Power supply build quality pictorial. part 2

[japlytic]

Infinity ATX P4-400W

Cheng 470uF 200V primary units.
Non-approved Y-class units .
13007 primary switchers.
FSDM311 +5VSB switcher.
HS8108B control IC.
1000uF 10V Bor Hurng before and after +5VSB inductor.
1000uF 10V Bor Hurng on +3.3 and +5V rails.
1000uF 16V Bor Hurng on +12V rail.
220uF 16V Bor Hurng on -12V rail.
S10C45C rectifier on +3.3V rail (16A).
S16C45C rectifier on +5V rail (24A).
S12C20C rectifier on +12V rail (16A).
KBP307C input rectifier.

[pdavid]

Two ATX psus.

Compaq 250W (Lite-On) Veeeery heavy
680uf 200V ????
Secondary has Rubycons...
Can't see much really


Ever-Power 235W (Channel Well)
2x 330uf 200V Teapo
2x 1000uf 10V Capxon
2x 1650 16V Fuhjyyu
Oh well...

4 Amp rect. bridge
2x 13007
2x P40NE mosfets
1x D83-004 Schottky

Both working good. They were just dusty.

[everell]

The EverPower is a ChannelWell AT power supply with add on 5vsb board to make it a ATX power supply. They also did this for Antec. Here are some pictures of an Antec ATX power supply made by ChannelWell with the 5vsb retrofit board.

[pdavid]

When I opened up the EverPower I remembered other psus having similar add on 5vsb circuits. It seems the KA393 gets power from a small board that says ps-on.

I haven't looked too much into the 250W Compaq. This one looks way more intersting with it's quaint buildup.

[mariushm]

This is a power supply I have nice memories of - well not really this psu but the computer it came from, a HP Vectra with a 486/66 Mhz that was gifted to me so that I could have a computer in the first year of university.

As you can see from the pictures, it's a Delta, pretty much plain AT... 5v@13.5A, 12v@4.5A with a max output of 100w
Minibea fan 12v@0.3A , can't tell the actual mosfets/diodes because I'd have to desolder the inpu wires or remove the heatsinks but the capacitors are all quality :

Rubycon 270uF 400v on the primary,
1 United Chemi Con or Nippon Chemi Con (it's that particular logo on them) sxe 2200uF 16v and 2 2200uF 10v

Rubycon YXB 1000uF 10v on secondary,
some SXE 2200uF 10v
a few small KME 10uF 50v, a 220uF 3.5v
a lonely taicon 1000uf 16v on the 12v line and a few small ones around the center of the pcb, 0.47uf, 47uF

It still worked fine about 3-4 years ago but it should still work great now.. For something made in '96 it still looks good, nothing bulging or leaking...

I'd like to know the topology but I assume it's a half bridge or something before that...

[mariushm]

Recently had to open up my Samsung ML-1710P printer so I thought about doing some power supply snapshots.

center IC is str g6353 ... the one on the left I can't reach

big cap is Lelon 400v 100uF LS series
black cap on the right si a Samwha 1000uF 16v (s3c) , green one is a Lelon 1000uF 35v RGA series, blue one is a Lelon but too lazy to bend it and read the value

Don't see the need to recap it as it works fine... well as fine as a drum with 10k pages lifetime works after 15k pages printed on it (there's some toner graying out a few millimeters on the right side of the pages, like you would see on some fax machines)

[momaka]

Originally Posted by mariushm

I'd like to know the topology but I assume it's a half bridge or something before that...

Hard to say exactly... there's a small transformer that could be the driver if this is a half-bridge design. But there's also only one mains cap, so this could also be a MOSFET-based single-transistor forward design (with IC1 being the PWM controller and IC2 and IC3 providing feedback to primary). Most likely it's a single-transistor forward. How many transistors are there on the primary side, and what are the various IC?

[japlytic]

SLB ATX-300W

All electrolytics are ChengX.
Half of the bridge rectifier are 2A units, with the other two 3A units.
470uF 200V primary units.
1000uF 10V before +5VSB inductor and 470uF 10V after it.
2200uF 10V before and after +3.3V inductor.
2200uF 10V before and after +5V inductor.
2200uF 25V before +12V inductor and 470uF 16V after it.
470uF 10V on -5V rail.
470uF 16V on -12V rail.
2SK2141 primary side MOSFET (this is not an old-school style power supply!)
2SC5027 primary side transistor for +5VSB.
SBR2040CT rectifier for +3.3V.
SBL3040PT rectifier for +5V.
MUR10100CT rectifier for +12V .
TD3843 (=UC3843) controller with LP7510 voltage supervisor.

[b700029]

Leadman/Powmax!

I like how they used a huge axial diode in a place silkscreened for a tab-mount (but the heatsink was cut too short to reach it)

[Th3_uN1Qu3]

Originally Posted by japlytic

2SK2141 primary side MOSFET (this is not an old-school style power supply!)

You know i was preparing to ask where the TL494 is... But i guess i know why they did it like that - look at that tiny transformer! Also, a currentmode controller makes the thing harder to blow up. That means, motherboard goes out first.

[japlytic]

Omni PS-ATX 500W

680uF 200V CapXon LP primary units.
3300uF 10V Xunda (bulging) before +3.3V inductor with 1000uF 10V XJ after it.
3300uF 16V Nova on +12V rail.
2200uF 10V XJ before +5V inductor with 3300uF 10V Xunda (bulging) after it.
470uF 16V XJ before and after +5VSB inductor.
470uF 16V XJ on -5V and -12V rails.
MJ13009 primary switchers.
6A 800V bridge rectifier.
SBL3040PT rectifier each for +3.3V and +5V rails.
U20D20C rectifier for +12V.

[c_hegge]

Another Omni PS-ATX 500W (they are obviously not that consistent with their designs).

680uF 200V Nippon Chemi-Con KMG primary caps

2x 1000uF 10V SD caps on 5v Rail (one before and one after Jumper wire), ST3045CW Schottky.

2x 1000uF 16V <BH> on +12V rail (one before and one after Jumper wire), STTH2002CT rectifier

The 3.3v rail is uses the same parts as the 5v rail (only, it has a PI filter coil)

ST W13009 primary switchers.
6A bridge rectifier.

I have another identical PSU, which I re-capped, added a fan, and replaced the ceased up fan with a CM blue LED fan, this one, on the other hand, I'm just gonna explode.

[SIDMX]

Pixxo Jaguar PW series 600W (K-Mex)


Bridge Rectifier - KBL06 600v-4A
Primary Caps - 2x HEC 560uf-200v
Primary Switchers - 2x MD 13009K 400v-12A
Primary Side Mosfet - CET CEF-02N65D 600v-1.5A
Schottky Rectifier - MOSPEC S30D40CS 40v-30A
Schottky Rectifier - MOSPEC S20C40C 40v-20A
Switch Mode Rectifier - MOSPEC F16C20C 20v-16A
Supervisor+Regulator+PWM - System General SG6105ADZ
Output caps for +3.3v- 2x YC 2200uf-10v
Output caps for +5v - 2x YC 2200uf-10v
Output cap for +12v - Chengx 2200uf-16v
Output cap for -12V - KHC 470uf-16V

There's no way it can do 600W, maybe 400W Max ??, what do you think guys?

[lti]

That looks like a 200W power supply to me.

[goodpsusearch]

300-350W max

edit: With that transformer I don't think it can make more than 300W...

[momaka]

Originally Posted by SIDMX

There's no way it can do 600W, maybe 400W Max ??

More like 300W continuous. 350W on a good day. It will do 400W only when it explodes and shorts momentarily before the fuse blows.

I bet you the over-power protection on the SG6105 is set to that as well (about 300W or so).

Originally Posted by SIDMX

What do you think guys?

Let's see, here...
60A on the 5v rail with only 1x 30A rectifier? - I don't think so. Maybe peak for 0.1 ns. The 3.3v and 12v rails claiming 36A with 20A and 16A rectifiers respectively is also a complete lie.
What else? - the spacing of the traces on the PCB on the primary side is wrong - it should be much more than that. The transformers are undersized and UL logo is forged (no number).

[c_hegge]

It looks very similar to the Rexpower PL400 which I re-built, so the OEM would be KeyMouse. On the PL400, the OPP was set at only 200W. Whether this would be the same or not, I don't know. the only way to find out for sure is put it on a load tester (or send it to me )

EDIT. I do also have a K-Mex branded PW450 unit. It will do up to 350W, but this one isn't quite as beefy.

[b700029]

They didn't even bother to change the ratings for the rails with different wattages!

The limiting component here is the primary rectifier, 4A at 120V gives 480W in but with 75% efficiency (typical) that's 360W out. The secondary is good for 400W so the first thing to go will probably be the primary rectifier.

EI-40 transformer is sufficient for a 350W unit, and is even found in some real 400W.

[Th3_uN1Qu3]

Originally Posted by b700029

The limiting component here is the primary rectifier, 4A at 120V gives 480W in but with 75% efficiency (typical) that's 360W out. The secondary is good for 400W so the first thing to go will probably be the primary rectifier.

I've heard that a million times. And no, the primary diodes almost never blow even though in theory they are way undersized for the task. None of the power supplies with the "4-diode treatment" tested by Hardware Secrets ever blew the primary rectifiers. It's the primary switchers which almost always give up due to overheating, and the 12v rectifiers sometimes.

[goodpsusearch]

Delta DPS300MB 300W

5vsb = 4.2V

1 bad Capxon GL 1500uF 16V and 1 Ltec 220uF 25V that had high esr

Attached photos with the Capxon removed: