Power supply build quality pictorial. part 2

[momaka]

Re: Power supply build quality pictorial. part 2

The smallest coil is shorted by PCB.
Here, have some pics that show it better:

Okay, now that ^ is weird. Even if that torroid coil wasn't shorted, I still don't see what purpose it serves as one of its sides would still be connected to ground.
This PSU is pure

Sometimes, it really does help to look at things on a 22" screen and not a tablet.

I have no problems on a 19" CRT .

[370forlife]

Re: Power supply build quality pictorial. part 2

Yes, I agree about that.

I am refering to the connector they actually put after the 4 diodes:

https://www.badcaps.net/forum/attach...1&d=1388768415

See? It doesn't make any sense. Whatever you put there (PFC coil, jumper, etc) it is permanently shorted and out of the circuit. I think they really messed up with their designs. I can't find another reasonable explanation.

So they can plug a fake PPFC coil into it that doesn't even need to have a real loop of wire. They just saved 1/2 a cent by doing that per unit.

[Pentium4]

Re: Power supply build quality pictorial. part 2

Here, we have a typical well built Macron. First one I've seen with PFC, a non Power Logic fan (although I can't complain about a ball bearing Globe Fan), and without a proper AC plug.

The AC receptacle starts off with four Y caps, two 0.47uF X caps, and a coil. It then continues on the PCB with another coil, 0.47uF X cap, and one MOV.

The bridge rectifier has no heatsink but is rated for 12A

The PFC circuit uses two ST 23NM50N MOSFET's in the TO-247 package, and a LiteON LTTH1506D PFC diode. They're controlled by a Fairchild SG6932. Unfortunately, Fuhjyyu @85C for the PFC cap…

Double forward topology, uses two ST 23NM50N FET's (Same as PFC) rated for 17A, 500V

Unfortunately it still uses two transistor for 5VSB, but at least it uses a 4A MOSFET, Unisonic 4N60

Transformer is good sized…I can't see any of the secondary silicon specs without disassembling, except for the SBR30A40CT 30 Amp “Super Barrier Rectifer” for 3.3V. I can see that it has two 12V rails each with an OCP shunt.

Supervisory chip is a Weltrend WT7525

All the caps are Fuhjyyu except for that lonely Teapo on the 12V (SC series, 2200uF 16V on CPU 12V rail)

If the temperature was kept down I bet it could do 550W depending on the 12V silicon, I'm hoping they're at least 30A schottky's.

[ncovert]

Re: Power supply build quality pictorial. part 2

That looks like a good power supply, just needs de-Fuhjyyu'd.
I'd recommend Matsushita for the primary and UCC for the secondary.

[c_hegge]

Re: Power supply build quality pictorial. part 2

Niiiice. Now we just need to find a Macron with Japanese caps and we'll hvae a complete collection

[Pentium4]

Re: Power supply build quality pictorial. part 2

That looks like a good power supply, just needs de-Fuhjyyu'd.
I'd recommend Matsushita for the primary and UCC for the secondary.

Exactly! I do have a nichicon 400V cap I will use if/when I recap it

Niiiice. Now we just need to find a Macron with Japanese caps and we'll have a complete collection

Haha I don't think such a thing exists though. Such a shame they plague their well designed units with the most garbage caps available.

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

They can't put japanese caps. They would get expensive and the mass will buy the new Raidmax/Logisys/Allied/Sunpro instead.... I would advise them to use exclusively Teapo and low esr caps only. I have seen Thermaltake using Teapo GP caps on pi filters... Then you can't blame the Teapo caps for failing.

[Pentium4]

Re: Power supply build quality pictorial. part 2

They can't put japanese caps. They would get expensive and the mass will buy the new Raidmax/Logisys/Allied/Sunpro instead.... I would advise them to use exclusively Teapo and low esr caps only. I have seen Thermaltake using Teapo GP caps on pi filters... Then you can't blame the Teapo caps for failing.

You're right. But it doesn't even have to be Japanese. Think of how much longer OST would last over Fuhjyyu, JEE, etc. Yeah can't blame them for failing...common to see SEK in PSU's, even some of the newer ones

[Pentium4]

Re: Power supply build quality pictorial. part 2

Disassembled, and I was very very pleased to see that the Macron I posted above uses two 40A TO-247 schottkys for the 12V. SWEET!

[mariushm]

Re: Power supply build quality pictorial. part 2

Another "Delux" power supply. Must have been a more expensive computer case because it has enough components in it to to save it from the "hall of shame" thread.

Fan going out, vibrating the whole case. Replaced it.
1000uF 10v capacitor on 5v swollen, probably due to heavy 5v load and bad fan.
Replaced both with 1500uF 6.3v Nichicon caps, replaced the 12v caps with 1500uf 16v as well as a precaution.

[ben7]

Re: Power supply build quality pictorial. part 2

Another "Delux" power supply. Must have been a more expensive computer case because it has enough components in it to to save it from the "hall of shame" thread.

Fan going out, vibrating the whole case. Replaced it.
1000uF 10v capacitor on 5v swollen, probably due to heavy 5v load and bad fan.
Replaced both with 1500uF 6.3v Nichicon caps, replaced the 12v caps with 1500uf 16v as well as a precaution.

Hm those look like fake NCC capacitors on the primary xD

[mariushm]

Re: Power supply build quality pictorial. part 2

Yeah, they probably are. But at least it has proper filtering so it redeems itself a bit.

[momaka]

Re: Power supply build quality pictorial. part 2

1000uF 10v capacitor on 5v swollen, probably due to heavy 5v load and bad fan.

And that large 27 Ohm resistor next to it . (It's dissipating nearly 1W of power - not good.) Try to change it for something that will dissipate 0.5W or less and the new caps will be very happy .

[Behemot]

Re: Power supply build quality pictorial. part 2

Or maybe do flying resistor mod.

[Pentium4]

Re: Power supply build quality pictorial. part 2

It's crazy how much Coolmax overrates their units, and they're super overpriced! This one was in a customers' Pentium D system. Looks like the typical 350W Sun Pro. On this one though, I'm impressed with the size of the main toroid I fixed it up nice for the customer, and I believe it will last quite a long time.

Those Tongjia caps were actually 470uF! I was shocked I replaced the standard ceramic "Y caps" with certified ones, added an X cap, and added two MOV's. It has TO-247 13009's for the switchers, nice heatsink.

Naturally, I recapped the entire thing. Besides the input, all the caps were Ricon. Sun Pro not giving a shit, the 5VSB minimum load resistor (22Ω) was physically touching the 470uF Ricon cap (I replaced the resistor with a smaller one), and it got completely toasted! It read 40Ω ESR!!! Can't imagine the ripple the board was getting, since it was the cap after the PI coil.

The LY fan spins well, and had plenty of grease in the bearing. 30A schottky for 5V and 3.3V. 12V had two 16A ultra fasts in parallel.

The voltage selector switch was giving a lot of tension on the input caps, and was touching the primary heatsink, so I just jumpered it and left the switch on the back, I don't trust the customer to not switch it to 230V

Redid a lot of the soldering, a lot of big blobs and joints starting to go cold.

[c_hegge]

Re: Power supply build quality pictorial. part 2

I'd say about 350W would be it's max. It looks decent with the upgrades, though.

[Pentium4]

Re: Power supply build quality pictorial. part 2

I'd say about 350W would be it's max. It looks decent with the upgrades, though.

Thanks!! Now that I think about it, maybe 300W since the fan likes to spin so slow, and the thermistor is inside the 3.3V toroid which I'm guessing doesn't get as hot as the main toroid or the secondary heatsink. I think the main reason the PCB got so dark, is because the fan spins slow and this thing was absolutely packed with dust. It was disgusting, actually.

[c_hegge]

Re: Power supply build quality pictorial. part 2

Yeah, Sun Pro always seem to put the thermistor in the 3.3v toroid coil. It's a bad idea, since it means that the fan is regulated based on 3.3V load (which it won't have a lot of in most modern PCs). What's more of a problem, though, is that they sometimes, don't use enough glue, and the thermistor comes out of the coil and just sticks up. Then, the fan never comes out of idle.

[Pentium4]

Re: Power supply build quality pictorial. part 2

What's more of a problem, though, is that they sometimes, don't use enough glue, and the thermistor comes out of the coil and just sticks up. Then, the fan never comes out of idle.

That's what would have happened in this case because there was no glue holding it in there, but the minimum load resistor on the 5VSB ironically got so hot that it melted to the heatshrink of the thermistors wire and held it in place

[c_hegge]

Re: Power supply build quality pictorial. part 2

Ah, yes, I can see the damage to the wires in the pictures now.