Power supply build quality pictorial. part 2

[ben7]

Re: Power supply build quality pictorial. part 2

Prepare your vomit pail

This is older revision, I will also talk about newer one which I got to my hands a year ago. This thing is still sold for like 24 USD/19 Eur incl. VAT. I tried to get local inspection to forbid that, you will see how many things it violates, but I think the „engineer“ dealing with me was both morron and on paycheck of the manufacturer.

Pic 1 and 2 shows nothign spectacular, just remember the values

Pic 3 - overall looks. This thing even has PFC, the newer revision got absolutelly tiny coil and ceramic cap - most likely PFC imitation, usefull for 5 W max PSU. According to inspection, PFC is not a must, the European norm actually talks about harmonics radiation - you can use just good ol' input filtration for this! No PFC is required in EU! Thats common mistake from those who actualyl never read the norm But when they write PFC in the name, there should be one, right

Anyway, filtration is on daguhter board, but nobody would be harmed if they used the free spots too. Thermistor and varistors nowhere to be found. I will add all the things missing. Newer revision actually has two varistors.

Tiny diodes, newer version even has rectifier, but 2 A labeled with rubber electrodes. The things is really small and starts dropping current at just 50 °C. Basically no difference here.

Input caps are 470 uF, Fhy, never ever heard of that. Never revision has similar crap. Transistors are Toba T3866 and T13007-R. Found only bullshit, no actuall datasheet. Transistors seem OK, both heatsinsk really thin (2 mm) and tiny.

Ordinary diode on +5 V SB. 20A schottky for +5 V, 12A fast-recovery for +12 V and 10A schottky for +3,3 V. Do you still remember the values from pic 1? But according to inspection: it is OK, the values on casing are just peak ones, manufacturer cannot expect you will connect it somewhere else than ordinary computer so it does not need to provide the wattage specified for longer period of time (longer than like 1 second).

Only coils there are on +3,3 V rail, otherwise plain wire or stuff. Notice the termistor on common coil - seems that rectifiers are not the hottest things in this PSU, considering no coils and small Asia'X caps -5 V rail is not occupied by components

Overall soldering quality is mediocre, PCB is not worst. Newer revision has even thinner PCB and casing. They most likely have one engineer for researching how these things can be cheaper because I found something new every time in there

lol wow

AsiaX and 'FHY'?

FHY = F***ing Hate You?

and LOL, 'Toba' transistors? Toshiba ripoff?

[Th3_uN1Qu3]

Re: Power supply build quality pictorial. part 2

Anything with 13007 in the name is MJE13007 copy, everyone and their mom makes them. Just like 13009 = MJE13009.

Man, just look at how they threw those silpads on there. Those rectifiers could short together anytime. But hey, at least it has a thermal fan controller.

[Behemot]

Re: Power supply build quality pictorial. part 2

Will add some rectifiers in parallel, hopefully it will also reduce heating and together with 12cm fan the tiny heatsink will be enough. Than rectifier instead of diodes, transistors are enoguh for 350 W hopefully. Also add all missing pieces

I think the efficiency will rise at least 5 %

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

13007 = 300watt max

250W to be more realistic

[c_hegge]

Re: Power supply build quality pictorial. part 2

I've never been able to pull more than 300W from a PSU with 13007s, and that's on a good day. Usually it's more like 250W.

[Behemot]

Re: Power supply build quality pictorial. part 2

Your ruining me guys Why is that? Aren't they rated for 8 A? Considering 325 V, its 1,077 A to get 350 W. Even just 340 W, considering some 10 W is supposed to go through +5V SB

[Th3_uN1Qu3]

Re: Power supply build quality pictorial. part 2

The primary voltage isn't gonna be anywhere near 325v under load. Plus, in half bridge, the primary voltage swings between +/- 1/2 of the input voltage measured across the primary caps, because the caps form a divider and the transformer is hooked to the midpoint of this divider.

Hence the instant peak current in either transistor is double the current drawn from the AC line, because they swing the transformer primary between half the input rectified mains voltage (the midpoint between the two primary caps), and the + or - of the rectified mains, in turn.

[tom66]

Re: Power supply build quality pictorial. part 2

Also the crappy drive circuits used in these PC power supplies means high switching losses.

[Behemot]

Re: Power supply build quality pictorial. part 2

So, with power factor of 0,7, poor efficiency of 70 % and losses, the load will be more than 8 A? Damnit.

Well, will have a look for two 400+ V/10+ A NPN transistors, could have some arround…

[Wester547]

Re: Power supply build quality pictorial. part 2

I've never been able to pull more than 300W from a PSU with 13007s, and that's on a good day. Usually it's more like 250W.

What about two Fairchild 13009 switching transistors with two 470uF/200v (85C) primary capacitors, and a PSU otherwise based on half-bridge topology (definitely no PFC in this case)? Does that sound like a 300W continuous to anyone or is it more like 250W continuous, 300W max? ^^;

[Pentium4]

Re: Power supply build quality pictorial. part 2

Surprised it has tansient filtering. Does it have Passive PFC or is that thing just a weight?

[b700029]

Re: Power supply build quality pictorial. part 2

What about two Fairchild 13009 switching transistors with two 470uF/200v (85C) primary capacitors, and a PSU otherwise based on half-bridge topology (definitely no PFC in this case)? Does that sound like a 300W continuous to anyone or is it more like 250W continuous, 300W max? ^^;

http://www.hardwaresecrets.com/print...ply-Review/557

350W continuous 400W max should be comfortably within limits. Efficiency might be a bit low.

(The article in that link shows efficiency at less than 110V in. It will be much better, even 80%+, with 240V.)

But according to inspection: it is OK, the values on casing are just peak ones, manufacturer cannot expect you will connect it somewhere else than ordinary computer so it does not need to provide the wattage specified for longer period of time

This is also partly the reason why so many would be satisfied with a $15 "650W" Powmax or similar low-end Deer/L&C: their systems never draw more than ~150-180W at full load, and idle much lower (50-60W), which is well within the capabilities of even those gutless wonders.

[Wester547]

Re: Power supply build quality pictorial. part 2

http://www.hardwaresecrets.com/print...ply-Review/557

350W continuous 400W max should be comfortably within limits. Efficiency might be a bit low.

(The article in that link shows efficiency at less than 110V in. It will be much better, even 80%+, with 240V.)

That PSU has 680uF/220V primary capacitors. I meant a PSU with only 470uF/200V primary capacitors, which I imagine knocks down the maximum power output to 300W, continuous 275W or something like. And well, I live in the United States, so I'd have to set it to around 115V. ^^;

And I agree that it's hard to use more than 200W without a beefy graphics card or many, many add-in cards, (or likewise, USB devices), or without overclocking heavily.

[momaka]

Re: Power supply build quality pictorial. part 2

Man, just look at how they threw those silpads on there. Those rectifiers could short together anytime.

Whatever. QC PASS. Next!

This is also partly the reason why so many would be satisfied with a $15 "650W" Powmax or similar low-end Deer/L&C: their systems never draw more than ~150-180W at full load, and idle much lower (50-60W), which is well within the capabilities of even those gutless wonders.

Exactly.

One of my friends was building an i5 last month for another friend. For some reason, the computer wasn't working properly (would turn on and then turn off after 10 seconds... and occasionally display the BIOS boot screen then turn off). We tried different RAM, different graphics card, different CPU... basically everything. Narrowed it down to either motherboard or PSU. The PSU he had inside was a 750W Corsair, so it was unlikely that that was the problem. Regardless, we decided to try another PSU. Since he didn't have any other PSUs, we took the computer to my house to try it with one of mine. I hooked it up to my 300 Watt Inno Power (Macron Power) MPT-301. He didn't believe me that my PSU would even run - but it did. The symptoms were exactly the same as before... Next week he got a board replacement (another Gigabyte UD3) and everything worked so it was the board that was causing the problem, as we confirmed after trying my PSU.
The point of the story, however, is that even my old Macron could run that PC with its onboard video. And the 12V rail on my Macron didn't even dip below 12V (nor was the 5V too high). Considering the CPU is 95W TDP, I think it did pretty well. Moreover, the PSU didn't even blow any hot air, and my room was at close to 86F/30C that day.

Speaking of which, I don't think I ever posted pictures of that PSU. It was recapped long time ago and used for testing quite a bit. Works great with both 5V and 12V -heavy PCs.

[Behemot]

Re: Power supply build quality pictorial. part 2

Surprised it has tansient filtering. Does it have Passive PFC or is that thing just a weight?

I can see capacitor in there as well as half a ton of copper wire. I guess it is actually PFC but dont want to open it up really

Well, can change inputs in one 350W Fortron (Fuhjyuus 680 uF/200 V ATM) for Chemi-Cons and even repair it when I will be doing it (again, it does not eat small caps, probably ESR related stuff) and use Fuhs in here. The thing is, I will have to rise price for the Forton at least 50 % and sell this crap for more than 1 CZK/W

[Behemot]

Re: Power supply build quality pictorial. part 2

My dear pricey Chemi-Cons are safe! I realized I got another dead Colorshit arround. Its 350 W (actuall wattage is lower) and not really worth the repair. But has 680uF input caps which I can take…

[Pentium4]

Re: Power supply build quality pictorial. part 2

Got quite a few to share with you guys I paid $12 total for all of the following PSU's. Up first is this Allied I picked up. The guy at the computer shop wanted $10 for it (LOL) but I showed him there were bad caps through the vent and bargained him down to $3. I'm actually surprised by it! It's manufactured July, 2002. All the failed caps are DON.



I know they are YC's but for the low wattage and being 680uf, 200V they should be just fine. Oh wait...what is that?! A full transient filter on an Allied? WHAT?!


Look at those pathetic caps haha. They couldn't have puked any harder. I have some caps, I'm actually going to recap this thing. The caps on the secondary are DON, 1 Jun Fu, some G-Luxon and a brand I can't see until I desolder it. PWM controller is a DR-C2002. I have some UCC KY's that are the same uf and Voltage, but shorter and fatter, should they work to replace the DON and Jun Fu?




Up next is this LiteOn unit. I paid $8 for it, and I'm glad I did. It looks great! look at those beefy heat sinks!!!! And gotta love the PFC




Primary and Secondary caps are NCC, the rest are Taicon, Ltec, and I think a Teapo hiding down there.






So I went into this computer store and I asked if they had any power supplies that they didn't want. The guy said "Yeah we got some dead ones here that we are going to throw away" I asked to go through it and found this Antec! They just let me have it. Unfortunately they cut off all the wires Well, I can just solder on some new ones from one of my other PSU's. All it has is ONE dead OST on the secondary! Primary cap is a Hitachi. Semi modular.









This FSP I got from the same guys. This thing is HEAVY. Like 6 lbs....Holy balls, look at that 5V.




Secondary caps are Teapo


Those are some badass Primary caps!!!

Is this thing worth soldering wires to or just salvaging it? What do you guys think of the others?

[ben7]

Re: Power supply build quality pictorial. part 2

The FSP and the liteon look lovely!

Both are worth fixing

[Pentium4]

Re: Power supply build quality pictorial. part 2

They do don't they The LiteOn runs fine. There's nothing wrong with it!

[ben7]

Re: Power supply build quality pictorial. part 2

They do don't they The LiteOn runs fine. There's nothing wrong with it!

Haha LOL!

Did you test the delta yet?