Sign in with
Sign up | Sign in
Your question

cpu killed power supply, why?

Last response: in Overclocking
Share
Anonymous
a b ) Power supply
a b à CPUs
a b V Motherboard
a b K Overclocking
October 4, 2004 12:37:52 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

Hi all

Hopefully this is the best news group to post this query. I have/had
amd athlon 1600+ cpu, chaintech motherboard and 512mb of ram running
windows xp. Friend of mine had a computer that died after a power
surge and wanted to check what bits were dead by component swapping
with my machine. it was a msi motherboard with a 900 duron cpu.

Hard drive was fine then he tried the duron900 cpu in my motherboard.
When turned on the power supply was sort of going half speed up and
down for about 3 seconds then smoked up. It killed my power supply and
as yet I don't know what else doesnt work as I have not got a new one
installed.

How can a spiked duron900 wipe out a power supply like that?

Just interested is all.

thanks for any replies

darrin
new zealand :) :) 
Anonymous
a b ) Power supply
a b à CPUs
a b V Motherboard
a b K Overclocking
October 4, 2004 3:35:40 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

"Gimmee" <thecrowatnz@hotmail.com> wrote in message
news:41610a12.5480858@news.nzwide.ihug.co.nz...
> Hi all
>
> Hopefully this is the best news group to post this query. I have/had
> amd athlon 1600+ cpu, chaintech motherboard and 512mb of ram running
> windows xp. Friend of mine had a computer that died after a power
> surge and wanted to check what bits were dead by component swapping
> with my machine. it was a msi motherboard with a 900 duron cpu.
>
> Hard drive was fine then he tried the duron900 cpu in my motherboard.
> When turned on the power supply was sort of going half speed up and
> down for about 3 seconds then smoked up. It killed my power supply and
> as yet I don't know what else doesnt work as I have not got a new one
> installed.
>
> How can a spiked duron900 wipe out a power supply like that?
>
> Just interested is all.
>
> thanks for any replies
>
> darrin
> new zealand :) :) 

It cant..
Chances are the motherboard or power supply had problems..

Just because it doesnt turn on doesnt mean it's not the power supply.. If a
spike went through the system then putting bad parts in good working kit
can have problems.. So a duff processor might kill the board..
But a duff processor wont kill the board in a way that it will kill the
processor..
Anonymous
a b ) Power supply
a b à CPUs
a b V Motherboard
a b K Overclocking
October 4, 2004 11:35:11 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

To kill a power supply, the supply had to be defective when
purchased. Review even Intel specifications for PSUs. It
lists wir size to short out all power supply outputs - and
still the power supply must not fail. But today, too many
computer experts assume they don't need no stink'in training.
Therefore they don't even know basic electrical concepts nor
do they even read the Intel requirements. This has created a
lucrative swamp for Asian power supply manufacturers to dump
inferior supplies at even higher profits.

For example, what do the written specs with that power
supply say. Did it specifically list these and so many other
necessary criteria? And was this provided in writing?
Specification compliance: ATX 2.03 & ATX12V v1.1
Acoustics noise 25.8dBA typical at 70w, 30cm
Short circuit protection on all outputs
Over voltage protection
Over power protection
100% hi-pot test
100% burn in, high temperature cycled on/off
PFC harmonics compliance: EN61000-3-2 + A1 + A2
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs

I doubt it which is why the power supply failed. Any computer
assembler with minimal electrical knowledge appreciates how
important things in this list are to computer integrity.
Apparently your supply did not have at least some of these
functions and therefore probably did not even provide written
numerical specs.

Now we get to an even more dangerous fact. A failing power
supply must never damage any computer hardware. If your
supply was missing essential functions, then another function
not installed may have been the circuit that keeps power
supply from damaging or overstressing any other hardware.

We build and repair computers only to learn. We don't do
this to save money. We do it to be smarter. Take this
problem and learn from it. Learn the many functions that must
be in all power supplies - even 30 years ago.

Fourth point: swapping to fix things is foolish especially
when simpler, faster, and more accurate methods are
available. Two discussions demonstrate how to perform this
all faster and more accurately with the ubiquitous 3.5 digit
multimeter:
"Computer doesnt start at all" in alt.comp.hardware on 10
Jan 2004 at http://tinyurl.com/2t69q or
"I think my power supply is dead" in alt.comp.hardware on
5 Feb 2004 at http://www.tinyurl.com/2musa

Again, gain from your problems. Learning from mistakes is
essential. Learn why any computer assembler with basic
knowledge knows and uses even a 3.5 digit meter to even find
problems before they happen.

Gimmee wrote:
> Hi all
>
> Hopefully this is the best news group to post this query. I have/had
> amd athlon 1600+ cpu, chaintech motherboard and 512mb of ram running
> windows xp. Friend of mine had a computer that died after a power
> surge and wanted to check what bits were dead by component swapping
> with my machine. it was a msi motherboard with a 900 duron cpu.
>
> Hard drive was fine then he tried the duron900 cpu in my motherboard.
> When turned on the power supply was sort of going half speed up and
> down for about 3 seconds then smoked up. It killed my power supply and
> as yet I don't know what else doesnt work as I have not got a new one
> installed.
>
> How can a spiked duron900 wipe out a power supply like that?
>
> Just interested is all.
Anonymous
a b ) Power supply
a b à CPUs
a b V Motherboard
a b K Overclocking
October 5, 2004 10:51:27 PM

Archived from groups: alt.comp.hardware.overclocking.amd (More info?)

w_tom wrote:
> To kill a power supply, the supply had to be defective when
> purchased. Review even Intel specifications for PSUs. It
> lists wir size to short out all power supply outputs - and
> still the power supply must not fail. But today, too many
> computer experts assume they don't need no stink'in training.
> Therefore they don't even know basic electrical concepts

So true, so true ... I think it's much worse in the overclocking scene
though, where people are doing voltmods without even knowing what a resistor
is.

> nor
> do they even read the Intel requirements. This has created a
> lucrative swamp for Asian power supply manufacturers to dump
> inferior supplies at even higher profits.
>
> For example, what do the written specs with that power
> supply say. Did it specifically list these and so many other
> necessary criteria? And was this provided in writing?
[...]
> I doubt it which is why the power supply failed.

The first part of this statement is true, but the second part is vacuously
true. Taking a look at a couple of what I would classify as damn high-end
PSUs:
[1] PCP&C Turbocool 510 series. You can grab the specs at:
http://www.pcpowercooling.com/products/power_supplies/m...
ols/hp_510atx.htm
[2] Antec True550. Specs are at
http://www.antec.com/specs/true550_spe.html
[3] Forton 550W (FSP550-60PLN-R). 24-pin ATX, but close enough :)  Specs from
http://www.fsusa.com/Product/FSP550-60PLN.pdf

There are possibly pages with more info than these, so feel free to correct
me :) 

To do the cheap end, I used the box for my Hyena ( = Foxconn = Allied =
millions of other names) "500W" PSU, and the bit of paper that came with it.

Note: I may have got some of the first thee PSUs mixed up below, but the
Hyena was done later so the values in the Hyena's column should all be
correct.

Stepping through the requirements:

> Specification compliance: ATX 2.03 & ATX12V v1.1
PCPC: [Y], Antec: [N], Fortron: [N/A (EPS12V)], Hyena: [Y]
I'm presuming Antec just forgot to list it, or presumed it was obvious.

> Acoustics noise 25.8dBA typical at 70w, 30cm
[N], [N], [N], [N]

> Short circuit protection on all outputs
[Y], [N], [Y], [Y]
PCPC: implied through overcurrent protection presumably

> Over voltage protection
[Y], [N], [Y], [Y]

> Over power protection

[Y], [N], [N], [N]
PCPC Listed as over-current protection

> 100% hi-pot test
[N], [N], [Y], [Y]

> 100% burn in, high temperature cycled on/off
[N], [N], [Y], [Y]

> PFC harmonics compliance: EN61000-3-2 + A1 + A2
[N,N,N], [N,N,N], [N,N,N], [N, N, N]

> EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
[Y, N, Y], [Y, Y, Y], [Y, Y, Y], [Y, possibly, Y]

"Possibly" because I have no idea what the CISPR22 symbol looks like and
there's a great big line of them on the side of the box, so possibly it's in
there.

Incidentally, it's quite fun hunting for CE symbols on obscure things. So
far I've found it on my ski goggles, a model london bus, and a can of
flyspray :) 

> Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
[UL, ULC, TUV], [UL, TUV, CB, VDE, D, N, S, FI], [UL, TUV, N], [D, N, S, FI,
CB, TUV, UL]

> Hold up time, full load: 16ms. typical
[Y], [N], [Y], [N]

> Efficiency; 100-120VAC and full range: >65%
[Probably], [Y], [N], [Y]
PCPC only said "70%" without stating conditions or whether it was a minimum.

> Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
[Y], [N], [N], [N]

> Dielectric withstand, input to output: 1800VAC, 1sec.
[N], [N], [N], [N]

> Ripple/noise: 1%
[Y], [Y], [Y], [Y]

> MTBF, full load @ 25°C amb.: >100k hrs
[Y], [N], [Y], [N]


So three of what I would regard as very close to top-of-the-line power
supplies failed a multitude of the tests. The generic Hyena didn't do to
shabbily, but this is possibly due to having a bit more space to print all
these things. Not to say I think the Hyena is anywhere near any of those in
terms of quality. I'd swap it in the blink of an eye for any of those if I
could.

I would be very surprised if any PSU (besides the Seasonic 400W supply that
you copy'n'pasted this from, complete with grammatical mistakes :)  )
actually list all of the above (and possibly even complied with, given the
range of safety and EMI standard that are listed). This doesn't mean it's a
bad power supply, as demonstrated above. If a power supply *did* have all
the features listed above (ignoring the non-safety related ones like EMI,
acoustic, ripple, etc) then I would agree that it's pretty much only
possible to blow the PSU if the PSU is defective (which can't be ruled out
for any manufacturer).

[...]

--
Michael Brown
www.emboss.co.nz : OOS/RSI software and more :) 
Add michael@ to emboss.co.nz - My inbox is always open
!