Power supply build quality pictorial. part 2

[Behemot]

Re: Power supply build quality pictorial. part 2

http://www.youtube.com/watch?v=FytMNBUR2CQ

[Heihachi_73]

Re: Power supply build quality pictorial. part 2

I'm assuming it's only 85, since there is no rating on the sleeve.

[Heihachi_73]

Re: Power supply build quality pictorial. part 2

A very dusty Macase MG-250EP AT 250W unit.

Board is dated 04/01/95, QC test dated 04/12/97 and the internals is covered in cobwebs and spider exoskeletons. "Quaranty Lapses After Removal"

The insides started off looking good with a Delta AC socket and Panasonic primaries, as well as sockets for the AC input and fan, then went downhill from there by exclusively using Fuhjyyu throughout the rest of the board. Eighteen of them. Surprisingly, none have bulging tops yet, except for one suspect-looking primary (although this could simply be the plastic 'lid' on the cap). I just can't see how a Panasonic could be outlasted by those tiny Fuhjyyus though!

C5: Matsushita UP 470uF 200V 85C 6928
C6: Matsushita UP 470uF 200V 85C 6929C
C10: Fuhjyyu TNR 1uF 50V 105°C
C11: Fuhjyyu TNR 1uF 50V 105°C
C14: Fuhjyyu TNR 220uF 25V 105°C
C15: Fuhjyyu TNR 220uF 25V 105°C
C16: Fuhjyyu TNR 2200uF 16V 105°C
C17: Fuhjyyu TNR 1000uF 16V 105°C
C18: Fuhjyyu TNR 470uF 16V 105°C
C19: Fuhjyyu TNR 470uF 16V 105°C
C20: Fuhjyyu TNR 1uF 50V 105°C
C21: Fuhjyyu TNR 10uF 50V 105°C
C24: Fuhjyyu TNR 47uF 35V 105°C
C25: Fuhjyyu TNR 4.7uF 50V 105°C
C26: Fuhjyyu TNR 4.7uF 50V 105°C
C27: Fuhjyyu TNR 47uF 50V 105°C
C28: Fuhjyyu TNR 4.7uF 50V 105°C
C29: Fuhjyyu TNR 2.2uF 50V 105°C
C91: Fuhjyyu TNR 10uF 50V 105°C
C92: Fuhjyyu TNR 10uF 50V 105°C

I still have the original tower case for it as well, and it has enough slots to house five hard drives (six if you swapped out the floppy). Two hard drive bays rest right underneath the PSU. There are three 5.25" drive bays.

It would have been a decent server case back in its day, however being of AT design there isn't really any use for it now - you can't even fit a MicroATX board in there because it fouls the rear HDD/PSU area (I wouldn't care about dremeling out the keyboard port to fit the back of an ATX mainboard).

[Shocker]

Re: Power supply build quality pictorial. part 2

The insides started off looking good with a Delta AC socket and Panasonic primaries, as well as sockets for the AC input and fan, then went downhill from there by exclusively using Fuhjyyu throughout the rest of the board. Eighteen of them. Surprisingly, none have bulging tops yet, except for one suspect-looking primary (although this could simply be the plastic 'lid' on the cap). I just can't see how a Panasonic could be outlasted by those tiny Fuhjyyus though!

Especially when the Fuhjyyus are TN series (I think the R is meaningless), which are General Purpose!!!!!

Anyway, try pushing down on the plastic top. If it goes flat, they're probably fine.

[Heihachi_73]

Re: Power supply build quality pictorial. part 2

FSP200-60ATV 200W OEM power supply from my Acer T310 desktop. This PSU worked without failure running 24/7 for 6.5 years (late Dec 2005) up to June 3 2012, where a power cycle rendered it unable to do much more than turn on the power LED for a second (after draining the leftover power from the caps and turning it back on, it was completely dead). There were no faults in operation prior to shutting down the PC to add/test a PCI SCSI card.

In the following images, the fan has been removed - I clipped the wires since I couldn't be bothered desoldering it. The 80MM fan is a Yate Loon Electronics D80SM-12 (0.14A), which didn't actually spin that fast compared to other 12V fans I have (Globe fans are quite fast by comparison).

Every single cap is from Teapo, with the exception of a single Jamicon hiding amongst the transformers. Even the miniature one on the fan control board is a Teapo 10uF 35V. There are four bulging caps, probably more have failed but are not physically deformed. There were two resistors (R47, next to possibly-failed 1uF 50V C39; R51, next to failed 3300uF 10V) which had got very hot during its last minutes of use, leaving scorched dust marks on the case, and the glue around R47 and its miniature friends shown in picture 7 had turned brown and crispy (I have since removed it). No PCB markings yet, since there's that nasty yellow glue all over the board; I can't make out most locations and there doesn't seem to be a schematic for the 200W version.

The big boys are Teapo LXK 470uF 200V 85°C 05/04.

[Wester547]

Re: Power supply build quality pictorial. part 2

Nice power supply there.

In the following images, the fan has been removed - I clipped the wires since I couldn't be bothered desoldering it. The 80MM fan is a Yate Loon Electronics D80SM-12 (0.14A), which didn't actually spin that fast compared to other 12V fans I have (Globe fans are quite fast by comparison).

That might be because the D80SM-12 maxes out at 2600RPM/34 CFM, so I doubt (being a thermally varying fan) that it ever gets much faster than 2000RPM. While usually I would opine that's too low (and it may be for a power supply in the long run, at least an average one with Teapos or equivalent capacitors, as many 80mm PSU fans I see are rated at least 3000RPM/38 CFM), sleeve bearing fans definitely last longer when they don't run at such high speeds. I also consider Yate Loon a good brand of fans.

Every single cap is from Teapo, with the exception of a single Jamicon hiding amongst the transformers. Even the miniature one on the fan control board is a Teapo 10uF 35V. There are four bulging caps, probably more have failed but are not physically deformed. There were two resistors (R47, next to possibly-failed 1uF 50V C39; R51, next to failed 3300uF 10V) which had got very hot during its last minutes of use, leaving scorched dust marks on the case, and the glue around R47 and its miniature friends shown in picture 7 had turned brown and crispy (I have since removed it). No PCB markings yet, since there's that nasty yellow glue all over the board; I can't make out most locations and there doesn't seem to be a schematic for the 200W version.

At least there aren't any Fuhjyyu in there. I've seen Fuhjyyus lay death to otherwise great FSP PSUs before. :P But I suppose Jamicon isn't much better. Does that yellow glue pose a threat to PSUs? I've seen that stuff in Huntkey PSUs (Dynex rebranded, in this case) before.

The big boys are Teapo LXK 470uF 200V 85°C 05/04

Considering I usually see 680uF/200V Teapo primaries in 300Ws from FSP (and I imagine 560uf/200V primaries in their 250W PSUs), that works quite well.

If you don't mind me asking, how many power cycles did that machine have with that PSU in it (in total)?

[c_hegge]

Re: Power supply build quality pictorial. part 2

Every single cap is from Teapo...There are four bulging caps, probably more have failed but are not physically deformed.

And that would be why I hate teapo.

[Heihachi_73]

Re: Power supply build quality pictorial. part 2

If you don't mind me asking, how many power cycles did that machine have with that PSU in it (in total)?

Probably less than ten per year (mostly power outages, the only time I have ever shut down the PC via Windows is to install graphics cards/RAM etc). The amusing part is that it died during the start of Australia's winter rather than the middle of summer. The original 40GB hard drive is still the C: drive, which has clocked up 51781 (CA45 in hex) hours according to SpeedFan. Apparently the hard drive has a power cycle count of 465 (01D1) though - I did turn it off daily during the first year or so however, then I got sick of it since I couldn't download overnight!

So far my next [victim] PSU has been on 24/7 since June 3 (although it's a 2010 model with everyone's favorite Fuhjyyu primaries visible through the honeycomb grille at the rear). See how long it fares, even if it makes 2 years it will be long enough for me to upgrade from this dinosaur to the next generation of PCs. I'm also thinking of dropping Windows like a ton of bricks, since my reaction was when I was loaned a Windows 7 laptop overnight earlier this year. I already can't get used to the default XP themes, the first thing I do with a junk-find PC is change it to the 'Classic' Windows 2000 look! Kubuntu time, if the GUI doesn't crash every 5 minutes and kick me to the bash screen whilst failing to reload the interface without a reboot (bad experience with Kubuntu 10 on an IBM ThinkCentre 8191KMQ - both the full install and the live DVD did the exact same thing).

[SIDMX]

Re: Power supply build quality pictorial. part 2

Delta DPS-266AB Rev.01 (266W Max.) ...custom size PSU from VAIO PC

I own this PSU it is right now powering a P4 based PC (Cedarmill), this photos are before the recap ...like 7 months ago, sorry didn't take pics after recap was done, maybe next time it need some cleaning.

Bridge Rectifier is GSIB 1560
Primary Cap 1x 1500uf/200V Nichicon GN Series
Primary Side MOSFET W20NK50Z
5VSB Switcher TNY267P
Schottkys:
2 x MBR3060PT
2 x STPS3045CW
1 x STPS20100CT

Didn't have the time to trace what rail use each schottky...any guesses here?

Original caps:

+5VSB - 1 x 2200uf/10v Taicon PW - 1 x 1000uf/10v Ltec LZG

+3.3v -2 x 2000uf/6.3v Rubycon ZL

+5v - 2 x 2000uf/10v Ltec LZG

+12v -1 x 2200uf/16v Rubycon ZL - 1 x 1000uf/16v Rubycon ZL


Caps used for recap:

+5VSB - 1 x 2200uf/10v Nichicon HE - 1 x 1000uf/10v Rubycon ZL

+3.3v -2 x 2000uf/6.3v Rubycon ZL

+5v - 2 x 2000uf/10v Rubycon ZL

+12v -1 x 2200uf/16v Rubycon MCZ - 1 x 1000uf/16v Rubycon ZL
I know... overkill but that was the only one I have at hand in that moment.

[c_hegge]

Re: Power supply build quality pictorial. part 2

I doubt if you had to replace the Ruby ZLs. Even junk caps like CapXon and LTEC usually do OK in delta power supplies

[goodpsusearch]

Re: Power supply build quality pictorial. part 2

MCZ is a very bad choice for psu recapping.

[SIDMX]

Re: Power supply build quality pictorial. part 2

I doubt if you had to replace the Ruby ZLs. Even junk caps like CapXon and LTEC usually do OK in delta power supplies

True, but date codes from all Rubys were 2003 (T034x-S034x) they already earned indefinite vacations don't you think ?

MCZ is a very bad choice for psu recapping.

Indeed, I will never do anything like this in a customer psu, i know (learned it here in BC) that impedance/ESR is important for proper output filter tune but i needed to assemble a second PC without spend any money and this was the only way, all parts were left by customers who don't care about remaining parts after i upgraded their PCs. Is there any other thing that i'm missing here? (other than higher ripple in output beacuse of MCZ's ESR)

[Pentium4]

Re: Power supply build quality pictorial. part 2

And that would be why I hate teapo.

To be fair, they lasted 50,000+ hours. But yes I'm sure a Japanese cap would still be working, still a good life if you ask me.

[c_hegge]

Re: Power supply build quality pictorial. part 2

Yes, but most of the failures I see happen much sooner.

[Pentium4]

Re: Power supply build quality pictorial. part 2

This PSU is out of my parents Dell rig. Pentium D 3.4GHz/4M/800, 2.5GB 667, Reference GeForce 6800GT 256MB, 64GB SSD. I remember setting this thing up for my parents when it arrived on our doorstep on 3/3/06. I figured it was time to check the PSU and make sure that all the caps were okay..plus I was curious I thought it was going to be a Delta unit....but it looks more like a Liteon PSU...that big ass thing, is that passive PFC?


I was surprised to see 820uf primary caps. Unfortunately they're OST @ 85C...


All the secondary caps are OST and Ltec. Looks like they have some pressure between the vent stamps but not a full on bulge. Should still be working okay?







I'm surprised to see the caps in pretty good shape after almost 6 and a half years, about 8 hours of use a day. Only been cleaned a few times

[b700029]

Re: Power supply build quality pictorial. part 2

Those might not be real bulges, but an artifact of the stamping process that creates the tops of the cans.

Still, given the brand, I wouldn't be surprised if they were.

[dmill89]

Re: Power supply build quality pictorial. part 2

This PSU is out of my parents Dell rig. Pentium D 3.4GHz/4M/800, 2.5GB 667, Reference GeForce 6800GT 256MB, 64GB SSD. I remember setting this thing up for my parents when it arrived on our doorstep on 3/3/06. I figured it was time to check the PSU and make sure that all the caps were okay..plus I was curious I thought it was going to be a Delta unit....but it looks more like a Liteon PSU...that big ass thing, is that passive PFC?

Definitely a LiteOn. Can tell from the model # L375P (the L at the front means LiteOn). Yes that big coil on top is for the passive PFC. The caps look ok (although OSTs are famous for failing without any visual signs), the "bulges" between the vents appear to be from the stamping process. However Ltec and OST are not particularly great brands (far from the worst but mediocre at best) and that secondary appears to run relatively hot, there is little room for airflow, and caps are right up against hot components such as coils and heat sinks (those 2 small Ltecs right against the heat sink could definitely be cause for concern). These Dell units with the fan in the front (inside of chassis) don't get the best airflow and do tend to run a little hot.

[Pentium4]

Re: Power supply build quality pictorial. part 2

Definitely a LiteOn. Can tell from the model # L375P (the L at the front means LiteOn). Yes that big coil on top is for the passive PFC. The caps look ok (although OSTs are famous for failing without any visual signs), the "bulges" between the vents appear to be from the stamping process. However Ltec and OST are not particularly great brands (far from the worst but mediocre at best) and that secondary appears to run relatively hot, there is little room for airflow, and caps are right up against hot components such as coils and heat sinks (those 2 small Ltecs right against the heat sink could definitely be cause for concern). These Dell units with the fan in the front (inside of chassis) don't get the best airflow and do tend to run a little hot.

That's why I'm amazed nothing has died yet! Yesterday was its first full cleaning...and that PFC thing sits on top of the secondary. But since they normally just surf the internet, they don't see too much load. Even when that Pentium D is at high usage, it's not working it that hard since it's rated for 360W continuous on the 12V. Been a pretty good PSU. But yeah, I really don't like the placement of the fan at all, just doesn't seem very effective. Nice beefy heatsinks though.

[Behemot]

Re: Power supply build quality pictorial. part 2

Whats its rotation speed? These things often do not have speed controll and such a fan on full speed can do lot of airflow.

[Wester547]

Re: Power supply build quality pictorial. part 2

Nice power supply there, overall.

I was surprised to see 820uf primary caps. Unfortunately they're OST @ 85C...

Aren't many primary caps rated at 85C (if I'm not mistaken)? ^^; I don't think that would bee too much of a problem in the long run (provided the primary caps aren't on the echelon of Crapxcon or Failujju ) because (correct me if I'm wrong on this as I may well be), to my knowledge, primary capacitors are not subjected to as much heat as secondary capacitors (or as much of a risk). Nevertheless, the heatsinks are very close to the fan, too close for my tastes. Usually I see the heatsinks further away from the fan than that in PSUs. Anyway, every old (and even high quality) PSU I've seen from say, Newton Power, Astec, Hipro, or whichever brand has had the primary capacitors rated at 85C, so I don't think it should be too much of an issue if ever.

All the secondary caps are OST and Ltec. Looks like they have some pressure between the vent stamps but not a full on bulge. Should still be working okay?

Well, an old Dell Dimension 8200 (from the end of 2001) I have has a NPS-250GB (Newton Power) with a mix of Nichicon and Ltec capacitors (and Taicon as well; Nichicon on the primary, Ltec/Nichicon/Taicon/maybe Rubycon on the secondary), along with a Sunon sleeve bearing fan (, but at least it's a powerful 80mm fan, rated at 3000RPM/40CFM, and the heatsinks are very thick), and it's never given me any sort of instability (though the fan does have an usual rattling sound after 20 minutes of being on, but it's been that day since day one). It even becomes a bit hot at the fan grille after a while of intense usage. It's had ~17,000 hours of on time and ~4,500 power cycles, still doing fine (though the fan sounds a bit worse for wear so I won't be surprised if it seizes at any fleeting moment). I see no indications of failed capacitors in that PSU, though it doesn't mean they can't fail very silently, of course, but hopefully that puts your mind to ease. If I'm not mistaken (PCBONEZ said something about this, to my recollection), in 2001, Ltec 'stole' a formula from Rubycon and that improved upon the reliability of their capacitors. Even then, I would imagine the longevity of their capacitors depends greatly upon heat and stress as with any mediocre capacitor.

I'm surprised to see the caps in pretty good shape after almost 6 and a half years, about 8 hours of use a day. Only been cleaned a few times

As has the Newton above. The fan still looks kinda baked with dust. :-P

One other thing I noticed is the yellow glue on those primary capacitors, in your Liteon. Is it safe to have that in a PSU? I've heard it can eventually eat components off the PCB, kinda like the tan/brown glue in some PSUs, though I hope that doesn't hold so (not to scare anyone, just something I read).

Also, even though the heatsinks are bunched up together, that fan does look and sound nice. What brand is it (if I may ask/if you know)?