Power supply build quality pictorial DISCUSSION

[c_hegge]

It's a linkworld. Belongs in the Gutless PSU hall of shame.

[momaka]

Originally Posted by mariushm

Gonna test these two another day... might even work, who knows.

They most likely are. Probably shorted output cables. I had to fix a PA-10 for a friend before - the output cable was shorted right where it exits out of the adapter. Still works fine as far as I know after doing a little rewiring on the cable.

I'm surprised that Linkworld isn't a half-bridge design.

[Behemot]

Cabling is most comon problem on thee adapters at all. And if not cable, than connector is screwed up.

[mariushm]

Ok... I've tested both laptop chargers.

The Dell works fine, outputs 19.56v DC, no problem at all. I suspect the issue is the chip inside that IDs the adapter as Dell charger. As I don't have a Dell laptop, I'm not concerned about it and I've decided it's too nice of a charger (90w rated) to open it and investigate further.

The HP adapter didn't output any voltage at the connector so I opened it up. Very nicely shielded as you can see in the pictures.
The inside is also nice, only Nichicon caps... (edit: primary nichicon, 120uF/400v and secondaries 2 x 470uF probably 25v ucc/ncc something ) can't tell the transistors because it's covered in that white glue.
But anyway, AC in ok, over the primary cap there's 297v so that's reasonable.. at the solder pads where the cable goes there's 18.5v as it should be.
So I guess the connector is a mess. I'm waiting for the primary cap to discharge and then I'm probably going to solder a regular hdd connector to it and have this as a 18.5v power supply.

Pictures attached with the insides... click to zoom

11.jpg
12.jpg
13.jpg

[Pentium4]

So lately when I recap PSU's I've just been leaving the single cap for the -5V and -12V rail. If those caps eventually die can it cause issues for any of the other rails? Because I don't like to waste caps if I don't have to

[momaka]

^ well, if the caps become open-circuit or develop very high ESR, they can mess up the output voltage, which can in turn mess-up the regulation of the whole power supply. I usually leave those too, though, unless the PSU has been overheated or the caps are really crappy (I'm looking at you CapXon).
I actually did have a Bestec ATX-250-12Z drop out or regulation and turn off under load because the cap on the -12V rail was bad (a 1000uF 16V CapXon).

[Behemot]

Check this Super Flower Golden Green HX crap, especially the input filtration. Do you miss something in there?

Other than that it is Crapxon only but single Chemi-Con on +5 V SB. You see that filtration cap on the PWM Vcc? it is soldered to the Vcc legs directly. These things easily get 100+ °C! I had better opinion about SF

[stj]

it's very clean and well assembled - unlike antec!
as for soldering the cap on the chip - that will reduce the ESR and inductance

[Behemot]

You don't really miss anything in the filtration?

There's nothing clean about the assembling - several hand soldering repairs, many of them not still not nice, the flux is not cleaned and it's lackquered just over it. There is quite a quantity of flux and cleaning stuff residue, has really long way to go for „clean“. I found a solder ball in there.

As for the c(r)ap, the reduction is by no way worth the extra heat the cap will get and it WILL fail because of it as long as it's crapxon.

[stj]

i cant see *any* flux in that foto - it looks like it was washed with ipa.

[370forlife]

Theres no NTC Thermistor on the input?

[Behemot]

stj - zoom in, you can see it everywhere around the hand repairs, especially cable soldering (input and output) and the DC-DC daughterboard soldering. Not some huge quantities but still it is there. But they used isopropyl or some other alcohol, there is what seems to be a mixture of dissolved lackuering especially on the secondary, you see those light stains?

370forlife: there is NTC, on those long legs just next to the input cap.

OK I'll tell ya. Nobody sees not only missing varistor but especially bleeding resistors? This thing can be deadly…fortunatelly it discharges relativelly fast in this case, probably through the bridge rectifier into PFC, but still there are even positions for them, not used though…

[Behemot]

Found it! They first pointed me to current shunts but afterwards finally showed the right location…the bleedign resistor is under the primary heatsink Than they tried to told me Y caps are varistors, but anyway…

[momaka]

The only thing I see very wrong is a 10uF 50V CapXon right next to a heatsink (third picture, left center).
... and the fact that it's a CapXon cap.