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

[Pentium4]

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

I want to replace the 12A fast recovery diode on the 12V with a 20A schottky, would that throw it out of whack at all? Cause I think that would give it a little efficiency increase. I also have a transformer from a Bestec (different topology) that lines up perfectly pin wise, not sure if it would work though. It doesn't have a fat grounding wire coming out of it

[Behemot]

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

From my experiments, I think you'd need to modify the other silicon as well. What's the rest now, on +3,3 and +5 V, are both shottkies or superfasts?

If there is separate ground, it is most likely half-bridge. You need to keep the topology, the windings for different topologies are completelly different, look at http://pe2bz.philpem.me.uk/Power/-%2...ps.htm#forward and further.

[370forlife]

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

I believe it depends on the topology. I have never had a problem replacing fast recovery diodes with schottky's but those are almost always found in half-bridge units. If the unit uses anything other than half-bridge though, they typically will not cheap out as much and use actual schottky's on all rails. (obvious exception being Leadman and their various spawn,) but even then I use a single transistor forward LP-7700 with a 20A schottky on the 12v rail as a lab supply with no problem.

EDIT:

Regarding transformers, they hardly ever interchange. Typically only the cheap half-bridge units have interchangeable transformers across manufacturers. Not sure why, maybe its just the best way to do it, or they just all copy eachother.

[Pentium4]

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

Thanks for the replies! The 5V has a 30A schottky and the 3.3V has a 20A schottky, so one for the 12V should work perfectly? So I'm stuck with this transformer? Shoot so if it's only good for 250W I should leave the 13007's in there too. I've soldered on a SATA cable, a 4 pin (to make the main plug a 24 pin) added a 0.47uF X cap, 2 huge Y caps, and replaced the input coil with a bigger one. I also plan on replacing the 470uF CS input caps

[Behemot]

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

For soem low-power system, why not…

If the +12 V goes too high (though extra 0,6-0,7 V should not matter that much here), add other rectifiers in parallel with those on +3,3 and +5 V, it will lower their's drop and should bring voltages to normal.

[Pentium4]

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

For soem low-power system, why not…

If the +12 V goes too high (though extra 0,6-0,7 V should not matter that much here), add other rectifiers in parallel with those on +3,3 and +5 V, it will lower their's drop and should bring voltages to normal.

Thanks, I will try that tomorrow for the 12V, would a higher rated rectifier increase efficiency over a 20A? Like say if I had a 30A or 40A schottky there even though I wouldn't want to push it past ~250W

[Wester547]

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

Thanks, I will try that tomorrow for the 12V, would a higher rated rectifier increase efficiency over a 20A? Like say if I had a 30A or 40A schottky there even though I wouldn't want to push it past ~250W

It depends on what the datasheet says about both rectifiers. A number of things have to be taken into account: the voltage drop (voltage drop * amps = how much power is wasted), the temperature both rectifiers are rated at, the junction-to-case (or sink) thermal resistance, etc... but I don't think going higher would hurt for a PSU like this.

[Pentium4]

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

It depends on what the datasheet says about both rectifiers. A number of things have to be taken into account: the voltage drop (voltage drop * amps = how much power is wasted), the temperature both rectifiers are rated at, the junction-to-case (or sink) thermal resistance, etc... but I don't think going higher would hurt for a PSU like this.

Thanks for the input, I will check out the datasheets tomorrow I know they are high quality parts, and either of them is a big improvement over a MOSPEC 12A fast recovery diode...

[Wester547]

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

MOSPEC actually do produce quality parts (assuming they are of a decent current rating), it's just the diodes on brackets and noname manufacures that you want to avoid.

[Heihachi_73]

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

A cooked 200W Sirtec High Power SI-A200M2 dated 0333, found in an Acer Pentium 4 of the same vintage. The PC's case was of ordinary quality too, when I took it apart late last year I ended up slicing my finger on the razor-sharp edge at the rear of the case. However, the mainboard itself still works after testing another PSU (an almighty 110W HP OEM Delta DPS-110MB full of Chemi-con LXJ and two Rubycon 200V 470uF mains caps), and booted into Windows XP like nothing happened.

Cause of failure: Seized Sunon KD1208PTS2 fan, cooked the PSU and took many caps with it, although they were probably failing already going by the brand. The PSU squeals when powered on and won't even spin a replacement fan.

C7: Teapo SK 560uF 200V 85°C 07/03
C8: Teapo SK 560uF 200V 85°C 07/03
C18: Teapo SEK 10uF 50V 105°C 06/03
C46: Teapo SEK 10uF 50V 105°C 06/03
C48: Teapo SEK 10uF 50V 105°C 03/01 [old stock?]
C100: Teapo SC 2200uF 16V 105°C 02/03
C102: Teapo SEK 1000uF 10V 05/03 [bulged, actual uF/volt rating obscured by glue, but looks identical to C250]
C103: Teapo SC 2200uF 10V 105°C 07/03 [bulged]
C250: Teapo SEK 1000uF 10V 105°C 05/03 [bulged]
C251: Teapo SEK 100uF 25V 105°C [bulged]
C253: Teapo SEK 10uF 50V 105°C 06/03
C303: Teapo SC 3300uF 6.3V 105°C 05/03
C501: Jenpo CE-WR 0.47uF 50V 105°C
C503: Teapo SEK 0.22uF 50V 105°C 06/03 [cooked glue on top]
C508: Teapo SEK 0.22uF 50V 105°C 06/03
C901: Teapo SEK 47uF 50V 105°C 06/03
C904: Teapo SC 1000uF 16V 105°C 06/03 [bulged]
In fact, even the three X caps are Teapo...

[Wester547]

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

And that X capacitor is rated at 85*C.... but I guess it doesn't make much of a difference whether X capacitors are rated at 85*C or 105*C?

[PeteS in CA]

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

Probably polypropylene dielectric. You'll see the same rating regardless of the brand, as it's a characteristic of polypropylene. IIRC, polyester can go a bit hotter.

[Heihachi_73]

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

For those interested, the PC was an Acer Veriton 5600G. As I don't have any room for another Pentium 4 system, and the fact it's a desktop case which can't fit a standard sized ATX power supply, this 200W PSU will be scrapped rather than repaired.

I forgot to mention, this PSU had surface-mounted stuff on the underside, including an 8-pin TL3843.

All is not lost though, I will gain a bridge rectifier (amazingly it didn't have 4 diodes) and a handful of coils, TXs and other goodies. The big 560s might be a keeper as well, depending on their condition. I might also have a bit of fiery fun in the background c_hegge style, courtesy of Teapo.

[c_hegge]

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

^
Hehehe. have fun. make sure to get the tiny vent-less caps. Those always go kapow, rather than fizzle

[Heihachi_73]

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

I find it amusing that Teapo themselves approve of this in section 1-11 of their own catalog.pdf (formerly "GENERAL INFO SEARCH.pdf").

"Teapo smoke, don't breathe this"

Random findings... there were three different bung designs on the Teapo caps I pulled. One type had multiple rings inside one another, which seems to be the most common design (2200uF 10V), another had a 'split' bung similar to Nichicon (3300uF 6.3V), and the third had a single circle on a flat bung (2200uF 16V).

[tleu8472]

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

Referring back to my earlier post #412: I finally got around to recapping this FSP unit, however now it does not start up! I think it used to start up before I recapped it. What should I check for - a blown fuse or a short somewhere? Any advice would be appreciated.

[Behemot]

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

Short, accidentally reversed cap, bad cap type.

[Pentium4]

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

The curiosity got the best of me and when I saw this brand newly listed on Tigerdirect (Manufacturer associated with it was Coast to Coast Fulfillment on the website) I bought the cheapest one ($19.99).....I knew it would suck but I had some money to burn... The label has so many lies on it. APFC? Uh-huh....78% + efficiency? There's no way.... It has a 16A part for each rail, only fast recovery is the 12V....on the other side of the fast recovery there is a slot for another rectifier. The heatsinks are so bad though! This thing would get really hot around 200W I'd imagine. Transformer is maybe size 28.

There are no PI coils for the main rails but there are spots there for them. All the secondary caps are Asia X.

470uF Sun primary caps, no MOV's, 2 X & Y caps (Slots for 2 more Y), one coil (If you look on the box it shows 2 input coils and 2 different X caps through the grille....), four 2A diodes, they're pretty small though. Knock-off 13007's. It has an Ever Cool fan, and the thermal sensor for it is glued onto the larger toroid coil. Wires are 20AWG

[cheese007]

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

I like how the power supply on the box isn't the one inside. Notice how the lack of filtering on the one that the user actually receives.

[cubecompmtdx]

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

I figured this unit would be a nice addition to the hall of shame. This is an Allied (Deer) AL-A300ATX unit (many different revisions carry that model number, I have an old pink labeled one kicking around). Decided to pop the cover off this power supply to inspect it. While these Allied/Deer power supplies are oftentimes known to have bad capacitors, this one was either

a. overloaded
b. improperly cooled, or
c. both

This unit does have a thermal fan controller, and the thermistor for that fan controller wasn't touching the secondary side heatsink like it's supposed to. I believe the fan never sped up to a fast enough speed to keep things cooled. As a result, many components, particularly diodes and resistors, got way too hot, causing massive burn marks and discoloration in the PCB. The fan is not seized, though I don't know if it was getting power or not. Almost all the capacitors on the secondary and 5vsb areas are bulging/leaking. It's a good thing my uncle took this unit out of service, because not only was it a risk for the attached hardware... this thing has become a fire hazard! My original plans were to replace the bad capacitors, but after seeing how things look with the cover removed, I'm just going to harvest the useable parts from it. Also, I noticed there was a 5 amp fuse installed. The PCB actually states that a 5A fuse should be installed in a 235W unit!