How can I check how many watts a PS is outputting?

[zer0nix]

...also, what are the signs of PS failure, aside from fluctuating voltage (which my ps does not have)?


this topic could alternatively be titled 'Help! i think my PS is dying, but i'm not so sure...'

-=WARNING=- long, semi-related paragraphs ahead...


i think my PS is outputting an insufficient amount of power. ever since i installed a sunbeam rheostat, other installed components have been failing left and right; i lost THREE hdd this year, all attached (one at a time) on the same power line as that rheostat.

5 minutes ago, i thought my dvd burner (the only component now connected on that power cable) also died; it wasn't responding when i pressed on the eject button, it wasn't lighting up and explorer wasn't detecting it. i pulled open the case, rechecked the connections and when i tried to reboot, the burner lit up and ejected as per my commands -but now suddenly the pc was acting as if all the hdd were missing. i immediately shut off the system and detached the burner and viola, now i'm booted in and typing on the pc. all this leads me to believe there is *somehow, suddenly* insufficient power for my burner and all my hdds -and this accursed rheostat.

previously, this pc had been running with 3 hdd and 1 burner for nearly 3 years straight.


--the reason i suspect my psu is failing is because i've actually encountered all this before:

shortly before my old psu (from an even older, 1995 era pc) erupted in sweet white smoke, the same afflictions haunting my current setup haunted my old rig: hdd were dying left and right, the pc acted slow and unstable despite being 48h prime95 stable and virus free, and one of the last components to fail was my old, trusty cd burner; one day, after inserting a cd i had burned, suddenly it didn't seem to be reading the disc and it wasn't ejecting as per my commands.

it wasn't long before the pc died. some of the hardier components continue to function beautifully in a new pc (ram, cdrom, floppy drive). others, not so much (ALL my hdds, *********!! ********* TO HELL!!).

before that psu died, i also added a new component to that system: some new ram and a storage hdd. calculations showed my psu OUGHT to have been able to power this rig, but it was shortly after this that problems began to appear, and **** started to die.

-of course, it was summer and this article:
http://www.motherboards.org/articles/guides/1487_8.html
has been enlightning in suggesting that perhaps the elevated temperature was causing my psu to provide far less power than it was rated for.


my current (aging) setup:

mobo: nf7s v2 w microcool northpole nb1 (40mm fan)
cpu: xpm2500 @ 200 MHz x12, 1.725v
cooler: 90mm panaflo
ram: 2x 512mb pc3000 kingston hyper x @ pc3200, 2.6v
gpu: 64mb radeon 9500 @ stock
hdds: 1x200gb seagate barracuda + 1x80gb western digital (on seperate power cables)
optical: 1x lite on ld3540a dvd burner
network card: 1x readynet 10/100 ethernet adapter
floppy drive: 1x floppy drive
case fans: 2x 80mm panaflo
rheostat: sunbeam rheostat powering the casefans and northpole (draws power from same cable as dvd burner; is this bad?)
psu: 300 watt antec sl300s


multimeter readings:

12v: 11.93-11.92v
5v: 5.10-5.11v
3.3v: 3.19-3.18v


so, what are your thoughts ocforums?

EDIT: oh wow, there's a swear filter now. neato.

 

[Oklahoma Wolf]

Your SL300 has bad capacitors - I'd put money on it. Known problem on those older Antecs. Replacing with a decent new 300W-350W should solve all issues.

 

[zer0nix]

aside from the obvious failure issues, how do you know it's bad caps when the voltage is *mostly* stable and in a 'safe' range?


EDIT: the reason i ask is, antec has a policy where they might actually send this defective psu right back if they deem it to have no problems -and i will be out the shipping costs.

i want to find something i can point at to confirm there's something wrong.

 

[JamesXP]

Capacitors control ripple (smoothen and even it out) you cannot measure this with a DMM, only with an ossicillioscope... (spelling anyone?)

 

[Oklahoma Wolf]

I know because bad caps causes ripple and noise to jump to insane levels, which cooks hardware. The voltages aren't stable or safe - you just can't tell with a normal DMM. You need an oscilloscope to see it. See my review here (bottom of the page) to see what I'm talking about: http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story4&reid=71

Look through the PSU grilles at the capacitors. You should see one or more bulging or leaking.

 

[zer0nix]

oh wow, thanks for that! damn, another $30 wasted getting that multimeter oh well.

i am rmaing this thing RIGHT NOW.

 

[ratbuddy]

aside from the obvious failure issues, how do you know it's bad caps when the voltage is *mostly* stable and in a 'safe' range?


EDIT: the reason i ask is, antec has a policy where they might actually send this defective psu right back if they deem it to have no problems -and i will be out the shipping costs.

i want to find something i can point at to confirm there's something wrong.

I have an Antec Smartpower 2.0 450 watter. I read about bad caps in these things after buying it, and called Antec to ask if I needed an RMA. They had me check the serial number on the side of the PSU (had to remove it from the case to check this) and they would be able to tell from the manufacture date whether it had the bad ones or not. If it did, they would swap it out no questions asked. Why not give them a call? I got right through to someone who spoke english well and was familiar with what I was asking..

Can't hurt to try, right?

 

[Oklahoma Wolf]

They had me check the serial number on the side of the PSU (had to remove it from the case to check this) and they would be able to tell from the manufacture date whether it had the bad ones or not.

It's important to note that even if Antec doesn't recognize it based on the date of manufacture, that does not necessarily mean the caps are fine and dandy. This was an epidemic reaching from the old pp303x and pp412x days to the Truepower II days, and all CWT built Antecs from those days are at risk.

I've seen them fail in as early as 10 months in the case of the Smartpower 2's, and as late as 3 years on those that have been well cooled.

 

[ratbuddy]

It's important to note that even if Antec doesn't recognize it based on the date of manufacture, that does not necessarily mean the caps are fine and dandy. This was an epidemic reaching from the old pp303x and pp412x days to the Truepower II days, and all CWT built Antecs from those days are at risk.

I've seen them fail in as early as 10 months in the case of the Smartpower 2's, and as late as 3 years on those that have been well cooled.

They actually weren't able to tell me if mine was OK. It was made on the borderline month when they definitely stopped using the Fuhjyyu caps. He said if it had been 1 month later he could give it a clean bill of health. I wanted a new PSU for my GX2 anyway so I just put the thing back in the box to save as a spare.

 

[RollingThunder]

oh wow, thanks for that! damn, another $30 wasted getting that multimeter oh well.

i am rmaing this thing RIGHT NOW.

Zero,

Buying a multi-meter is never a waste for computer enthusiasts. It should be part of your day-to-day toolbox. I use mine all the time.

 

[Oklahoma Wolf]

Yeah - calling a DMM useless because it won't do a scope's job is a little like calling a car useless because it won't do a pickup's job

 

[ratbuddy]

Yeah - calling a DMM useless because it won't do a scope's job is a little like calling a car useless because it won't do a pickup's job

*cough*

 

[Oklahoma Wolf]

A DMM with scope functions costs a fortune in relation to the ordinary DMM. Had the Brat been a Rolls-Royce vehicle, the analogy might apply better