Power supply build quality pictorial. part 2

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

400W would be the limit but they should be TO-247 package to do that for more than a minute.

[c_hegge]

Re: Power supply build quality pictorial. part 2

Yup. I've pulled 400W successfully from a few 13009-powered power supplies before. They had 1000uF caps, all schottky rectifiers on the secondary and very thick heatsinks.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

An LC-Power 500W 'Gold' thing...

Somehow I doubt it's really a gold because the primary coil looks a little small...

[Pentium4]

Re: Power supply build quality pictorial. part 2

Wow, yeah that thing looks very small for 500W. You should see how hot it gets with a decent load on it. The rest of it looks nice.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Saw similar small PFC coils on some 230VAC only units.

So I'd expect the unit to be finde but 230VAC only...

[c_hegge]

Re: Power supply build quality pictorial. part 2

Wow. That would have to be the smallest PFC coil I've seen on a "500 Watter". I wouldn't want to push it much past 350.

EDIT: Actually, my Antec VPA500P's PFC coil was barely bigger than that, and it did 500W OK, so maybe it is fine as a 500W on 230V

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

The FSP Hexa+ 500W has a similar small PFC coil.

So I'm not worried about that. It seems that the coil can be a bit smaller in a 230VAC only unit...

[Behemot]

Re: Power supply build quality pictorial. part 2

You don't have to boost so much from 315 to 380 V like from 160 to 380 so you have smaller currents and lower boosting frequency. But it may still get somewhat hot and die soon. But with them secondary caps, that doesn't matter I guess

These 230V only also have smaller-rated silicon in the PFC section so you may save a dollar or two on such unit.

Here is the certification so it definitelly did provide 500 W for some time at least, and on 115 V grid. Certification for 230V is madness now, Ecova just took higher 115V rating and placed it for 230 V. So 230 V bronze is basically 115 V silver, that's insane, you won't make 2 % difference just by switching from 115 V to 230 V.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

But with them secondary caps, that doesn't matter I guess

You mean the Teapo SC series caps?
Or do you mean the two JunFu?
I'm really not sure what to make of those two, because there are two rather big (1500uF, 6.3V) polymer caps before those two...

With the ripple I think there isn't much left after those two polymer caps...
The question is: How high is the ripple & noise with and without those caps...
It seems that the ripple & noise is rather low in the Test of the 400W unit. So it may still be in spec without those two...

Certification for 230V is madness now, Ecova just took higher 115V rating and placed it for 230 V. So 230 V bronze is basically 115 V silver, that's insane, you won't make 2 % difference just by switching from 115 V to 230 V.

Well, the minimum here in Europe is 82/85/82% efficiency...

Well, it depends on the design of the unit and the wattage how much you gain from the higher voltage...

Computerbase.de did test that...
If you know a little bit of German you may want to take a look at some tests.

Buttom line: at lower loads 115VAC is more efficient. But at higher loads, you may gain the 2-3% needed for the 230VAC 80plus certification.

[Behemot]

Re: Power supply build quality pictorial. part 2

My mistake, that's even 3 %, and full-range. Bronze certified at 115 V (82/85/82) would need 85/88/85 to be 230V Bronze. You won't do that, you need true 80 PLUS Silver@115 V unit. Than aprox. 2 % for Silver and Gold. Check the new table udner "What is 80 PLUS certified?" button. It is just crazy over the October 2011 table.

If I am looking right, those polymers are filtering input for the DC-DC modules. If other filtering caps get bad on +12 V, I don't think it will be in spec, those caps are not in the right place to smooth all the +12V ripple. Besides, many PSUs with heavily bloated caps are not even able to run, they just do not turn on at all, or shut down during higher loads.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Ahm, what caps do you mean?

The Junfu Ones? There are just two of them, one each for the +3,3V and +5V rail.
On the DC-DC module there are 4 polymer caps on each module: 2 470uF/16V for Input Filtering and two 6,3V/1500uF ones for Output.

Everyhting else is Teapo. Don't know who made the Polymers though.

Here a picture of the +12V...
There seems to be two 470uF Polymers and three Teapo SC with 3300uF it seems. ANd of the DC-DC module where you can see the other two caps.

[Behemot]

Re: Power supply build quality pictorial. part 2

Those 3300uF ones are 16V, right? If they go bad, only couple 470uF polymers will have to handle the ripple. I don't think that will work out.

The polymers itself are Teapo CG.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

When they go bad, but I think that won't happen for a couple of years.

The design doesn't look that bad as they are not near some hot things...
And the worst ripple is taken care of by those two polymers...

[Behemot]

Re: Power supply build quality pictorial. part 2

I think they'll pass away before anything could happen to that PFC coil. That means soon

[Majestyk]

Re: Power supply build quality pictorial. part 2

Thanks for the additional information about the Casing Macron / MPT-301!
I remember there´s a number on the downside of the PCB starting with MPT...
So the PSU doesn´t seem to be too bad after all
The Fuhiyyus were all within specs, but the PSU had never been used. Must of the secondary caps´ capacities are below the ratings though.

Here´s a picture of the PFC-coil:

[Pentium4]

Great wall gw-4000 (85+)

So, I saw this thing on eBay and offered the seller a lower price because it looked nice, and he accepted. Never seen a Great Wall PSU in the states here before. It's labelled as a 300W 80 Plus Bronze unit. So, this has to be the most generously rated chinese PSU I've ever seen. The bridge rectifier is 10A, with a huge heatsink on it. Nice silicon on the PFC and switching sections. There's even a tiny little heatsink on the 5VSB chip

The 12V has two rails, and each one has a 60A schottky The two minor rails have dual 40A schottkys in parallel. The first 12V rail which feeds the CPU 4+4 pin and single PCIE 6 pin connector is filtered by a 2200uF nicon KME, and a 1000uF 16V nicon KME. The rest of the 12V is filtered by a single 1000uF nicon KME. These caps have very low ripple current capabilities, so I will replace them with better parts, like KY for 2200uF and ZLH for 1000uF. HEC primary cap read 355uF, but I'm not a fan of HEC caps even on the primary. I have a brand new 270uF Elite GM that I will use to replace it. The rest of the caps are pretty unknown. The filters caps are all nicon (except for a 2200uF 16V Teapo SY on the 5VSB before the coil), but all the small ones consist of ZHN, and SDC. The soldering is somewhat sloppy, but I've seen worse. The Ever Cool fan was heavily overgreased. It filled the entire shaft all the way up to the seal! I think this thing could do 400W comfortable, maybe more. However, I think I will use this in one of my servers to lower the power bill a little.

[ben7]

Re: Great wall gw-4000 (85+)

So, I saw this thing on eBay and offered the seller a lower price because it looked nice, and he accepted. Never seen a Great Wall PSU in the states here before. It's labelled as a 300W 80 Plus Bronze unit. So, this has to be the most generously rated chinese PSU I've ever seen. The bridge rectifier is 10A, with a huge heatsink on it. Nice silicon on the PFC and switching sections. There's even a tiny little heatsink on the 5VSB chip

The 12V has two rails, and each one has a 60A schottky The two minor rails have dual 40A schottkys in parallel. The first 12V rail which feeds the CPU 4+4 pin and single PCIE 6 pin connector is filtered by a 2200uF nicon KME, and a 1000uF 16V nicon KME. The rest of the 12V is filtered by a single 1000uF nicon KME. These caps have very low ripple current capabilities, so I will replace them with better parts, like KY for 2200uF and ZLH for 1000uF. HEC primary cap read 355uF, but I'm not a fan of HEC caps even on the primary. I have a brand new 270uF Elite GM that I will use to replace it. The rest of the caps are pretty unknown. The filters caps are all nicon (except for a 2200uF 16V Teapo SY on the 5VSB before the coil), but all the small ones consist of ZHN, and SDC. The soldering is somewhat sloppy, but I've seen worse. The Ever Cool fan was heavily overgreased. It filled the entire shaft all the way up to the seal! I think this thing could do 400W comfortable, maybe more. However, I think I will use this in one of my servers to lower the power bill a little.

WOW that is an impressive unit for being an off-brand. Maybe you could also check the heatsink temps while it is under load for us? I'm curious as to how well those heatsinks dissipate heat, as I don't recall seeing very shiny ones like that before.

[Behemot]

Re: Power supply build quality pictorial. part 2

Great Wall is chinese OEM with increasing volumes, they are making some impressive high-power units and also some units for known brands. I think they have highest-power ATX PSU at all ATM, something around 2 kW. You are limited with less there in US because of lower voltage.

[stj]

Re: Power supply build quality pictorial. part 2

that would be nice if they could put the parts flat against the pcb before soldering.

although it's not going to effect the running, i cant stand seeing that.
when i see that i have to desolder, re-seat and then solder parts in correctly - it's just me!
i had to even do it with the control board in my hot-air station.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

That one looks like the same plattform as the LC-Power GP3...
http://www.tomshardware.de/LC-6550GP...-240781-8.html