the gutless, bloated, and fried power supply hall of shame

[Dan81]

Re: the gutless, bloated, and fried power supply hall of shame

Well the transformer core size is credible for 400-500W. With heatsinks, both thickness and surface are matter. Thickness matters for the thermal resistance within the heatsink - can the heat flow from the point of origin to the farthest fins. Surface area is important where exposure to cooling air is concerned.

What of the switch devices and rectifiers?

Current state:


Main switchers - 13007+2N60
Secondary - 12v- 16A rectifier,5v -30A rectifier,3.3v-20A rectifier


This is what it haves now. I might do a complete overhaul of it ('cept heatsinks,those seem thick enough) so those 13007s will probably go as soon as possible.

[PeteS in CA]

Re: the gutless, bloated, and fried power supply hall of shame

Current state:


Main switchers - 13007+2N60
Secondary - 12v- 16A rectifier,5v -30A rectifier,3.3v-20A rectifier


This is what it haves now. I might do a complete overhaul of it ('cept heatsinks,those seem thick enough) so those 13007s will probably go as soon as possible.

Yeah, those 13007s are OK for 250W, but not for 380W. Likewise those rectifiers. I'd suggest 40A devices for the +5V and +3.3V, and 25A for the +12V.

[momaka]

Re: the gutless, bloated, and fried power supply hall of shame

Well the transformer core size is credible for 400-500W.

Concur.

Likewise those rectifiers. I'd suggest 40A devices for the +5V and +3.3V, and 25A for the +12V.

Nah, no need for 40 Amp rectifiers on the 3.3V and 5V rails, unless you really are planning to use this in a dual Athlon MP system with a Radeon 9700/9800 video card, or something equally 3.3V/5V-heavy. 30 Amp rectifiers for the 3.3V and 5V rails is usually more than enough.

Only the 12V rectifier is important for modern computers. That said, there may not be much of a point in upgrading it if the wire on the output filter toroid for the 12V rail is not very thick. And if it's a small wimpy toroid, forget about pulling more than 15 Amps on the 12V rail, if even that.

[Dan81]

Re: the gutless, bloated, and fried power supply hall of shame

Nah, no need for 40 Amp rectifiers on the 3.3V and 5V rails, unless you really are planning to use this in a dual Athlon MP system with a Radeon 9700/9800 video card, or something equally 3.3V/5V-heavy. 30 Amp rectifiers for the 3.3V and 5V rails is usually more than enough.

Only the 12V rectifier is important for modern computers. That said, there may not be much of a point in upgrading it if the wire on the output filter toroid for the 12V rail is not very thick. And if it's a small wimpy toroid, forget about pulling more than 15 Amps on the 12V rail, if even that.

Well,the output filter toroid doesn't seem to be that tall,but it seems quite thick.

As for the output rectifiers,I'm agreeing with that - 30A for both 3.3 and 5V are enough power to work with.

But now,just out of curiosity,how can one make a 5v heavy PSU (like those Deer,Leadman,Sun Pro etc.) into a 12v heavy PSU for modern applications?

[PeteS in CA]

Re: the gutless, bloated, and fried power supply hall of shame

Well, if you aren't going to be loading the 5 and 3.3 near their rating, 30A would be fine. OTOH, 40A parts would also improve efficiency.

What momaka said about the inductor wire gauge is true. You have to look at all the various things that impose practical limits on realistic current capacity - switch devices, main transformer core and windings, rectifiers, inductor wire gauge, PCB traces, O/P harness wire gauge, O/P connector terminals.

[momaka]

Re: the gutless, bloated, and fried power supply hall of shame

Well,the output filter toroid doesn't seem to be that tall,but it seems quite thick.

If you want higher power output, the output toroid needs to be both big and have thick wires on it.

But now,just out of curiosity,how can one make a 5v heavy PSU (like those Deer,Leadman,Sun Pro etc.) into a 12v heavy PSU for modern applications?

Put a Schottky rectifier on the 12V rail (with a 60V rating or more). Schottky diodes have 0.4-0.7V lower Vf (forward voltage drop). This helps to keep the 12V rail higher under load. That said, once you modify a 5V-heavy PSU with a Schottky diode on the 12V rail, it may no longer be possible to use the PSU for a 5V-heavy PC until you remove the swapped Schottky on the 12V rail and put back the original fast-recovery (FR) rectifier. For example, my Inno Power MPT-301 (by Macron) used to have very balanced 5V and 12V rails both in 5V-heavy and 12V-heavy PCs. But I was worried about the small heatsinks with a heavier 12V load and I wanted better efficiency, so I swapped the original F16C20 FR rectifier on the 12V rail with an STPS20s100 (20 Amp Schottky). This made the voltage on the 12V rail go up quite a bit. On a 12V-heavy PC, the 12V rail went from 11.98V to 12.3V or more. And on a PC with a heavier 5V load, the 12V rail actually went over spec at 12.6-12.7V (which is something that never happened before).

Now, if you do mod a 5V-heavy PSU with a Schottky rectifier on the 12V rail, and the 12V rail still seems weak, that's probably a sign that the PSU just isn't made for a 12V-heavy load.

Another thing you can do - but this is only if you are going to use the PSU with a 12V-heavy PC that puts minimal load on the 5V rail - is swap the 5V Schottky rectifier for a FR rectifier and swap the 12V FR for a Schottky. This should bring the 12V rail up even more and lower the 5V rail further. But don't even dare use such a modded PSU on a 5V-heavy PC.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

Yes, I knew I've seen the D304X before. In Linkworld! It almost melted but delivered 295 W before the OPP kicked in. OPP of this "420W" unit is set to 300 W

[PeteS in CA]

Re: the gutless, bloated, and fried power supply hall of shame

Yes, I knew I've seen the D304X before. In Linkworld! It almost melted but delivered 295 W before the OPP kicked in. OPP of this "420W" unit is set to 300 W

A very amusing review. Sounds to me, though, that the turns ratios of the main transformer didn't allow the PSU to deliver full rated power at full pulse width.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

The turns ratio between the secondary taps are particularly wrongly set as even according to its own power distribution table you can see what the resulting regulation is…what about the primary vs. secondary ratio, I can not tell. I still have it but have no desire to undwind the transformer ;-) Will disassemble it for parts one day or something like that.

[Dan81]

Re: the gutless, bloated, and fried power supply hall of shame

I think I have a candidate for the worst power supply (relative to "rated" wattage). Ok, maybe not as bad as an A-Power or Echo Star (or powmax), but it's pretty crap in every aspect.

We have a dead Okia 550W unit, seems to be a 5VSB issue.

The even funnier part is that they reused an Deer casing (the pattern on the back and the voltage switch,along with the AC receptacle and switch were the giveaway. Oh,and the fan grille being built to be screwed INSIDE the PSU)

[Pentium4]

Re: the gutless, bloated, and fried power supply hall of shame

Check out this nonsense! I spotted this after testing the PSU. It powered up just fine. When I carefully removed the screw, the Y cap lead instantly broke! It was melted to the Y cap, so I can't really tell if this happened during manufacturing or if it fell in there. I think it happened in the factory, because it looks really wedged in there and I think it would be difficult for a screw to fall in like that.

[momaka]

Re: the gutless, bloated, and fried power supply hall of shame

Oh Deer! ????

Probably the screw made Live arc to Ground through it through the Y-cap leads, and that's why one of the leads melted to it. I've used this very same method to weld thermocouple wire together. (Though, I did it through a space heater in series, so the current was more controlled.)

[jazzie366]

Re: the gutless, bloated, and fried power supply hall of shame

Just realised all the glue on the caps, but they're going to need more to stop them leaking.

Those input filter caps, the HV ones on the primary with the ,"F" inside of what looks like saturn's rings. What brand is that? I know they're absolute shit because Sans Digital uses them in their PSUs and they fail in under a year to up to 3 years was the longest they lasted. The company also tries to charge you 100$ for a new PSU, meanwhile they use the same shitty caps. However, the PSU other than the caps is build very well, lots of filtering on the input and output with a beefy transformer mean a happy psu.

[jazzie366]

Re: the gutless, bloated, and fried power supply hall of shame

It's ATX, it has 3.3v (the orange wires) and 5vsb wire (purple) as well as ps-on (green wire) and ps-ok (grey wire).

The 5vsb is probably produced by the circled transistor or maybe IC in to-220 package.

Where did you find this!? That's a KenTek PSU which the manufacturer of is currently unknown, they should be found and shamed and also made liable for all the PCs that have been destroyed by them.

[Dan81]

Re: the gutless, bloated, and fried power supply hall of shame

Where did you find this!? That's a KenTek PSU which the manufacturer of is currently unknown, they should be found and shamed and also made liable for all the PCs that have been destroyed by them.

If anywhere on the Kentek it does say 8868 (or a LP- marking),it's a Leadman of the crappiest kind.

[goontron]

Re: the gutless, bloated, and fried power supply hall of shame

Not much detail here... My media center/file server started locking up. It had failed POSTs, random segfaults, and udev hangs for a time before this incident. Those were signs i was ignoring. Today it hard locked, upon reboot it took 45 minutes to finish POST. I go through the usual steps, none work. So i pull out my trusty Tek 'scope and probe things. 3.3v is good. 5v has .3v of ripple, not so good. I get to the 12v rails and they have .5v of ripple.... HALF A VOLT! I cracked this thing open. Not the worst, but not good for a 750w PSU.

To start off, Blackberry camera shot of the nameplate....

mushkin MKNPSJL700.

Now for the overhead

Does have fan control. Heasinks are a little anorexic. Brown snot. Asia X caps.

Solder side up!


Not bad....

Lifted pad on the PFC cap... (Again, sorry, i need to get my Moto fixed. This BB is shit)

Damn... There goes the idea of fixing it...

[momaka]

Re: the gutless, bloated, and fried power supply hall of shame

Does have fan control. Heasinks are a little anorexic. Brown snot. Asia X caps.

I wouldn't say the heatsinks are anorexic. If the PSU is efficient, those heatsinks should be good enough for the ratings. The Antec EA-500 I have has much smaller heatsinks than that, and it will do its 500 Watt rating no problem.

The crap caps and the tan/brown conductive glue are a bummer, though - especially the glue.

Damn... There goes the idea of fixing it...
https://www.badcaps.net/forum/attach...1&d=1486176440

Wait, you are not fixing the PSU?
Seems like a decent unit. I would, at least. New high quality caps and an hour of work are well worth the peace of mind in knowing that you now have a PSU that won't crap out on you... well, IMO, anyways.

[Wester547]

Re: the gutless, bloated, and fried power supply hall of shame

It looks like the PCB snapped just a tad when he removed it from the PSU chassis. That's what he meant by not being able to fix it anymore.

[Agent24]

Re: the gutless, bloated, and fried power supply hall of shame

Doesn't look that bad to me... I fixed a radio PCB that was completely snapped in half. Had to run about 50 wires but it worked. I can't see what's so bad about this one?

[goontron]

Re: the gutless, bloated, and fried power supply hall of shame

I will fix it I suppose. I don't like how the board is warped...