Bad Motherboard?? ...short?

  • Thread starter Thread starter Jeff
  • Start date Start date
J

Jeff

....have an older intel socket 478 motherboard in a 4 year old machine that I
was working on today.

The thing ran fine yesterday. Because this was an older spare machine, I
wasn't as careful as I normally am. I drilled a hole (with a cordless drill)
in the case to install another case fan. When I attempted to reboot, things
weren't working correctly.

As soon as the power supply is plugged in, the machine starts up even with
the case power switch completely disconnected. It won't boot and won't even
post. I get nothing on the video at all. If I pull the ram out, I get
warning beeps that seem to mean that there is no ram. ...but that's it. I've
pulled out and unplugged everything, but nothing helps. I've pulled out the
bios battery to reset. Again, the case power switch doesn't work and instead
the thing just comes on when the machine's power supply is switched on. I
didn't have a spare power supply to swap in, but I can't see how the psu
would matter. When unplugged the psu from the MB, the psu [properly] does
not come on.

As near as I can tell, running the drill into the front of the case must
have caused something to short in the board. okay, I know this was stupid,
and I would not have done this on a new machine.

Is there any other explanation other than a bad MB?
 
"""running the drill into the front of the case .."""

It looks like must be this drilling...perhaps you damaged the front panel
Power Switch /
Reset button / Led / USB wires n sockets / Audio .....
whatever is on the front... try disconnecting the front panel wires
....unplug all of them ..
The only one you have to have is the power On/Off ..operate that (now)
disconnected
line by shorting the contacts as needed on the M.Board.
I suppose you MIGHT have damaged a Hard Drive / Floppy / CD so as to stop
the
start up...
My bet goes on the Power or reset switch damaged open / or shorted.
OR some scrap o metel got into the M.Board...turn it upside down & kick it
???
(a little bit).
Mouse
@@@
 
Trimble Bracegirdle said:
"""running the drill into the front of the case .."""

It looks like must be this drilling...perhaps you damaged the front panel
Power Switch /
Reset button / Led / USB wires n sockets / Audio .....
whatever is on the front... try disconnecting the front panel wires
...unplug all of them ..
The only one you have to have is the power On/Off ..operate that (now)
disconnected
line by shorting the contacts as needed on the M.Board.
I suppose you MIGHT have damaged a Hard Drive / Floppy / CD so as to stop
the
start up...
My bet goes on the Power or reset switch damaged open / or shorted.
OR some scrap o metel got into the M.Board...turn it upside down & kick it
Click to expand...

Thanks. I should have specified that I ran the drill into a thin bare strip
of metal that was no where near the actual power switch. ...and I've pretty
much tried everything that you mentioned - although there is no sense in
having to short the contacts, because as I mentioned, the power turns on
immediately even with the front panel wires all completely unplugged. I
actually removed the entire MB and blew the case and MB out with air.
....still the same. I can get a spare power supply to the thing tomorrow, but
I don't really see that this will do any good as the power supply does shut
off when it is disconnected from the MB. I also disconnect all floppy/CD/
and hard drives. The thing should post without the drives even if it goes no
further.
 
Jeff said:
...have an older intel socket 478 motherboard in a 4 year old machine that I
was working on today.

The thing ran fine yesterday. Because this was an older spare machine, I
wasn't as careful as I normally am. I drilled a hole (with a cordless drill)
in the case to install another case fan. When I attempted to reboot, things
weren't working correctly.

As soon as the power supply is plugged in, the machine starts up even with
the case power switch completely disconnected. It won't boot and won't even
post. I get nothing on the video at all. If I pull the ram out, I get
warning beeps that seem to mean that there is no ram. ...but that's it. I've
pulled out and unplugged everything, but nothing helps. I've pulled out the
bios battery to reset. Again, the case power switch doesn't work and instead
the thing just comes on when the machine's power supply is switched on. I
didn't have a spare power supply to swap in, but I can't see how the psu
would matter. When unplugged the psu from the MB, the psu [properly] does
not come on.

As near as I can tell, running the drill into the front of the case must
have caused something to short in the board. okay, I know this was stupid,
and I would not have done this on a new machine.

Is there any other explanation other than a bad MB?
Click to expand...

When you drill a metal case, you remove all hardware, and keep a vacuum
cleaner handy. That way, you can collect all the bits of conducting metal
that the drilling operation will create. If those fall onto the motherboard,
there is no way to predict what will happen.

Disassemble the thing, clean up inside, and reassemble it. Maybe you'll
get lucky, or something.

Paul
 
Paul said:
Disassemble the thing, clean up inside, and reassemble it. Maybe you'll
get lucky, or something.

Paul
Click to expand...


Thanks. ...but I already tried this. I think that I have learned my lesson.

....good I didn't do something like this on a new, expensive machine.

I'll try a new power supply tomorrow, but I don't expect much. At least a
used board from ebay is only about $40.
 
Thanks. ...but I already tried this. I think that I have learned my lesson.
...

I'll try a newpower supplytomorrow, but I don't expect much. At least a
used board from ebay is only about $40.
Click to expand...

Why would you now replace a power supply? For example, all ATX
power supplies must have all outputs shorted together and never be
damaged. If a metal chip caused a short, then protection circuits
standard inside power supplies must protect everything without damage.

What were the beep codes? That idenifies some items as working. And
if you got beep codes, the power supply is working - at least as best
you can tell until those voltages are measured with a 3.5 digit
multimter.

A spare power supply is wasted time and money. If swapping things
to fix it, then you may have exponentially complicated the problem.
Start by establishing what is working. The 3.5 digit meter and this
procedure is a good start - without anythng removed or disconnected
from the machine:
"When your computer dies without warning....." starting 6 Feb 2007 in
the newsgroup alt.windows-xp at:
http://tinyurl.com/yvf9vh

Is BIOS battery defective? Again, don't even remove the battery. In
but seconds, meter could have measured battery and determined if
battery was connected. Removing and reseating BIOS battery could have
adversely changed some CMOS settings - just another example of
exponentially complicating a problem.

Testing a power supply disconnected from a load and without numbers
from that meter says little. Even some defective supplies will come
on during that test. You are doing tests that only say maybe. You
need diagnostic procedures that say definitely good or definitely
bad. Shotgunning can even make things worse.

Did a metal fragement short something? Well once that short is
removed, then board will work again. ICs work just fine after a short
circuit is removed - an IC design standard dating back to late 1960
designs. If a metal fragment caused the problem and board still does
not work; then metal fragment is still on the board. You have
apparently assumed a metal fragment causes damage in direct
contradiction to how digital ICs work.

But then your every test still does not even say the power supply is
OK. Your each test has only said maybe this might be working OK. You
need tests that say "yes it is OK without a doubt so I will move on to
other suspects". Get the meter. Stop guessing.
 
A spare power supply is wasted time and money. If swapping things
to fix it, then you may have exponentially complicated the problem.
Start by establishing what is working. The 3.5 digit meter and this
procedure is a good start - without anythng removed or disconnected
from the machine:
Click to expand...


Thanks w_tom. ...wasn't going to buy a psu, I had a spare that I was going
to plug in.
....didn't even need to remove anything from the case. I would agree with you
that buying one would be a waste of time and money.
.... at this point with most everything removed, I can simply plug in the
spare psu quicker than I can test the old one.

Is BIOS battery defective? Again, don't even remove the battery. In
but seconds, meter could have measured battery and determined if
battery was connected. Removing and reseating BIOS battery could have
adversely changed some CMOS settings - just another example of
exponentially complicating a problem.
Click to expand...

okay, I already have a meter, but didn't have one where I was working on the
machine. I'll probably go back today and take the meter with me. The bios
settings were all default, other than the boot drive options, so I wasn't
really worried about them.

But then your every test still does not even say the power supply is
OK. Your each test has only said maybe this might be working OK. You
need tests that say "yes it is OK without a doubt so I will move on to
other suspects". Get the meter. Stop guessing.
Click to expand...

I agree with most of what you suggest. Here is what I don't know 100%, but I
think I have it correct - perhaps someone can confirm or correct me.

1) When idle, the PSU supplies a minimal voltage to the MB that simply
allows the case power button to work.
2) When the case power button is pushed, a circuit is completed in the MB,
3) The completed circuit sends a signal to the PSU and the PSU starts fully.
4) The PSU fan comes on at this point. The PSU sends a signal to the MB that
it is now on.
5) The bios then starts and attempts its POST.
6) The POST starts and will provide beep codes if something prevents it from
continuing to the point that you can get video output.


What is happening in my case is that the first 3 steps above are being
circumvented. As soon as 110 voltage is provided to the PSU, the entire
machine goes on without touching the case power switch and this occurs even
if the case power switch leads are removed from the MB. ...but if the PSU is
disconnected from the MB, then the PSU does not start - i.e., the fan does
not come on. I assume that whatever occurs in step 2 above, where something
tells the PSU to turn on completely, is the problem. I assume that this
occurs due to something within the MB itself and not within the processor or
elsewhere. As far as I know, if I would completely remove the processor, I
should still get to about step 4 in the normal way before halting.

I assume that the behavior causing steps 1, 2, and 3 to be skipped can only
be due to a bad MB and that this does not have anything to do with the bios
(so the bios battery can't be the problem?) and that it can't be due to
something within the PSU, since the PSU does properly stay off when
disconnected from the MB. I also assume that it can't be due to the case
power switch because it still occurs when the case power switch is
disconnected.

I would normally just scrap the thing as it is a bit old. ...but what is
going on is that this was my old primary machine that I just upgraded and I
was going to keep using it as a file server. So, I just spent two days
setting up an inexpensive raid 1 array in it, loading W2K3, doing all
patches, configuring it just the way I wanted, etc.. I don't want to repeat
all of that work (okay, mostly sitting around waiting for things to install,
but I can't walk away for long periods), and I can't just move the raid
array to a new machine. So replacing the MB might be the most time
efficient, since I should then be able to just keep the raid array as it is.

I would like like to narrow things down so that I know that it must be the
MB and nothing else before I spend the time and money getting a new one. I'm
not 100% positive, for example, that a bad bios battery can't in some way
cause this type of thing.
 
Okay,

One more/final update - the meter showed that the bios battery was okay at 3
volts.

Plugging in another power supply didn't do anything. The board still starts
immediately instead of waiting for the case power supply button to be
pushed.

If I completely pull out the processor, the board does not start (e.g., the
case fans don't go on), but if I put the processor back in, once again
everything starts up as soon as the 110 power is turned on.

I would assume that the MB is designed so that a processor is required for
anything to start so the above doesn't indicate anything about whether or
not the processor is okay. ...I've never experimented on any other machine,
however.

The only real remaining question is whether there is any remote possibility
that the problem could be in the processor instead of the MB. I'm about 99%
certain that it's the MB alone, but could someone confirm this before I
attempt to buy another board that I'll likely not be able to return.

Jeff
 
...have an older intel socket 478 motherboard in a 4 year old machine that I
was working on today.

The thing ran fine yesterday. Because this was an older spare machine, I
wasn't as careful as I normally am. I drilled a hole (with a cordless drill)
in the case to install another case fan. When I attempted to reboot, things
weren't working correctly.

As soon as the power supply is plugged in, the machine starts up even with
the case power switch completely disconnected. It won't boot and won't even
post. I get nothing on the video at all. If I pull the ram out, I get
warning beeps that seem to mean that there is no ram. ...but that's it. I've
pulled out and unplugged everything, but nothing helps. I've pulled out the
bios battery to reset. Again, the case power switch doesn't work and instead
the thing just comes on when the machine's power supply is switched on. I
didn't have a spare power supply to swap in, but I can't see how the psu
would matter. When unplugged the psu from the MB, the psu [properly] does
not come on.

As near as I can tell, running the drill into the front of the case must
have caused something to short in the board. okay, I know this was stupid,
and I would not have done this on a new machine.

Is there any other explanation other than a bad MB?
Click to expand...

Try taking it out of the case and running it. I was given a system
which was supposedly dead but it ran with no problem when I removed
all the parts and connected the PSU out of the box. I assume there was
a short against the metal casing. Putting it all back together in
another case solved it.

km
 
...
1) When idle, the PSU supplies a minimal voltage to the MB that simply
allows the case power button to work.
2) When the case power button is pushed, a circuit is completed in the MB,
3) The completed circuit sends a signal to the PSU and the PSU starts fully.
4) The PSU fan comes on at this point. The PSU sends a signal to the MB that
it is now on.
5) The bios then starts and attempts its POST.
6) The POST starts and will provide beep codes if something prevents it from
continuing to the point that you can get video output.

What is happening in my case is that the first 3 steps above are being
circumvented. As soon as 110 voltage is provided to the PSU, the entire
machine goes on without touching the case power switch and this occurs even
if the case power switch leads are removed from the MB. ...but if the PSU is
disconnected from the MB, then the PSU does not start - i.e., the fan does
not come on. I assume that whatever occurs in step 2 above, where something
tells the PSU to turn on completely, is the problem. I assume that this
occurs due to something within the MB itself and not within the processor or
elsewhere. As far as I know, if I would completely remove the processor, I
should still get to about step 4 in the normal way before halting.

I assume that the behavior causing steps 1, 2, and 3 to be skipped can only
be due to a bad MB and that this does not have anything to do with the bios
(so the bios battery can't be the problem?) and that it can't be due to
something within the PSU, since the PSU does properly stay off when
disconnected from the MB. I also assume that it can't be due to the case
power switch because it still occurs when the case power switch is
disconnected.
...
Click to expand...

The purple wire provides power to the power supply controller. Green
wire tells power supply to turn on. Power supply has but two seconds
to determine all voltages are OK; then tell controller via gray wire.
This and more is summarized in "PC doesn't start" on 15 Feb 2007 in
microsoft.public.windowsxp.hardware at:
http://tinyurl.com/25fvc7

If any one voltage does not try to start, then that would be the
suspect circuit. In your case (because that other supply is probably
good), then the voltage that does not try to turn on is where you look
for the problem on motherboard or peripherals.

Of course, because computer is powered even when powered off (see
voltage on purple wire), always remove power cord from AC wall
receptacle before connecting or disconnecting anything.

Pulling out processor will not discover anything useful. Processor
even with its own power supply rarely fail. Since the beep codes
exist, then processor and some other (limited) motherboard functions
are working.

If power supply only powers with some things removed, well that is
useful information even is computer does not boot. Also important are
those power supply voltage numbers and responses. They can provide
additional information that you don't realize. Numbers that make
others useful help.
 
Try taking it out of the case and running it. I was given a system
which was supposedly dead but it ran with no problem when I removed
all the parts and connected the PSU out of the box. I assume there was
a short against the metal casing. Putting it all back together in
another case solved it.

km
Click to expand...


Good suggestion. I tried removing the entire MB from the case and I get the
same. There seems to be some sort of short in the MB where the thing is
completing the circuit telling the powersupply to power immediately (after
110 voltage is applied to the supply) rather than wait until the case power
switch is turned on. As soon as I plug the psu into the 110 (actually turn
on the on/off toggle on the PSU itself) the bare MB that is now outside of
the case starts up with absolutely nothing connected to it other than the
processor.

I can't see how such a thing would occur, but I see no other explanation
other than a bad MB. I would hope that whatever fried the MB, however, left
the processor alone.

Jeff
 
Jeff said:
Good suggestion. I tried removing the entire MB from the case and I get
the same. There seems to be some sort of short in the MB where the thing
is completing the circuit telling the powersupply to power immediately
(after 110 voltage is applied to the supply) rather than wait until the
case power switch is turned on. As soon as I plug the psu into the 110
(actually turn on the on/off toggle on the PSU itself) the bare MB that is
now outside of the case starts up with absolutely nothing connected to it
other than the processor.

I can't see how such a thing would occur>>snip<<
Click to expand...
I'm sure you did not mean "see" in its literal meaning; however, reading
above posters'(w_tom & others, who seem knowledgeable) suggestions plus
having had one pertinent occurrence, I'd suggest finding a good magnifying
glass and carefully inspecting the mobo in the area where the drill likely
contacted it--don't forget both underside and topside. (I found an
invisible-to-my eye speck on a rather expensive mobo, freebie from owner who
bought a new one, where a trace had been cut when a tool apparently slipped
while installing a cpu-cooler. Inexpensive fix and mobo still operating
fine 2+ years later.) HTH & good luck, s
 
sdlomi2 said:
I'm sure you did not mean "see" in its literal meaning; however,
reading above posters'(w_tom & others, who seem knowledgeable) suggestions
plus having had one pertinent occurrence, I'd suggest finding a good
magnifying glass and carefully inspecting the mobo in the area where the
drill likely contacted it--don't forget both underside and topside. (I
found an invisible-to-my eye speck on a rather expensive mobo, freebie
from owner who bought a new one, where a trace had been cut when a tool
apparently slipped while installing a cpu-cooler. Inexpensive fix and
mobo still operating fine 2+ years later.)
Click to expand...


The drill didn't come anywhere near the MB. The board is in the back of the
case, and I simply poked a hole in a completely empty area on the front of
the case where the closest component (the hard drive) was at least 3 inches
away.

....but that's the only thing that I could imagine that might cause the
problem. ...for all I know, it was a coincidence that the thing went out
when it did.

....I mean others have recently mentioned not to even use a vaccuum to suck
out dust due to the static electricity and components come with warnings
about static.

I've already pulled the board and thoroughly blew off all dust and debris.
It continues to act in the same way.
 
Back
Top