Enlight 300W

[Spacedye69]

I KILLED IT!! A hard drive power plug was loose and shorted, killing the drive and the ps. I opened it up and the fuse was blown. Of course, replacing the fuse, it blew again. Nothing is burned, smells, or looks like it fried. There are small, round blue flat caps marked yc, Y caps? 3 of them and all three read inf resistance. Is this normal for Y caps? X caps read ok, as well as the diodes on the heat sinks. I'm looking for something simple before I pull the tranistors to check them. Seems +12v to gound. Any thoughts?

[linuxguru]

Re: Enlight 300W

Hmm, a short from either +12v or +5v to ground in the molex plug probably won't kill the drive, though a short from +12v to +5v probably can - but this is difficult, since they're on opposite sides of the connector. Are you sure the molex wasn't reversed and forced in?

Of course, if the connector was unplugged and came into contact with the underside of the drive electronics, various failures are possible. But if the fuse blows repeatedly, it's almost certainly a primary-side short, either the switching transistors and/or MOVs and/or bridge rectifier.

[yanz]

Re: Enlight 300W

i have a cheap psu doing the same symptomp. my friend (an electrician) found out that the culprit are those switching transistors. he said that the fuse isn't for protecting the psu from outside, but to protect the outside from the psu itself. now i know.

[Tom41]

Re: Enlight 300W

Similar thing happened to me; I was trying to power up a CD-ROM drive (old 2x speed drive) by plugging in the molex while the PSU was on. Unfortunately I was trying to insert the plug upside down, and as the pins briefly made contact 12v went into the 5v line - blowing the drive. The PSU survived, but the drive was completely dead. Probably a blown internal fuse, but it's very hard to get to on IDE CD-ROM drives.

Have you tried more than one replacement fuse? It's possible you tried a defective or wrong rated fuse when you replaced the one in the PSU, causing it to blow again. When replacing fuses, I need at least 3 replacements to blow straight away before concluding that the equipment is faulty

[starfury1]

Re: Enlight 300W

As to caps....low values should read infinite ohms(or there about depending on resistance range and meter)
The reality they charge up but too quick for any meter to catch

Those caps would probably fall into this category
(and probably as you said Y or X Which ever it is type caps)

Large Electro (as you probably know) will charge up but slower so you can see the charging action.
So at the beginning consider them a short circuit then as they charge to full they appear (ohms wise) infinite or open circuit.

Not the voice of experience here and guesstimate on my part.

If the psu is blowing mains fuses then it must be seeing a short circuit therefore I would follow Linuxgurus advice on it being a primary side failure and as yanz also said most likely the switching fet.

But check (bridge) rectifier diodes etc

Since part of the design of a switcher is to keep mains side electrically isolated from the secondary side (part of why they use HF transformer and opto-couplers)
There is a nice term for this but can't remember what it is.

I possibly wouldn't rule out a secondary side fault that killed the primary side because of the original short in the molex...
so would check what you can on that side too

If you have a fault on the secondary side it may well kill the primary side again.

Just my thoughts on it and not the expert on these things ...still on the learning curve ok

HTH

Yanz ...interesting way of looking at it "fuse blows to protect mains supply from evil of short circuit in psu " I like that

[Spacedye69]

Re: Enlight 300W

Transistor 2sc2625 - 2 of them both read 0 all directions on both resistance and diode - can they both be bad? They are on first heatsink.

[Spacedye69]

Re: Enlight 300W

Bridge rectifier test fine.

[Spacedye69]

Re: Enlight 300W

Replacements from Mouser are 12A and 100W instead of 10A 80W, but almost $9 each!

[Spacedye69]

Re: Enlight 300W

MOVs good.

[yanz]

Re: Enlight 300W

Yanz ...interesting way of looking at it "fuse blows to protect mains supply from evil of short circuit in psu " I like that

your line is better than mine

but almost $9 each!

holy sh**!

[starfury1]

Re: Enlight 300W

both...at $9 US I'd hate to think what I would pay here in oz for them.

possible I suppose you could be reading a coil or winding...don't know the circuit but in all directions very much doubt it.

I would guess both are cactus pull one and test, then the other
If they read zero out of circuit (check to be sure) then they are stuffed.....

some what expensive to replace, check as best you can for any other problems
you don't want to blow them again.

someone here with better experience on this will hopefully post.

I would hate to see the new ones get fried

Yanz .... thankx U get my PM about pole fuse ?

[Spacedye69]

Re: Enlight 300W

They were tested out of circuit. They are FRIED!!! I will be looking around to see if I can get some cheaper. Also, don't the switching tansistors run in parallel? That would explain why they are both dead.

[PeteS in CA]

Re: Enlight 300W

If the PWM is a TL494 or clone thereof, it's probably a half bridge inverter, which means that they take turns conducting. That said, if one transistor gets zorched without blowing the fuse, the other will quickly get zorched.

[linuxguru]

Re: Enlight 300W

>Transistor 2sc2625 - 2 of them both read 0 all directions on both resistance and diode - can they both be bad?

Usually when one transistor shorts in a half-bridge, the other will also get fried within a few milliseconds.

I just replaced a couple of 2SC2625 primary switchers with a pair of 2SD2498 horizontal-deflection transistors. The specs are slightly different, but it's fine for PSU use up to about 300W - 400W, half-bridge forward, with no modifications to the base-drive circuitry. The D2498 is widely available locally at very reasonable prices - I'd estimate median retail of $2 or lower, for the Toshiba OEM part.

The specs of the C2625 are 10A, Vceo 400 V. The replacement D2498 is 6A, Vceo 800 V - I believe it will be more durable in a half-bridge application, even with the lower current rating - Vcbo on the D2498 is 1500 V, which is quite a useful margin for turn-off spikes. hFE(min) seems to be 10 in both cases. The package is slightly different - the D2498 is fully plastic encapsulated and has a lower power dissipation rating of 50W, but that's perfectly fine for this switching application. It's also slightly taller and the position of the leads is offset by about a millimetre wrt to the C2625, so I had to bend the leads and do some mounting adjustments on the heatsink, but it worked out OK for me.

I'll doing an in-system test shortly, but I expect no problems. It's running cool as a cucumber with an external dummy load. The PSU is a generic 'PSA 12-300C', which is slightly better than a typical Deer '300 W' unit - this one looks like it's OK for 200W - 250W.

[Spacedye69]

Re: Enlight 300W

Thanks for the info. My wife should be having a baby any minute now. After that I can pick up where I left off.

[PeteS in CA]

Re: Enlight 300W

Another little Elmo? That's great! Congratulations.

[PedroDaGr8]

Re: Enlight 300W

His wife is having a baby and he is on here checking about his power supply AHHHHAHAHA

[PeteS in CA]

Re: Enlight 300W

When it comes time for delivery, she will kill him if he does within 10 feet (3 meters) of that computer, .