3 blown Power supply's in 6 months.

[RobW]

Hi all,
I have a P.C. that keeps blowing power supply's.
It is P4 2.4, Gigabyte GA-8IG100MK Main Board in a "whitebox" of
unknown make.
Low spec AGP video card
2 x Optical drives
1 x 80gig Seagate Barracuda
The last power supply it blew was a Unicase 400W Silent power supply.


I has been plugged into a middle of the road surge protector the whole
time (which seems to fine.)
This has all happened in the same house; however there hasn't been any
other power problems.

I'm thinking maybe it's the case because the problems happend when the
unit is being switched o every time.
Anyone else seen something like this?
TIA

Rob

 

[Jim Macklin]

Is the line power, from the electric company stable and
clean (no spikes, surges) ? Do you have any other
electrical problems? Is the computer on a circuit that is
isolated from heavy draw items such as motors (air
conditioners, washing machines) or do you see any
fluctuation in the lights?

Have you tried a "name brand" PSU from a company such as PC
Power & Cooling or Antec?


--
The people think the Constitution protects their rights;
But government sees it as an obstacle to be overcome.


| Hi all,
| I have a P.C. that keeps blowing power supply's.
| It is P4 2.4, Gigabyte GA-8IG100MK Main Board in a
"whitebox" of
| unknown make.
| Low spec AGP video card
| 2 x Optical drives
| 1 x 80gig Seagate Barracuda
| The last power supply it blew was a Unicase 400W Silent
power supply.
|
|
| I has been plugged into a middle of the road surge
protector the whole
| time (which seems to fine.)
| This has all happened in the same house; however there
hasn't been any
| other power problems.
|
| I'm thinking maybe it's the case because the problems
happend when the
| unit is being switched o every time.
| Anyone else seen something like this?
| TIA
|
| Rob
|

 

[RobW]

Is the line power, from the electric company stable and
clean (no spikes, surges) ? Do you have any other
electrical problems? Is the computer on a circuit that is
isolated from heavy draw items such as motors (air
conditioners, washing machines) or do you see any
fluctuation in the lights?

Have you tried a "name brand" PSU from a company such as PC
Power & Cooling or Antec?

Hi Jim,

Generally speaking the power from the electrical company is good.
No other electrical probs.
No heavy draw or fluctuation of lights either.
I haven't tried a name brand PSU as I have been too frightened it will
get fried like the others. Also, up to this point the supplier has
been replacing the PSU's under warranty.
They will no longer do this, lol.
I'm really wondering if it is a problem with the case.
Especially since the damage seems to be happening when switching on
the machine.

Cheers.
Rob

 

[Guest]

Try another powercable it might have a faulty fuse.

 

[Leythos]

Try another powercable it might have a faulty fuse.

Power cables don't have fuses, at least not the ones I've seen connected
to computers in the US.

If you are killing PSU's and you are SURE THE INCOMMING POWER IS
GOOD/Clean, then you've got a device inside the computer that's
overloading the PSU.

You need to determine what part of the PSU is dead - is it a fuse on one
of the DC lines or one the AC of the PSU or something else.

Your spec's don't seem to need more than 350W would provide, but you
could have a shorted/intermittent connection where the screws mounting
the motherboard connect to the chassis - many times the installer will
not use the proper screw size (screw heads) and will short a trace
without knowing it's been done.

Disconnect everything, reseat the computers motherboard on the chassis,
make sure that every mounting post is accounted for and that none are
shorting out traces on the board, same for the screws.

I've also seen power leads (for drives) without a cap over them short
out on the metal parts of a case and blow a PSU.

 

[Guest]

im from UK and every power cable has to have a fuse in UK so suppose it
depends where your from. if you use fuses it could well be a problem.

 

[Leythos]

im from UK and every power cable has to have a fuse in UK so suppose it
depends where your from. if you use fuses it could well be a problem.

I agree then, not knowing where the OP was from, it's a good assumption
to guess at the fused power cord if his country requires such.

 

[Yves Leclerc]

The "common" middle of the road surge protector can only handle ONE/TWO
surges. After this, the surge protection does not exist!

Your power in your room/office is definiately BAD! You are going to have
problems like this forever. The only true ways to fix this are:

1) Call an electrician and have them put in an "isolated" power circuit for
your PC. (Expensive!!!)

2) Buy an line conditioning power regulator, or a UPS. These type of units
will "filter" the power so that the out-bound power is with-in the "safe"
range of the power supplies.

And, always use a 'name-brand' power supply. Antec and Enermax always use
"good" components in their PSUs and the PSUs are mostly correctly rated for
their WATTAGE. Generic PSUs are not worth the monet when the power is
"dirty".

 

[w_tom]

Fuses don't protect electronics. After something electronic
has failed, a fuse may blow. This so the hardware failure
does not result in harm to humans. Do you think a fuse blows
to protect electronics? Electricity. First it flows through
everything in a circuit. Only then does something blow after
everything has been exposed to the same current flow. A 13
amp fuse is going to blow because more than 3 amps flowed
through a component rated only for 2 amps? Only when the 2
amp component shorts and starts drawing more like 13 amps.

 

[w_tom]

A properly designed power supply cannot be damaged by the
load. A properly designed power supply can have all output
wires shorted together. Still the power supply must work just
fine after those shorted wires are disconnected.

To dump power supplies into a market of computer assemblers,
some manufacturers 'forget' to include standard functions.
This means a less expensive power supply results in greater
profits. Yet many computer assemblers will just swap out
power supplies rather than first learn what has failed; or
why.

The best evidence is a dead body. What inside those power
supplies failed. Without that information, no one can tell you
why power supplies are failing - other than note discounted
supplies are routinely missing essential functions. Functions
that even mean power supply failure could damage motherboard
and disk drives.

A properly manufactured power supply will not damage other
computer components. Power supply that is missing essential
functions can damage motherboard and RAM. Just another reason
why it is foolish to buy power supplies only on price and
watts. First thing that power supply must provide is a long
list of written and numeric specs. If not, then you are only
putting other computer components at risk.

Will other electrical appliances damage a power supply?
No. Destructive spikes from washing machines or vacuum
cleaner is another myth - if the power supply contains
minimally essential functions. Will voltage dimming (a sag or
brownout) cause power supply failure? Again no if the power
supply is properly constructed.

If the supply does not retail list for at least $60, then
you know it is missing essential function. Such power
supplies must also forget to provide any numerical
specifications.

However do you know those power supplies are damaged. For
example, a power supply under too much load will shut off to
protect itself. Then others will declare the power supply as
defective rather than first learn why the power supply shuts
down. A computer assembler should verify power supply output
voltages with a multimeter when system is first constructed.

Just a few places to start fixing the problem. Again, best
evidence is inside the failed power supplies. Information
that could result in an immediate and concise answer.
Currently we can only post a long list of possible reasons for
failure.

 

[RobW]

I pulled the case and found a fuse that now resembles a used
firecracker which I guess came out of the PSU. (the PSU has a bottom
intake fan with large gaps).
I am going to take Leythos's suggestion and investigate which fuse on
the PSU is blown. I may have to get someone who knows what they are
looking at to examine it, .
I am in Australia and the mains is 240VAC.
The power leads have no fuses.
Thanks again and I will post the outcome for anyone who is interested.
Rob

 

[Mike Hall \(MS-MVP\)]

Tom

Take a look at this..

http://www.littelfuse.com/cgi-bin/r.cgi/application/Computer/3.html?LFSESSION=oNm8TFzMb5

Fuses are not just found in domestic, industrial and automotive power supply
lines..

--
Mike Hall
MVP - Windows Shell/User

 

[Jim Macklin]

Be careful poking around with/in power supplies, there are
rather large capacitors inside that can hold a lethal charge
for some time (weeks perhaps) and they need to be properly
discharged before you try to work inside. They will also
fry any meter you connect checking for resistance.

You can check for a shorted screw by using a DVM set to
resistance (ohms) and checking for continuity (0 ohms) where
the parts should be grounded together, or infinite
resistance between electric circuits and the case.


--
The people think the Constitution protects their rights;
But government sees it as an obstacle to be overcome.


|I pulled the case and found a fuse that now resembles a
used
| firecracker which I guess came out of the PSU. (the PSU
has a bottom
| intake fan with large gaps).
| I am going to take Leythos's suggestion and investigate
which fuse on
| the PSU is blown. I may have to get someone who knows what
they are
| looking at to examine it, .
| I am in Australia and the mains is 240VAC.
| The power leads have no fuses.
| Thanks again and I will post the outcome for anyone who is
interested.
| Rob
|
|
| On Mon, 09 May 2005 13:31:15 GMT, Leythos
|
| >In article
| >[email protected] says...
| >> im from UK and every power cable has to have a fuse in
UK so suppose it
| >> depends where your from. if you use fuses it could well
be a problem.
| >
| >I agree then, not knowing where the OP was from, it's a
good assumption
| >to guess at the fused power cord if his country requires
such.
| >
| >
| >>
| >> "Leythos" wrote:
| >>
| >> > In article
| >> > (e-mail address removed) says...
| >> > > Try another powercable it might have a faulty fuse.
| >> >
| >> > Power cables don't have fuses, at least not the ones
I've seen connected
| >> > to computers in the US.
| >> >
| >> > If you are killing PSU's and you are SURE THE
INCOMMING POWER IS
| >> > GOOD/Clean, then you've got a device inside the
computer that's
| >> > overloading the PSU.
| >> >
| >> > You need to determine what part of the PSU is dead -
is it a fuse on one
| >> > of the DC lines or one the AC of the PSU or something
else.
| >> >
| >> > Your spec's don't seem to need more than 350W would
provide, but you
| >> > could have a shorted/intermittent connection where
the screws mounting
| >> > the motherboard connect to the chassis - many times
the installer will
| >> > not use the proper screw size (screw heads) and will
short a trace
| >> > without knowing it's been done.
| >> >
| >> > Disconnect everything, reseat the computers
motherboard on the chassis,
| >> > make sure that every mounting post is accounted for
and that none are
| >> > shorting out traces on the board, same for the
screws.
| >> >
| >> > I've also seen power leads (for drives) without a cap
over them short
| >> > out on the metal parts of a case and blow a PSU.
| >> >
| >> > --
| >> > --
| >> > (e-mail address removed)
| >> > remove 999 in order to email me
| >> >
| >>
| >
| >--
|

 

[Leythos]

Be careful poking around with/in power supplies, there are
rather large capacitors inside that can hold a lethal charge
for some time (weeks perhaps) and they need to be properly
discharged before you try to work inside. They will also
fry any meter you connect checking for resistance.

Most computer PSU's use CAPS on the DC side and they don't have more
than 20+ VDC on most of them, the caps will also discharge in a short
amount of time due to the way a PSU is designed.

Now, if we were talking the HV circuit of a monitor/TV, I would agree
with you completely.

As with anything electrical - always disconnect power before working on
it and make sure there are no unsafe powers still available.

You can check for a shorted screw by using a DVM set to
resistance (ohms) and checking for continuity (0 ohms) where
the parts should be grounded together, or infinite
resistance between electric circuits and the case.

The screw, also connected to the chassis, should be a short to the power
supply ground, at least the screw eyes in many boards are also grounding
pads and will be the same as DC ground for the PSU. In many cases, due
to capacitors and other devices you can't get an infinite ohms reading
testing like you describe.

The best method is to take the motherboard back out of the case and see
if it's been improperly installed.

Since he mentioned a blown fuse inside the PSU, it would stand to reason
(since he said it's not on the AC Side) that he has an intermittent
short somewhere - as in a case/screw or wire missing insulation.

 

[w_tom]

Why did you provide a URL for circuit protection devices?
Those are not fuses? Did you assume that because it is a
Littelfuse product, then it must be a fuse? Or did you first
review the datasheets before posting? Cited are applications
for circuit protection devices .... which are not fuses.

Circuit protector devices and fuses are not the same
device. You have claims a device called TMOV is a fuse. That
is wrong. Go back and learn about TMOVs ... "available in 14
and 20 mm disc size with and without a monitor lead (to alert
you that the thermal element has opened). The 14 mm parts are
rated to 6kA and the 20 mm parts are rated to 10kA." Do you
claim these "6,000 and 10,000 amp fuses" would stop electronic
damage?

Again, fuses blow after the electronic damage has happened
so that the human is not put at risk. This being basic
electrical knowledge that even a computer assembler should
know.

 

[w_tom]

And again, fuse designed to blow so that "damage to the
lines/devices" created by component failure does not threaten
human life. Any device whose destruction can harm humans will
be disconnected by fuses and other equivalent 'one shot'
devices. The fuse is not installed to protect hardware. In
the process of protecting humans, a fuse may protect some
hardware - ie a power cord. But fuses are installed to
protect humans. Any protected hardware is incidental.

When a fuse blows in a power supply, a failed component
often causes that open fuse. Even shorting all power supply
outputs would not blow a fuse. The concept is called
fold-back current limiting. The fuse is for human
protection. It blows typically because a component inside the
supply has failed. Furthermore many power supplies don't use
fuses. Such human protection is found elsewhere in the power
supply design making the fuse unnecessary.

 

[Leythos]

Again, fuses blow after the electronic damage has happened
so that the human is not put at risk. This being basic
electrical knowledge that even a computer assembler should
know.

Sorry, the fuse it sized to prevent damage to the lines/devices, not to
blow after the device has been damaged (although that may also happen).

The fuse is sized to protect some part of the device, be it power lines,
transformers, etc....

 

[Guest]

One more option to check. Make sure the PSU is set for the correct incoming
voltage. There is a red switch to set for 110/120 or 220/240.

 

[Jim Macklin]

Fuses are rated by amps and time and are intended to open
the circuit before the wires emits smoke. In other words,
the fuse must blow before a fire starts. There are fast
fuses that protect some electronic circuits in devices such
as the resistance (ohms) circuit in a voltmeter.

There are also slow-blow fuses for use with motors.

There are ground fault detectors and there are Hall Effect
sensors which can detect problems in circuits.

One thing that the OP needs to tell, under what condition
and how fast does the PSU blow, does it work at all and
sometimes fail, or does it always fail immediately?


--
The people think the Constitution protects their rights;
But government sees it as an obstacle to be overcome.


| In article <[email protected]>,
(e-mail address removed) says...
| > Again, fuses blow after the electronic damage has
happened
| > so that the human is not put at risk. This being basic
| > electrical knowledge that even a computer assembler
should
| > know.
|
| Sorry, the fuse it sized to prevent damage to the
lines/devices, not to
| blow after the device has been damaged (although that may
also happen).
|
| The fuse is sized to protect some part of the device, be
it power lines,
| transformers, etc....
|
|
| --
| --
| (e-mail address removed)
| remove 999 in order to email me

 

[Leythos]

When a fuse blows in a power supply, a failed component
often causes that open fuse. Even shorting all power supply
outputs would not blow a fuse.

That depends on the power supply design, but in most switched power
supply designs you are correct, they can sustain a shorted output for a
sustained period.