On Sun, 25 Mar 2007 14:59:16 -0500, "Skeleton Man"
<invalid@guestwho.com> wrote:

>>So how much work are you willing to do on a PSU that, once
>>repaired, may have the same limitation as before and other
>>weaknesses that can similarly cause failure?
>
>If it's something I can accomplish with mininimal cost and labor I want to
>do it as an exercise more than anything.. if it works I'll keep it as a
>low-power spare.. (It wasn't meant for this kinda load anyway.. even the
>mainboard/cpu info reccommends a minimum of 450W..

That's pretty much a made-up number they list then, it has
everything to do with what else is plugged into the system.
The board/CPU alone would run from a 200W mATX. However, a
Wintech 350W may not be as fit to run a load long-term, as a
good 200W PSU would be.



>
>I have no problem taking measurements from a live supply.. a have two DMMs
>(one has frequency, diode, transitor, and capacitance) and a small analogue
>meter.. the only thing I don't have is a scope.. do they make handheld
>versions of these like they do multimeters ? Also can you get LCD/digital
>scopes as opposed to a CRO ?

Yes, there are handheld w/LCD, you'd have to look at what's
available in your price range to see what sacrifices you'd
make. Unless you really "need" it portable I suggest a
desktop model. If nothing else it will keep it from being
dropped or stolen (so easily) by staying on your bench.


>
>Here's the breakdown of the specs (WIN-350PS):
>
>+3.3V = 15A
>+5V = 25A
>+12V = 15A
>-5V = 0.6A
>-12V = 0.6A
>+5VSB = 5A
>
>Combined 3.3v/5v rails = 150W max

Ok but it's a Wintech, we can pretty much ignore those
numbers except to see that it may be more like a 250W PSU
than 350W, but one probably built as cheaply as possible in
some areas.


>
>The voltages were all well within spec, I can't remember the exact figures,
>but they were like +12.3V, +5.26V, +3.42v, etc.. (a hell of a lot better
>than the readings from this temporary psu.. not a single one is within
>spec.. all of them are 0.25 - 0.6v under par..)

??

5.26 is not within spec. 3.42 may not be either once the
ripple is accounted for. IOW, if you had a reading of 3.42
but 100mV of ripple, it's out of spec, rising 50mV to 3.47.

Where are you reading these temps? If not using the
multimeter, use it, not software.



>
>The testing method to check if the PSU was dead is simple.. plug in an old
>10GB HDD so there's some load and jumper PS-ON# to ground. I had it running
>that way before.. but after it died there was no response.. it still sounds
>like something charges though when I jumper PS-ON.. (like a very brief
>buzzing noise)..
>
>The psu came with my case about 3 years ago and I've never had a problem
>with it til now, so I guess I got my money's worth.. I'll be putting a
>brand name PSU in my PC next pay I think.. I was looking at the major
>players like Silverstone, Enermax, Antec, Fortron, etc.. I'm thinking about
>650W would be sufficient..

.... would be overkill unless you're running SLI'd video
cards. You are contrasting low-end, came-with-case PSU with
one that's more than you need. Just get a median wattage,
well made PSU. Your system will be fine with any decent
400W, maybe jumping to 450-500 just to cover some extra
current from overclocking. That it was running from a 350W
Wintech is in itself evidence it would run fine from a good
350W PSU (instead of the Wintech).



>
>This is my current setup, maybe you can recommend a particular model PSU:
>
>MSI K9N-SLI (nForce 500)
>Athlon X2 64 5200+
>2x 512MB Kingston HyperX PC6400 RAM
>Radeon X1600 Pro 256MB
>Western digital WD800JB
>Western digital WD2000JB
>Sony/Optiarc DVD-RW AD-5170A
>AOpen CDRW CRW5232
>

There are a lot of PSU that could run that, I suggest you
assemble a list of brands you would consider then buy
whichever goes on sale first... it's not like there's enough
difference to matter in reasonable quality PSU. The main
thing is to avoid the junk out there. A name-brand $50 PSU
is much closer to what you need than a repaired, generic
350W.


>You posted a lot of useful information, I'd be the last person to complain
>about too much detail.. even if I don't go into that much detail to fix the
>PSU, it's good to know incase I need to hack one up for a custom project..
>On that note.. why do UPS's not just supply 5, 12 and 3V rails instead of
>120V/240V ? It makes very little sense to convert 12 or 24V from backup
>batteries up to line voltage only to step it back down to 12V again anyway..

Batteries start above 13V and drop below 12V, they have to
be re-regulated anyway for good results. You could build a
custom UPS for one system, but it would be too great an
expense for the limited market segment which has a similar
enough system to use it.

You can buy PSU that run from 12V input, if that's what you
want, don't get an UPS... get a battery charger, battery and
12V input PSU. By making a traditional UPS as 120V output
it is the most compatible thing possible... and really, what
do they care if you have to pay them more for a larger UPS
to get the job done?

>
>I've seen people hack up PSU's into DC-DC supplies that run from a car
>battery, so the extra info helps..


It can be done, but to do it well it may be a lot more work
than it's worth since there are ready-made 12V input PSU,
now even purpose-built to be used on a car PC.