This happens with a goofy pcb layout. :P
Copper blown away between the large copper ball and the 000 part (jumper). There's also a segment blown off on the lower trace near the smaller Cu ball. Also a cleanly blown off segment to the right of the 000.
Trace detail under 000 part: V Trace to trace = 120VACrms 60Hz, line Gap ~0.7mm Trace width: 0.64mm I surge: complicated (Offline charging of 180uF through 6ohm surge thermistor.) I nominal:
Ooops.. Just found out what went wrong. Forgot to take off the scope ground before main power up :P This is a test board for an nonisolated offline convertor. My little traces become fuses that blew faster than the existing 2A slow blow fuse :)
D from BC British Columbia Canada.
You think PCB traces make good fuses? You know nothing. Semiconductor parts make FAR better fuses than either fuses or traces!
Maybe I could replace my 2Amp fuse with a 2N2222 :P
D from BC British Columbia Canada.
Essential for off-mains power supply work:
1) isolation transformer optional but very useful 2) HV diff probes 3) current probe Paul Mathews
Yeah I know. I just got socks and stuff for Christmas. Nothing practical like an isolation transformer :(
For now, I'm planning alternatives such as:
1) Making measurements off extra isolated windings on the power inductors. 2) Homebrew magnetic sniffer probe. 3) Using a Si8420 isolator to monitor digital signals. 4) Homebrew transformer probeOnly a matter of time till I crack and get the right equipment :)
D from BC British Columbia Canada.
Too bad you're not closer, I picked up a nice 3 amp medical-grade isolation transformer for 30$ in Montreal. It helps me feel safer when I use my Variac. (gloats) It would cost more than 30$ to ship it I think.
When I worked for English Electric many years ago I saw a power distribution cubicle that had been returned for repair with a copper bus bar about 3" by 0.5" thick that had melted. It must have had 1000s of amps through it.
Leon
Back when a computer display was a CRT, the joke was that a $300 CRT would protect a 10 cent fuse by blowing first...
The shipping cost for variacs and power isolation transformers (>100VA) often exceed the E*bay starting bid. :P
I tried to find an isolation smps on E*bay for less weight and less shipping cost. No luck yet. I suspect it's a rare item. It might be profitable if a designer sold isolation smps supplies (>100VA) on e*bay.
I might make a crude isolation smps to test the offline convertor.
Anybody know where to get used switchmode supplies with ~170VDC ~2A out?
D from BC British Columbia Canada.
Loss of the magic smoke is always a spectacular event.
That's the problem. Giving someone the web site catalogs for all the=20 electronics you want isn't really conducive to gift giving.
I wouldn't feel "too" bad. I once replaced a footswitch on a Pace vacuum solder extractor and the schematic "clearly" showed there was no mains on the footswitch pedal that controlled the vacuum pump. I'd had it with the crappy performance and the replacement footswitch we had ordered was sitting on the shelf begging to get installed....
So, sizing up the job, looked easy enough to swap out. The cord was way in the back, so I didn't bother to unplug it. (Like I said, schematic said there'd be no problem....)
Well, there was! A big "crack", a nice chip taken out of my new diagonals, and we're sitting in the screen room, alone, in the dark.... when Ed finally speaks up as says: "No, the other one!".
Smartass.!!
We finally found a 60-amp fuse at the hardware store and were back in business. Lesson learned: Schematics sometimes lie. And current paths often exist where you least expect them!
-mpm
What's all the blueish material about? Did the trimmer melt?
Ed
D from BC a écrit :
He, that's nothing.
On a fast power amp I designed we had a low inductance 1"x0.5" pcb shunt that had to handle more than 100A RMS (about 20W total dissipation). It was heat sinked to the liquid cooled heatsink through some sticky thermal gap pad. On the first production board, the prod people forgot to peal off the gap pad protection foil on one side. It took the test guys several boards to react: Pofff, pshhht.... Pofff, pshhht.... Pofff, pshhht.... Pofff, pshhht.... Pofff, pshhht....
The required nH inductance rendered the board unreparable.
At $7K the pofff-pshhht, it has been a pretty entertaining f*ck up :-)
Nope.. The trimmer didn't melt. The blue blob on the trimmer is left over glue gun glue. (Poor colour photography. It's supposed to be white.) The glue provided wire strain relief. The trimmer was disconnected (not used) before testing.
D from BC British Columbia Canada.
More than likely taht's from the flash on the camera.
Their let-through I^2 T is inversely proportional to their price.
This happened last week:
ftp://66.117.156.8/FriedBead.JPG
A 3-amp power ferrite bead fried somehow and took a bunch of trace with it. We can't figure out how this could have happened, unless maybe the customer did something weird.
John
Did anything else get whacked?
Have something to add? Share your thoughts — no account required.
Ask the community — no account required