Power supply build quality pictorial. part 2

[stj]

Re: Power supply build quality pictorial. part 2

yea, but the rectifier diodes are covered by the coils - i'd have to pull the whole damned heatsink assembly out - again.
i already pulled it once to replce a burned resistor & shorted ceramic cap behind it!

maybe over the weekend i'll pull it appart and think about it's future.
after i realised the input rectifier was 4a i was planning on scrapping it.

[stj]

Re: Power supply build quality pictorial. part 2

that one in the review link is exactly the same except it has a second fan.
so this pile of shit is (playing safe) a 320w psu.

i'm gonna trash it.

[Pentium4]

Re: Power supply build quality pictorial. part 2

320W is easily usable. And 4A is fine for the BR since you're on 230V

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

I would definitely keep it. Decent sized transformer, massive heatsinks, primary transistors good for 400W max, good ventilation with honeycomb style, pi coils on secondary, toroid coils good sized if we assume it is a 400W peak psu.

It would power without a problem any PCs without dual graphics cards or some super power hungry GPU or CPU. So, why toss it?

[c_hegge]

Re: Power supply build quality pictorial. part 2

320-350W may be it's limit, but it isn't too bad for a 350 Watter if you care to recap it. It will do that well and truly in spec.

[Pentium4]

Re: Power supply build quality pictorial. part 2

I recapped the Macron. When testing before the recap, it made a whining sound. Figured it'd be failing 5VSB caps or one of the other rails' filtering caps. I was absolutely amazed that all the filtering caps were in spec. Only two caps failed, the 10uF 50V GoldLink caps for the switching transistors. One read 8uF, 8Ω ESR, and the other one read 4.46uF, 40Ω. WHOA! That must have been causing the whining because after the recap it doesn't make that sound anymore. I have just one question, how did that cap with 40Ω not blow the switcher?

Also, those big Fuhjyyu caps read 1034uF and 1032uF. I was pretty impressed, it must be because they're rated for 105C. I used 1000uF 10V nichicon PW for the minor rail filtering because they beat the Fuhjyyu TN 2200uF 10V specs, and 8mm vs. 10mm allowed for better airflow in that area. Fixed a lot of solder joints, there were a lot going bad, especially on the 5VSB and the main transformer. I'm throwing it in a file server.
Edit: Almost forgot to mention, I added a 4 pin to make the motherboard cable 24 pin, and added a single string of 3 SATA connectors, so that I can load it up with a bunch of hard drives. When loaded to 150W mostly 12V based, the 12V drops to 11.77V. Seems to keep the hard drives a little cooler.

[stj]

Re: Power supply build quality pictorial. part 2

I would definitely keep it. Decent sized transformer, massive heatsinks, primary transistors good for 400W max, good ventilation with honeycomb style, pi coils on secondary, toroid coils good sized if we assume it is a 400W peak psu.

It would power without a problem any PCs without dual graphics cards or some super power hungry GPU or CPU. So, why toss it?

well, do you know how much W's id need for a sl8zh netburst/smithfield cpu,
with an intel 965 mobo and an average gpu + 4 harddrives?
i know the cpu peak is 127w!!!
average 95w

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

I actually run successfully a Pentium D 2.66GHZ (dynamic OC up to 3.2GHZ), 1.x SATA HDD WD, 7600GT MSI, 2 dimms of DDR2 and DVD-ROM, + DVD RW with an old half bridge but decent power supply.

That PC came with a Premier (Deer) "400W" psu that was 250W capable and the computer worked without the psu blowing. I can't imagine the ripple though...


BTW I have tested the power consumption of that PC with an online UPS and it maxed at 250W.

[momaka]

Re: Power supply build quality pictorial. part 2

well, do you know how much W's id need for a sl8zh netburst/smithfield cpu,
with an intel 965 mobo and an average gpu + 4 harddrives?

Absolute worst case (assuming 140W for CPU with VRM innefficiency, 50W for a mid-range GPU similar to GeForce 7600 or 8600, and 2A peak for each HDD as they spins up + 1A for all "other stuff"): 301W peak from the 12V rail, or about 25A peak. After the HDDs spin up, you're probably looking at about 18A with CPU and GPU fully loaded, and much less when they are not. So as long as the PSU has two 20A rectifiers in parallel, it should be able to handle this no problem. If you didn't have 4 HDDs, you can probably get away with a single 20A rectifier. This is for half-bridge topology.

[stj]

Re: Power supply build quality pictorial. part 2

well there are no parallel rectifiers i suspect because there are parallel holes on the back of the secondary heatsink area to fit one!

[electro7fan]

Re: Power supply build quality pictorial. part 2

You could put another rectifier if there's a spot... this PSU should be perfectly fine on a basic system but as long you don't put like a 125W CPU for example, common sense here.

[Pentium4]

Re: Power supply build quality pictorial. part 2

^ I agree that it's worth recapping. And with newer equipment, they really don't use that much power at all.

I recapped this Sun Pro (branded Winsis, posted a while ago) Pretty decent PSU now. Those CS caps measured ~810uF. I added MOV's, and relocated the fan control thermistor. It was in the 3.3V toroid coil, so I moved it to the main toroid. Fan still spins really slow, but it's nice and silent. It's in my main file server now which is Intel D946GZIS (All nichicon caps), Celeron D 3.46GHz Cedar Mill, 3GB DDR2, gigabit Pcie x1 card, two 2TB Seagate drives in RAID 1 via adaptec 1430SA.
Check out how much grease they put in the fan! I removed some of it. And, even worse....check out that 3.3V wire! It was never even soldered into the PCB. I imagine this could have caused weird issues? It leads to the motherboard 24 pin connector. Also, the soldering was pretty terrible. I probably redid 1/3 of all joints. Looks good now, pretty beefy for a Sun Pro

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

I have never seen something like this before. Usually they have no oil at all, or some 1 or 2 drops of it. This is amazing. What they were thinking? What brand is the fan?

[Agent24]

Re: Power supply build quality pictorial. part 2

check out that 3.3V wire! It was never even soldered into the PCB. I imagine this could have caused weird issues? It leads to the motherboard 24 pin connector.

Probably not, they are all just in parallel for the main connector so if one is not connected the others will just have to carry a bit more current.

[Pentium4]

Re: Power supply build quality pictorial. part 2

I have never seen something like this before. Usually they have no oil at all, or some 1 or 2 drops of it. This is amazing. What they were thinking? What brand is the fan?

Yeah I was surprised too. It's a Sun Pro, so naturally it's a Te Bao fan

Probably not, they are all just in parallel for the main connector so if one is not connected the others will just have to carry a bit more current.

That's good, and the 3.3V rail isn't really used much in modern computers

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Not a bad unit. No, the SG6105 doesn't support OCP. That is a single rail unit.

Huntkey does like to do funny stuff in case of the OCP of a unit (like they did with the X7-1200W).
Yes, the SG6105 doesn't support it. But what about the AS339P right next to it? What is that for?

[Wester547]

Re: Power supply build quality pictorial. part 2

Not a bad unit. No, the SG6105 doesn't support OCP. That is a single rail unit.

It does appear to have a separate set of +12V wires, so maybe it is a dual rail unit?

AS339P is a voltage quad comparator chip. I don't think it has anything to do with dual +12V rails. In most OEM units that I've seen, the thermistor is always mounted to the secondary heatsink for better or worse (so the temperature of the secondary heatsink determines how fast or slow the fan spins depending on how aggressively the thermistor increases or decreases its resistance in response to that). The cheapo units, not as much. I would think that the +5V/+12V "main" toroid should get hot enough to ramp up the fan speed unless the thermistor is designed to run the fan slowly.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Here something you could get for 14,95€ in Germany right now...

[Behemot]

Re: Power supply build quality pictorial. part 2

What's the power and efficiency? What is it actually?

[c_hegge]

Re: Power supply build quality pictorial. part 2

^
It's a Delta DPS-500QB - a 500W 80plus bronze unit. It isn't quite standard ATX size, but if it will fit in your case, it would be a very solid unit. Entirely Jap caps, DC-DC Conversion for the 5 and 3.3V rails, 2BB fan and all the good stuff you'd expect from server/workstation unit.