Power supply for USB to SATA adapter smoked on first use.

How large a fuse would have to be in

The problem is, except for very special energy-limiting fuses, the fust does NOT limit the current before it blows. It allows whatever current the unit will draw for some amount of time before blowing. So, it is possible for even a 1A fuse to allow fairly thick circuit traces to burn through when the device has a dead short. It would not be unusual for a 1 A fuse to allow maybe 100 A in a dead short situation, for a few milliseconds.

Jon

I haven't done any UL certification in over two decades, but I expect that the test REQUIRES that the traces don't fuse before the real fuse.

It's widely suggested that chinese crap doesn't test for anything. They just slap on a fake sticker and sell it cheap. Cuz that's what we cheapskates want...cheap...

** The thin wire of a 1A fast fuse would normally melt and open well before a PCB track a few mm wide can do so. ** What happens with large overloads, like 100 times, is the fuse wire vaporises and initiates an arc from end to end of the fuse. Such arcs show negative resistance and so are very destructive - substantial copper tracks and wires vanish instantly until another fuse link or breaker trips.

Of course, special fuses types exist that can handle fault currents of thousands of amps without arcing - microwave ovens and Fluke DMMs use them, but not small appliances.

... Phil

The light bulb in series is a bad plan; the power supply will, in order to deliver regulated output power, draw higher current due to the lower input voltage, and fry the fuse, even if it isn't faulty.

Most common problem that makes the fuses fry: shorted diode in the AC rectifier (usually full bridge).

Yup, that's the REAL problem, FAKE UL and other certification stickers! Very scary. My mother in law got some surge suppressor strips some time ago. They had UL labels. If you turned them over, the plugs would fall out, even the FIRST time they were used. No grip to the contacts in the outlets. I sent the info on these to the UL comments address, and never heard back from them. I expected they'd WANT to know about fake labels, but apparently there are tons of these.

Jon

The light bulb will limit the current to around what the bulb is rated for. That is a 100 watt bulb will limit at about 1 amp and a 60 watt bulb to about 1/2 amp. At low current the bulb is like a very low resistance and as more current is drawn and the bulb lights up it acts like a higher resistance.

I was thinking of connecting the bulb without a load just to make sure there were no direct shorts before I plugged it in the very first time, not to use it like that in a real-world situation.

Thanks for your reply.

--
David Farber 
Los Osos, CA

Wouldn't all of the current have to go through the shorted diodes (1N4007) and shouldn't they show some sign of physical destruction? Other than being electrically shorted, they were physically unscathed.

Thanks for your reply.

--
David Farber 
Los Osos, CA

If a diode fails short circuit, it can be lower resistance than associated wiring or fuses (i.e. it cooks everything else in the circuit, but stays cool).

an

e r

er.

e

s

I have one exactly like that and of course I tested it when it arrived. It' s been working fine for several years running external hard drives for back ing up files. Before using the new one I would connect a 12V bulb to the ou tput as a load and see what happens when you apply power. Chinese quality c ontrol is almost non-existant, but when you get a good one, it seems to las t. What do you expect for less than 10 bucks?

At any price, I would not expect it to blowup the first time I use it! If I do use a light bulb, I will start at the other end in front of the AC power input where the short was.

--
David Farber 
Los Osos, CA