so but it's possible).

Sadly I can't check the logo on the parts as they're in a WEEE recycling bin somewhere now but from memory it was like two waves superimposed one on top of the other, starting left, cresting right which doesn't look too close to the datasheet one but I guess there are probably plenty of cheap AMS1117 clones out there.

Will definitely be more careful next time.