Power supply build quality pictorial. part 2

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

Why do so many people write my name with PH?

Anyway, back to the Orion:
Was that a unit with passive PFC??
I've seen that a couple of times that some units omitted the input filtering with a PFC coil...

[dmill89]

Re: Power supply build quality pictorial. part 2

Anyway, back to the Orion:
Was that a unit with passive PFC??
I've seen that a couple of times that some units omitted the input filtering with a PFC coil...

The version sold in the US has no PFC, versions sold in other countries may have passive PFC (since some countries now require PFC).

This particular variant had no input filtering as well, though the ThermalTake Purepower 430W is essentially the same unit with a black paint job, input filtering, and a more realistic wattage rating.

[Wester547]

Re: Power supply build quality pictorial. part 2

Both this and the Bestec would last a very long time with good caps. It's just that the Bestec has some of that conductive glue, which needs to be removed.

On that note, be careful with LiteOn units, as some of them tend to have the conductive glue as well. HiPro and Delta usually don't.

YES.
Although, this does look like the conductive glue a little bit, I can assure you it is not. Delta uses a white silicone glue that tends to turn beige/yellow over time. But it does NOT go conductive, so it's fine.

Nope! Have a look again.
It's silicone glue and not the brown conductive stuff. You can tell by the round shapes too. The conductive glue tends to look more like vomit.

Wow. I had no idea. Guess you learn something new everyday - I thought the white silicone glue always remained white over time and that the yellow/light brown glue would always turn dark brown/black and conductive. At least all my old Newton Power Dell PSUs still have white glue (maybe it’s because they run so cool ). Sirtec / High Power also used the messy, brown/tan, truly conductive glue in some of their older PSUs.

So what you’re saying is, just because the glue is light brown or yellow, doesn’t mean it’s conductive. I should have known better to research that before I opened my mouth.

[jarvis7]

Re: Power supply build quality pictorial. part 2

Delta GPS 300 transistor has marking - P3NK80Z GKO6J CHN 619
and the IC next to Primary caps is - 3843B K9P623

So , i went to market -

I can swap this for HP-D2537F3H rev 3 for no extra money..

Not sure how much different R model is but here are the pics of D2537F3R by member Spacedye69 and a youtube video by anther member likely of badcaps

https://www.badcaps.net/forum/showpo...1&postcount=88

https://www.youtube.com/watch?v=odxR0bI-Rlc

i found this old post by Wester547 sir

Depends on which Hipros you're talking about. If you mean some of the cheaper ones like the ones that came out of eMachines or Gateways, (HP-P3527F3/HP-P352F3W/HP-P3507F5/W, etc), yes, those are built just well enough to do 300W @ 35*C ambient. However, the ones built for HP and Dell (HP-D2537F3x/HP-P2507F3x, etc) were always very overbuilt in my experience. Most of them did use linear regulation for +3.3V (but in such PSUs the +3.3V wires were almost always 16 gauge with the rest being 18 gauge save by the floppy wires of course) which was inefficient but it didn't really waste much power through the capacitors and did allow for especially low ripple and especially tight voltage regulation on that rail (they were known for extremely tight voltage regulation and low ripple all around, though).

Those power supplies were so overbuilt (with pretty much every type of protection circuitry you could want in a PSU, and always using a PWM TOPSwitch or Tinyswitch to generate +5VSB instead of two transistor +5VSB circuits) and so well designed that even crap like Asiacon, G-Luxon, and Sus'con lasted a long time in those (only exception to that... the Hipros that had Arcons, those failed very very quickly...). However, +5VSB still might have had failed capacitors, with enough time "plugged in" or "soft off" (or standby and hibernation).

Only other issue with those units was prematurely seizing ADDA sleeve bearing fans, and as noted above, sometimes conductive glue. Well, that, and sometimes loose fan thermistors. However, there is a point about a danger in using a power supply full of nothing but Teapo, because at that point at least one or two is bound to be in a hot location... that being said, I too could easily recommend a recapped Hipro without conductive glue (and with a good fan). Their older power supplies always had spots for 12.5mm capacitors and always used very thick heatsinks and very thick coils (on both the input and output).

i am not sure if "NO 2 transistor for 5vsb" is applicable to H version and whether glue is conductive or not.. , so is this D2537F3H good and fully safe, no 2 transistor 5vsb ?

I can pay 2$ more and Get

1) from hp machine- Lite ON PS 5301 08 -HP ( 12v -19A , 5v - 25A )

similar model

https://www.badcaps.net/forum/showthread.php?t=60822

2) from Dell machine - Lite ON L305P 01 ( 12va- 18A max , 12vb max - 18A , 5v-22A .. 12v a and b max combined 274 watts )

https://www.badcaps.net/forum/showthread.php?t=33442

3) From HP machine- D13 180P1A - ( 12va - 7A , 12vb - 10A , 5V -10A , 3.3v- 12A )

E143709 - Chicony ( has same number as HIpro)

input only 200-240V

this is very recent , has a datecode - 14/09 .. i looked through grill, caps were mostly Ltec and a few teapo , some glue was visible not sure if conductive or not.. i don't know whether this is APFC or not.. but i don't want to go for APFC preferably

( reason- UPS compatibility plus i already have 2x 200v panasonic primaries from 2005 bestec , as bulk caps are expensive and i require a very long hold up time of 80-90ms + with 70-80 watt load , so not sure if this 180 watts PSU may have a very low capacitance primary cap and i have read APFC stresses primary caps a lot so need good caps here, and on the contrary non pfc primaries are never stressed )

which one should i pick ? any comments, suggestions ?

Thanks a lot for your time , Help and sharing your wisdom

[jarvis7]

Re: Power supply build quality pictorial. part 2

with some luck and thanks to the poster, here is the internal shot of Chicony HP D13 180P1A PSU mentioned in the previous post

Please scroll down

https://item.taobao.com/item.htm?id=544329693438

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

Powertech ATX-530W

This power supply has the same name with these: 1 2, but this version is not gutless. I would say it's semi-decent. This is the 2011 version of ATX-530W, the gutless brothers of this psu are 2012 and 2013 versions.

Let's see what makes it not so bad:

-Full AC filtering, only MOV missing

-main transformer real 35 size

-not pathetic heatsinks. Good for 250W continuous, 300W peak.

-2 caps per rail for 3.3V and 5V, 1 cap and a coil for 12V.

-full pi filter for 5vsb

-group regulation coil ok sized for 250W

-13007 transistors and MBR2045 rectifier for 5V and 3.3V, 20F20C3 rectifier for 12V

-and the fan has not seized

But it's dead. The fuse has exploded as you can see in the pictures and 1 of the 4 input diodes that form the bridge rectifier is shorted. There is also some of that glue that turns conductive and foam instead of an insulation sheet bellow PCB that, according to some reports, could get conductive too.

Would you fix it?

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

But it's dead.

Doesn't look like you lost much...

It seems like the +5VSB area is rather close to the mains Input and not even a centimetre between the solderjoint of that cap and mains input.

Looks dangerous.

Big Clive might like it

[RJARRRPCGP]

Re: Power supply build quality pictorial. part 2

Powertech ATX-530W

This power supply has the same name with these: 1 2, but this version is not gutless. I would say it's semi-decent. This is the 2011 version of ATX-530W, the gutless brothers of this psu are 2012 and 2013 versions.

Let's see what makes it not so bad:

-Full AC filtering, only MOV missing

-main transformer real 35 size

-not pathetic heatsinks. Good for 250W continuous, 300W peak.

-2 caps per rail for 3.3V and 5V, 1 cap and a coil for 12V.

-full pi filter for 5vsb

-group regulation coil ok sized for 250W

-13007 transistors and MBR2045 rectifier for 5V and 3.3V, 20F20C3 rectifier for 12V

-and the fan has not seized

But it's dead. The fuse has exploded as you can see in the pictures and 1 of the 4 input diodes that form the bridge rectifier is shorted. There is also some of that glue that turns conductive and foam instead of an insulation sheet bellow PCB that, according to some reports, could get conductive too.

Would you fix it?

One of the transformers, indicates 2009, SMH...

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

Doesn't look like you lost much...

It seems like the +5VSB area is rather close to the mains Input and not even a centimetre between the solderjoint of that cap and mains input.

Looks dangerous.

Big Clive might like it

You are right! I missed that.


Power Innovator LC-B250ATX

This power supply looks like JNC LC-B450E I have posted before.

The secondary heatsink of Power Innovator is smaller to make space for PPFC coil. The coil is real. I have never seen a JNC/Allied/Solytech/Deer/L&C/Premier/Force/Turbo-X power supply with a fake Passive PFC coil.

Unfortunately, Power Innovator LC-B250ATX has no fan controller installed, but there is space on PCB for the parts needed.

On the other hand, there are all the necessary components for AC emi filtering, including an MOV for surges.

There are 6 coils, 1 for every voltage output and 2 caps for 5V, 5vsb and 3.3V.

I have removed 1 cap and glue all over the psu before taking pictures, sorry about that.

I would like to fix it, but the blown fuse indicates it's not gonna be an ordinary recap job.

[Stefan Payne]

Re: Power supply build quality pictorial. part 2

You are right! I missed that.

More pictures of that area pls =)

Power Innovator LC-B250ATX

You might laugh but I have the same unit, more or less...
Though mine looks a bit more trustworthy...

And the fan was a Globe fan IIRC...

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

More pictures of that area pls =)

Sure! The distance is less than 0.3cm (3mm)

You might laugh but I have the same unit, more or less...
Though mine looks a bit more trustworthy...

And the fan was a Globe fan IIRC...

Your unit has larger switching transistors (13009 maybe?), bigger transformers, proper fuse holder, and the thermistor is nicely attached to the secondary heatsink.

It will do 300W, I would replace the primary caps with 680uF caps though.

12V seems to have an insulated 10A fast recovery recifier though, that needs to be replaced with a schottky rectifier rated for 20A or more.

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

Gembird 400W dual fan

When I first saw the innards of this power supply, I decided that it is worth recapping, but now I am not sure..

First, some nice things about this power supply:

-full input filtering, including MOV

-dual fan with add on fan controller board

-PPFC

-good sized transformers. Main transformer size is good for 400W peak, 350W continuous

-big size group regulation coil

-2x E13009 transistors

-SBL3040PT for 5V and 3.3V, F16C20C for 12V

-plastic insulation under PCB

-heatsinks may look bad, but have a lot of fins

Reasons of concern:

-2x 470uF primary caps

-cannot tell who the OEM is. Maybe sun pro?

-It's dead. As soon as I connected it to AC, 5vsb came up, but 4 seconds later I noticed burnt silicon smell coming from the psu. I immediately disconnected it from mains before something worse happened

-obsolete design, 12V rated only 12A. And according to this site, 12V overcurrent protection is set at 144W (12A) which is too low for use in modern systems and upgrading the 16A ultrafast rectifier to 20A schottky wont do any good at all..

[momaka]

Re: Power supply build quality pictorial. part 2

I can swap this for HP-D2537F3H rev 3 for no extra money..

Not sure how much different R model is but here are the pics of D2537F3R by member Spacedye69 and a youtube video by anther member likely of badcaps

https://www.badcaps.net/forum/showpo...1&postcount=88

That looks a lot like my HP-D2507F3P, which I posted a very long time ago here in this thread. It's a very nice PSU: no conductive glue, huge heatinks, 12.5 mm spacing for most output caps, and an offline switch IC for 5VSB generation (i.e. NOT 2-transistor 5VSB). Mine also has passive PFC. All in all, I consider this one of HiPro's best PSUs.

That said, I'm still not sure if you should swap your Delta for that. Both the Delta that you have and the HiPro mentioned above are really good. But with the Delta PSU, you would definitely need to swap all of the caps for the 5VSB circuit, since it's a 2-transistor design. Other than that, it's safe to use.

Of course, if I had to be 100% honest, I'd go with the HiPro.
Just beware that the HiPro is more likely to have its output caps popped than the Delta. But after a recap, those HiPro PSUs are absolutely solid.

i am not sure if "NO 2 transistor for 5vsb" is applicable to H version and whether glue is conductive or not.. , so is this D2537F3H good and fully safe, no 2 transistor 5vsb ?

Yes.

I can pay 2$ more and Get

1) from hp machine- Lite ON PS 5301 08 -HP ( 12v -19A , 5v - 25A )

similar model

https://www.badcaps.net/forum/showthread.php?t=60822

2) from Dell machine - Lite ON L305P 01 ( 12va- 18A max , 12vb max - 18A , 5v-22A .. 12v a and b max combined 274 watts )

https://www.badcaps.net/forum/showthread.php?t=33442

3) From HP machine- D13 180P1A - ( 12va - 7A , 12vb - 10A , 5V -10A , 3.3v- 12A )

E143709 - Chicony ( has same number as HIpro)

input only 200-240V

The Lite-ON doesn't look bad, but I suggest to avoid them, because some of them do come with the conductive glue (but not all). I also don't like the snubber design on the primary PS in some of their models, which uses a 2.2 uF electrolytic capacitor (IMO not a very good strategy).

As for the HiPro above.... nah, recent HiPro/Chicony PSUs are not as good (IMO) as their older units. Yes, they will still do their rated power output. But I think the older units are just built better and have better ripple+noise suppression (thanks to their huge 12.5 mm output caps).

which one should i pick ? any comments, suggestions ?

Either keep the Delta and fully recap it (that means the tiny caps too) or swap for the HiPro HP-D2537F3H. I think these are the best choices, really.

Also, both of those PSUs have a heavier 5V rail, which means you can test legacy (Pentium 3 and AMD socket 462) systems without issues.

I thought the white silicone glue always remained white over time and that the yellow/light brown glue would always turn dark brown/black and conductive.

Well, the white silicone glue in HiPro PSU seems to stay white (dust and dirt aside). Meanwhile, the silicone glue in Delta PSUs appears to either turn beige/cream-colored over time or was just like that from the facroty (I can't tell, because on the most recent Delta PSU I opened, the silicone was a bit yellow inside too). Either way, it definitely was NOT the conductive glue.

[jarvis7]

Re: Power supply build quality pictorial. part 2

That looks a lot like my HP-D2507F3P, which I posted a very long time ago here in this thread. It's a very nice PSU: no conductive glue, huge heatinks, 12.5 mm spacing for most output caps, and an offline switch IC for 5VSB generation (i.e. NOT 2-transistor 5VSB). Mine also has passive PFC. All in all, I consider this one of HiPro's best PSUs.


Well, the white silicone glue in HiPro PSU seems to stay white (dust and dirt aside). Meanwhile, the silicone glue in Delta PSUs appears to either turn beige/cream-colored over time or was just like that from the facroty (I can't tell, because on the most recent Delta PSU I opened, the silicone was a bit yellow inside too). Either way, it definitely was NOT the conductive glue.

a million thanks sir

what about glue on hipro Sir ? in the pics and video , glue is light brown/cream , and on X capacitor it looks identical to my bestec brown glue

https://www.badcaps.net/forum/showpo...1&postcount=88

both the lite on models appear to have white glue..

some more detailed pics of a similar model of lite on Dell L305P-01

This is Dell L305P-00

http://forum.eepw.com.cn/thread/229308/1

Also ,regarding size of caps , as ordering jap caps is expensive here and veryyy hard to find for PSU use , i'd prefer Caps are standard/most common size for PSU , so in case this PSU fails , i can atleast use these caps on another modern PSU ..

Thanksssss

[Dan81]

Re: Power supply build quality pictorial. part 2

Gembird 400W dual fan

When I first saw the innards of this power supply, I decided that it is worth recapping, but now I am not sure..

First, some nice things about this power supply:

-full input filtering, including MOV

-dual fan with add on fan controller board

-PPFC

-good sized transformers. Main transformer size is good for 400W peak, 350W continuous

-big size group regulation coil

-2x E13009 transistors

-SBL3040PT for 5V and 3.3V, F16C20C for 12V

-plastic insulation under PCB

-heatsinks may look bad, but have a lot of fins

Reasons of concern:

-2x 470uF primary caps

-cannot tell who the OEM is. Maybe sun pro?

-It's dead. As soon as I connected it to AC, 5vsb came up, but 4 seconds later I noticed burnt silicon smell coming from the psu. I immediately disconnected it from mains before something worse happened

-obsolete design, 12V rated only 12A. And according to this site, 12V overcurrent protection is set at 144W (12A) which is too low for use in modern systems and upgrading the 16A ultrafast rectifier to 20A schottky wont do any good at all..

Gembird uses Sun Pro usually, along with Leadman's older LP-7700 platform.

Some of their PSUs, once recapped, are pretty good. Not a lot of companies that use Sun Pro that I've seen actually care enough to make something good out of Sun Pros. (yes Delux, I'm especially looking at you.)

[Dan81]

Re: Power supply build quality pictorial. part 2

Some shots of my fixed nJoy 500WGX "500W" PSU. Sorry if the pics look crapp, my LG G3 decided to die unexpectedly (loop of hanging a few seconds on animated bootscreen and restarting)

Only the switchers need replacing with better parts (planning to hunt down some new 2SC3320s) and the caps (I can't trust ChengX, and the primary caps aren't enough for 500W) too but otherwise it looks like a decent PSU.

Also, don't mind the RPC casing. That's because the nJoy casing was so dirty I had to throw it away.

[momaka]

Re: Power supply build quality pictorial. part 2

Powertech ATX-530W
...
But it's dead. The fuse has exploded as you can see in the pictures and 1 of the 4 input diodes that form the bridge rectifier is shorted. There is also some of that glue that turns conductive and foam instead of an insulation sheet bellow PCB that, according to some reports, could get conductive too.

Would you fix it?

Eh... maybe.
Really depends on the output wires. If it has at least 2x SATA, 2x Molex, 1x floppy, and a 4-pin CPU plug... and if the wires are not really thin unbranded/uncertified colorful spaghetti, probably yes.

One thing I really DON'T like about this PSU is how close the input AC line wires come to the secondary side components. Also, while there are 2 caps per output on the 3.3V and 5V rails, I don't see any PI coils, so that's another thing to consider. If if you have some really big capacity caps (like 3300uF or more) in 10 mm diameter and can fit them on the 3.3V and 5V rails and at least 2200 uF cap for the 12V rail, the ripple just might be okay for a very basic low-end PC. I really wouldn't push this unit past 120W for the sake of keeping ripple low. Otherwise, in terms of "raw" power, I'd say maybe 200 Watts continuous and 250W peak, simply because the fuse blew. There's no telling if that was just from marginal design or if the foam turned conductive. On that note, also scrap the foam and put a plastic sheet... or layer the bottom of the PSU with electrical tape (cheap man's plastic sheet ).

[momaka]

Re: Power supply build quality pictorial. part 2

Power Innovator LC-B250ATX
...
I would like to fix it, but the blown fuse indicates it's not gonna be an ordinary recap job.

It's a half-bridged Deer. Doesn't get more ordinary than that, actually.

With the number of other scrap h-bridge PSUs you have, this should be pretty each. All you need to do is check/replace the components in the BJT Base drive circuit and also the secondary-side driver transistors and associated components (typically two diodes and 2-4 resistors as the most critical stuff). I would guesstimate you can do the repair in less than an 1 hour if you know what you are looking for. If not, you just need to figure it out the first time. After that, all h-bridge PSUs will seem very easy to work on.

But of course, start by checking the 5VSB and 5VSB secondary side aux. output rail. Being a Deer/L&C, there's always a critical cap in these. So I usually start by checking/replacing that first, as well as the 5VSB output caps.

Anyways, this PSU looks rated pretty fairly otherwise. That's always good to see.

[momaka]

Re: Power supply build quality pictorial. part 2

You might laugh but I have the same unit, more or less...
Though mine looks a bit more trustworthy...

More trustworthy? With Canicon caps? HA! You almost got me conned.

Sure! The distance is less than 0.3cm (3mm)

Yeah, that's a bit too close to pass safety certifications. Well, the unit might pass due to the gap in the PCB there, which will stop an arc-over from going on continuously. However, the manufacturer should have extended that gap a little more to go around all of those 5VSB traces.

Gembird 400W dual fan

When I first saw the innards of this power supply, I decided that it is worth recapping, but now I am not sure..
...
-It's dead. As soon as I connected it to AC, 5vsb came up, but 4 seconds later I noticed burnt silicon smell coming from the psu. I immediately disconnected it from mains before something worse happened

Probably due to bad caps on the 5VSB circuit. Just because the 5VSB comes up to 5V doesn't mean it's working properly. I have a similar Sun Pro I posted in the gutless thread (as it was a very gutless version of this), and it had the same problem: though the 5VSB appeared to regulate okay without a load, a small load made it whistle a little. 5VSB still looked okay at around 5V, but the secondary side aux. rail that feeds the PWM chip was all over the place ranging from 5 to over 40V. These PSUs have a resistor that limits the voltage/current going to the PWM chip, so when the sec. side aux. goes over-voltage, that resistor tends to overheat and even smoke. So that's something worth looking on your unit.

And yes, this does indeed look like a Sun Pro / Leadman LP-7700 design. I'm not really a big fan of that platform due to the traces separation on the primary side, but so far I have not seen any arc over badly or catch fire because of that. The soldering was also atrocious on mine, with a ton of loose solder balls everywhere. But if yours has that many good components, it's probably not as bad.

Just keep in mind that these tend to have really low efficiency as well. dmill89 just tested a low-end one of these (very similar to mine) with a load, and the efficiency was 65-70% according to him... which frankly is quite low. I think even most Deers are better than that and usually stay in the very low 70% range (possibly peaking up to 75%?)

-obsolete design, 12V rated only 12A. And according to this site, 12V overcurrent protection is set at 144W (12A) which is too low for use in modern systems and upgrading the 16A ultrafast rectifier to 20A schottky wont do any good at all..

Eh. 12 Amps would be okay for a basic Core 2 Duo with onboard video and one HDD. I've ran a power-hungry Pentium 4 Prescott HT on a Bestec ATX1956, which is only rated for 10 Amps on the 12V rail. Also a 200W HiPro that IIRC was rated at 10A or 12A only. No problem whatsoever. The PSUs ran a bit warmer under full load, but handled it okay.

[Dan81]

Re: Power supply build quality pictorial. part 2

Probably due to bad caps on the 5VSB circuit. Just because the 5VSB comes up to 5V doesn't mean it's working properly. I have a similar Sun Pro I posted in the gutless thread (as it was a very gutless version of this), and it had the same problem: though the 5VSB appeared to regulate okay without a load, a small load made it whistle a little. 5VSB still looked okay at around 5V, but the secondary side aux. rail that feeds the PWM chip was all over the place ranging from 5 to over 40V. These PSUs have a resistor that limits the voltage/current going to the PWM chip, so when the sec. side aux. goes over-voltage, that resistor tends to overheat and even smoke. So that's something worth looking on your unit.

And yes, this does indeed look like a Sun Pro / Leadman LP-7700 design. I'm not really a big fan of that platform due to the traces separation on the primary side, but so far I have not seen any arc over badly or catch fire because of that. The soldering was also atrocious on mine, with a ton of loose solder balls everywhere. But if yours has that many good components, it's probably not as bad.

Just keep in mind that these tend to have really low efficiency as well. dmill89 just tested a low-end one of these (very similar to mine) with a load, and the efficiency was 65-70% according to him... which frankly is quite low. I think even most Deers are better than that and usually stay in the very low 70% range (possibly peaking up to 75%?)

Burnt silicon with 5vSB applied usually points to the whole switching transistors and 5VSB switcher being kaput. I had this happen to my PSU I posted above and entirely replacing the transistors (bar one burnt resistor inbetween the transformer and one of the switchers which I replaced, but YMMV and may not be the case there) finally brought it back to life.

Regarding that Deer, I'd point it to either the 4 diodes in the bridge being bad or switching transistors. I had to fix one of these and in that case it was the 4 diodes that were bad.

It did die after a long time of service but the cause was the traces literally popcorned under my 60W soldering iron when I desoldered some wires to work on it (the AC wires).