«

Yeah yeah, S.M.A.R.T. data is only any good for pass/fail. Any programme
that reads S.M.A.R.T. data and gives you anything *but* isn't worth the time
of day....

The flatmate's PC was giving her (me) problems. HD Sentinel kept reporting
more bad sectors every few days with the odd 'hang at checking IDE devices'
problem. HHD was cool, cable checked, replaced anyweay, molex swapped and
re-seated, dust removed around the IDE controller and sprayed with CO
cleaner, tried running the HDD off a PCI IDE/SATA card... Decided to RMA the
HDD.

The HDD I cloned onto and installed started reporting the same problems! The
PC is stable as a rock, an XP2500 Barton overclocked to XP3200 specs from
new. As the PC is so stable (re-checked with Prime95 last week) and the PSU
was a credible 520 watt mid-price unit ($130 in '04) I'd ruled it out as
being the probable cause. The PC wouldn't draw 250 watts on a bad day.

Ultimately I had to find out what was causing HD Sentinel to keep raising
the flag, telling me that the HDD was deteriorating almost daily. I am up to
139 reallocated sectors on this drive in two weeks. Yesterday I opened the
PSU and saw that the two 4700uf 10V big filtering caps were blown and
leaking. Have swapped the PSU (until I can get new caps, it's otherwise well
built) and have stop/started the machine several times with HD Sentinel
reporting no new problems. If it hadn't been raising warnings I wouldn't
have known to keep looking for problems. (The S.M.A.R.T. checking is on in
BIOS, reports all fine.)

I assume that the problem is on the 5V rail of the PSU or I would think that
the CPU and the GPU (7600) would have shown problems. The PSU has a single
28A 12V rail. The CPU and the GPU both have direct 12V cables supplying
their VRMs. As the HDD usues both 5V and 12V it had to be the 5V right? MBM5
gave the all-clear to all voltages once in Windows, I'm thinking the 5V was
spiking or sagging at startup.

Just thought I'd mention it. I think you can't have too much info and like
to monitor everything I can with my hardware.

Cheers,
--
Shaun.

DISCLAIMER: If you find a posting or message from me
offensive, inappropriate, or disruptive, please ignore it.
If you don't know how to ignore a posting, complain to
me and I will be only too happy to demonstrate...

Somewhere on teh intarweb "Nighthawk" typed:
> On Sun, 15 Jun 2008 00:45:18 +1200, "~misfit~"
> <> wrote:
>
>> Yeah yeah, S.M.A.R.T. data is only any good for pass/fail. Any
>> programme that reads S.M.A.R.T. data and gives you anything *but*
>> isn't worth the time of day....
>>
>> The flatmate's PC was giving her (me) problems. HD Sentinel kept
>> reporting more bad sectors every few days with the odd 'hang at
>> checking IDE devices' problem. HHD was cool, cable checked, replaced
>> anyweay, molex swapped and re-seated, dust removed around the IDE
>> controller and sprayed with CO cleaner, tried running the HDD off a
>> PCI IDE/SATA card... Decided to RMA the HDD.
>>
>> The HDD I cloned onto and installed started reporting the same
>> problems! The PC is stable as a rock, an XP2500 Barton overclocked
>> to XP3200 specs from new. As the PC is so stable (re-checked with
>> Prime95 last week) and the PSU was a credible 520 watt mid-price
>> unit ($130 in '04) I'd ruled it out as being the probable cause. The
>> PC wouldn't draw 250 watts on a bad day.
>>
>> Ultimately I had to find out what was causing HD Sentinel to keep
>> raising the flag, telling me that the HDD was deteriorating almost
>> daily. I am up to 139 reallocated sectors on this drive in two
>> weeks. Yesterday I opened the PSU and saw that the two 4700uf 10V
>> big filtering caps were blown and leaking. Have swapped the PSU
>> (until I can get new caps, it's otherwise well built) and have
>> stop/started the machine several times with HD Sentinel reporting no
>> new problems. If it hadn't been raising warnings I wouldn't have
>> known to keep looking for problems. (The S.M.A.R.T. checking is on
>> in BIOS, reports all fine.)
>>
>> I assume that the problem is on the 5V rail of the PSU or I would
>> think that the CPU and the GPU (7600) would have shown problems. The
>> PSU has a single 28A 12V rail. The CPU and the GPU both have direct
>> 12V cables supplying their VRMs. As the HDD usues both 5V and 12V it
>> had to be the 5V right? MBM5 gave the all-clear to all voltages once
>> in Windows, I'm thinking the 5V was spiking or sagging at startup.
>>
>> Just thought I'd mention it. I think you can't have too much info
>> and like to monitor everything I can with my hardware.
>>
> Out of interest, what make was the PSU with the faulty caps?

It's a Task TK520TX.

http://www.task.com.tw/product-power.htm

When I bought it I couldn't find it on Task's site, the closest I came was
this page:

http://www.task.com.tw/ps-atx-dual.htm

Mine is metallic blue but rated at 520W continuous, 590W peak. (+5V 34A,
+3.3V 28A and +12V 28A)

Not a common make I know but, before I found AcBel I was looking for a
quality mid-priced PSU. Overclockers NZ sell them and I found this user
report on NZ PC World forums just before I bought it (from b1naryb0y):

"I have that PSU and it is one of the best I have ever owned. I bought this
to replace my old $230 thermaltake that could no longer handle the load.

The 520W rating on the task PSU is the true watt output, they are capable of
upto 590W at peak load.

The killer feature for me was the 26Amps on the +12v rail, which is very
high considering most generic 400W PSU only have 10Amps and my Thermaltake
had 18Amps.

Running off the PSU I have:

1x AthlonXp 3200+
1x Radeon 9800pro
3x HDDs
1x Floppy
1x DVD Writer
1x CD writer
1x TV Tuner
1x Network Card

This should give you some idea of the load that the Task PSU can take.

Lastly, the Task PSU is very quiet and keeps my system nice and cool

"

(from: http://forums.pcworld.co.nz/archive/...p/t-50372.html )

It weighs nearly twice what slightly cheaper similarly rated PSUs of the
time weighed and the components inside look very high quality. Two full
length extruded heatsinks, huge toroidial inductors. good cabling... Now
that I've opened it I cxan see even more and have come to the conclusion
that they were unwittingly caught up in the 'bad caps' saga that swept Asia
from about '99 to '05. (Some say it's still going on.)

It seems that Task power supplies aren't sold in the US or UK, it was hard
to find anything out about them. However Steven Fan at OEM computers in
Manukau (where I bought it in late '04) assured me that they were good.
Steven likes me (he's the boss at OEM, whenever I go in to buy something he
sits in front of his PC, looks worried, then tells me his lowest price, even
though I'd already decided to purchase at their 'retail') so I don't think
he'd lie. In fact he's steered me away from the odd component that I'd
decided to buy saying "I won't say it's no good, just that you can do better
for your money" even when it means I go elsewhere to buy instead.

Gimme a sec...

Here ya go: http://community.webshots.com/album/563794226dBArNS some pics of
the Task PSU now.

> The PSU in my own machine is an 350W Enermax with 28A on the single
> 12V rail. The Enermax I got for the other demon PC is a 400W with two
> 20A 12V rails and two PCIe 6-pin connectors.

Yup. It's only really in the last 3 years that dual 12V rails have become
the norm. I've seen triple 12V rails on some PSUs. Two big ones for CPU and
CPU respectively and a third, lesser amperage rail for drives and what-not.

> The original Tt 430W
> unit was only 16A (tested, 18A claimed) on the single 12V rail but 28A
> on 3.3V and 30A on 5V.

Yeah, the low-end Tt PSUs really are crap. I'd never buy one (but have two
here that I was given).

> Modern mobos need plenty on the 12V rail(s).

Sure do, since the Pentium 4/Athlon XP mobos started running the CPU from
the 12V four pin (originally, my mobo has an 8 pin CPU power connector)
connector near the CPU socket, then GPUs started requiring their own 12V
connectors too. Older design PSUs are more biased towards 3.3V and 5V as
that's where the mobos got the power they fed the xPUs. 12V was really only
needed for drives back then.

Cheers,
--
Shaun.

DISCLAIMER: If you find a posting or message from me
offensive, inappropriate, or disruptive, please ignore it.
If you don't know how to ignore a posting, complain to
me and I will be only too happy to demonstrate...

Somewhere on teh intarweb "Richard" typed:
> ~misfit~ wrote:
>> Yeah yeah, S.M.A.R.T. data is only any good for pass/fail. Any
>> programme that reads S.M.A.R.T. data and gives you anything *but*
>> isn't worth the time of day....
>>
>> The flatmate's PC was giving her (me) problems. HD Sentinel kept
>> reporting more bad sectors every few days with the odd 'hang at
>> checking IDE devices' problem. HHD was cool, cable checked, replaced
>> anyweay, molex swapped and re-seated, dust removed around the IDE
>> controller and sprayed with CO cleaner, tried running the HDD off a
>> PCI IDE/SATA card... Decided to RMA the HDD.
>>
>> The HDD I cloned onto and installed started reporting the same
>> problems! The PC is stable as a rock, an XP2500 Barton overclocked
>> to XP3200 specs from new. As the PC is so stable (re-checked with
>> Prime95 last week) and the PSU was a credible 520 watt mid-price
>> unit ($130 in '04) I'd ruled it out as being the probable cause. The
>> PC wouldn't draw 250 watts on a bad day. Ultimately I had to find out
>> what was causing HD Sentinel to keep
>> raising the flag, telling me that the HDD was deteriorating almost
>> daily. I am up to 139 reallocated sectors on this drive in two
>> weeks. Yesterday I opened the PSU and saw that the two 4700uf 10V
>> big filtering caps were blown and leaking. Have swapped the PSU
>> (until I can get new caps, it's otherwise well built) and have
>> stop/started the machine several times with HD Sentinel reporting no
>> new problems. If it hadn't been raising warnings I wouldn't have
>> known to keep looking for problems. (The S.M.A.R.T. checking is on
>> in BIOS, reports all fine.) I assume that the problem is on the 5V rail
>> of the PSU or I would
>> think that the CPU and the GPU (7600) would have shown problems. The
>> PSU has a single 28A 12V rail. The CPU and the GPU both have direct
>> 12V cables supplying their VRMs. As the HDD usues both 5V and 12V it
>> had to be the 5V right? MBM5 gave the all-clear to all voltages once
>> in Windows, I'm thinking the 5V was spiking or sagging at startup.
>>
>> Just thought I'd mention it. I think you can't have too much info
>> and like to monitor everything I can with my hardware.
>
> I have had the 12v drop to about 9-10 on a PSU and the CPU and VGA
> kept on working.. The only clue was that I was getting coasters from
> the burner all the time, it took ages to recognise discs and was in
> general an unreliable piece of crap. Was only when I looked in the
> bios I saw that it was low. Swapped PSU over and the burner came
> right.

Yeah, I suspected the PSU early on but the BIOS, MBM5 and even my $15
multimeter all said that the 12V was within cooee of being spot on. I
suspect that, if the 12V was the rail that had the problems, the CPU would
have been unstable as, when I overclocked it a couple years ago, I found the
minimum vcore I could that would run stably. As the vcore is derived from
the 12V rail, I would have expected random crashes if it was sagging.

As those electros are for 'smoothing' I'm thinking that perhaps there were
maybe spikes on the 12V and perhaps sags/spikes on the 5V. Without a PSU
tester or an oscilloscope and dummy load I guess I won't know for sure. All
I do know is that something was rotten in the state of Denmark, as evidenced
by the blown and crusty electros.

http://community.webshots.com/album/563794226dBArNS

Cheers,
--
Shaun.

DISCLAIMER: If you find a posting or message from me
offensive, inappropriate, or disruptive, please ignore it.
If you don't know how to ignore a posting, complain to
me and I will be only too happy to demonstrate...

"~misfit~" <> wrote in message
news:...
> Yeah yeah, S.M.A.R.T. data is only any good for pass/fail. Any programme
> that reads S.M.A.R.T. data and gives you anything *but* isn't worth the
> time of day....
>
> The flatmate's PC was giving her (me) problems. HD Sentinel kept reporting
> more bad sectors every few days with the odd 'hang at checking IDE
> devices' problem. HHD was cool, cable checked, replaced anyweay, molex
> swapped and re-seated, dust removed around the IDE controller and sprayed
> with CO cleaner, tried running the HDD off a PCI IDE/SATA card... Decided
> to RMA the HDD.
>
> The HDD I cloned onto and installed started reporting the same problems!
> The PC is stable as a rock, an XP2500 Barton overclocked to XP3200 specs
> from new. As the PC is so stable (re-checked with Prime95 last week) and
> the PSU was a credible 520 watt mid-price unit ($130 in '04) I'd ruled it
> out as being the probable cause. The PC wouldn't draw 250 watts on a bad
> day.
>
> Ultimately I had to find out what was causing HD Sentinel to keep raising
> the flag, telling me that the HDD was deteriorating almost daily. I am up
> to 139 reallocated sectors on this drive in two weeks. Yesterday I opened
> the PSU and saw that the two 4700uf 10V big filtering caps were blown and
> leaking. Have swapped the PSU (until I can get new caps, it's otherwise
> well built) and have stop/started the machine several times with HD
> Sentinel reporting no new problems. If it hadn't been raising warnings I
> wouldn't have known to keep looking for problems. (The S.M.A.R.T. checking
> is on in BIOS, reports all fine.)
>
> I assume that the problem is on the 5V rail of the PSU or I would think
> that the CPU and the GPU (7600) would have shown problems. The PSU has a
> single 28A 12V rail. The CPU and the GPU both have direct 12V cables
> supplying their VRMs. As the HDD usues both 5V and 12V it had to be the 5V
> right? MBM5 gave the all-clear to all voltages once in Windows, I'm
> thinking the 5V was spiking or sagging at startup.
>
> Just thought I'd mention it. I think you can't have too much info and like
> to monitor everything I can with my hardware.
>

Let me get this straight...You had a faulty power supply, but HDS led you to
believe that the hard drive was bad, so you replaced that instead. And so
you're back here hustling this product because...?!?

On 2008-06-14, ~misfit~ <> wrote:
> Yeah yeah, S.M.A.R.T. data is only any good for pass/fail. Any programme
> that reads S.M.A.R.T. data and gives you anything *but* isn't worth the time
> of day....
[snip]
>
> Just thought I'd mention it. I think you can't have too much info and like
> to monitor everything I can with my hardware.

I have always taken the policy, that my backups will rule. HD flaking, or
not, then it is a case my data being worth more than that question.

HD are so common, like cars and SUVs, that one does not pay any attention to
the idea that they may fail.

Still, it is the same with the electricity power supply of this country.

Backup people, for for you have either needed them or you will. Insurance
companies were built on this. For material stuff, ships and the like.

Your data, who cares a flying pig, apart from you? How important is *your*
data to *you*.

Make it so. To quote a certain Captain.

(Hint. Star Trek The next generation.