From: Justin Shifflett on
I was told I should post here.

I don't want to have to replace the motherboard if I don't have to.
But I think I do.

My PC will turn on. The fans will run. But the PC won't detect the
keyboard. It will however detect the mouse (it's optical, and glows
red). Nothing will happen on the monitor, and the monitor light will
turn yellow like nothing is on-- but the monitor works fine, as it's
being used on the current PC I'm on.

I called HP (The company that made my computer). They went through
the steps with me-- getting Static Electricity out of the system,
reseating the Ram, listening for beeps (there weren't any as far as I
could tell.) ANd they said they thought it was a Motherboard issue.

I don't know what else to do. I've tried just about everything I can
think of. It's weird how the mouse will be read, but nothing else
will. I dont' know if it's the motherboard, the CPU, or what.

Any tips?
From: Paul on
Justin Shifflett wrote:
> I was told I should post here.
>
> I don't want to have to replace the motherboard if I don't have to.
> But I think I do.
>
> My PC will turn on. The fans will run. But the PC won't detect the
> keyboard. It will however detect the mouse (it's optical, and glows
> red). Nothing will happen on the monitor, and the monitor light will
> turn yellow like nothing is on-- but the monitor works fine, as it's
> being used on the current PC I'm on.
>
> I called HP (The company that made my computer). They went through
> the steps with me-- getting Static Electricity out of the system,
> reseating the Ram, listening for beeps (there weren't any as far as I
> could tell.) ANd they said they thought it was a Motherboard issue.
>
> I don't know what else to do. I've tried just about everything I can
> think of. It's weird how the mouse will be read, but nothing else
> will. I dont' know if it's the motherboard, the CPU, or what.
>
> Any tips?

I like to start with the model number. Like "Pavilion 1234" or the like.

Using the model number, I like to check to see whether that computer
has any "favorite" failure modes. Knowing the approximate age also
helps, as a particular power supply failure, ruins the motherboard and
things like the hard drive.

One test is to:

1) Remove all AC power to the computer. Unplug it to be safe.
2) Remove the RAM sticks and place them in an antistatic bag for
safe keeping.
3) Turn on the computer and listen for beeps.
4) If you hear beeps, it means the CPU is working. The CPU is what
helped make the beeps.
5) If turning off all power, adding one stick of RAM back, and powering
up yields no beep, after step (3) gave beeps, it could be a bad stick of
RAM. If the RAM is bad in low memory locations below 640K, sometimes
that causes the BIOS to die before there are beeps. (That has happened
to me on an Nforce2 motherboard.)

A bad CMOS battery has been known to prevent a PC from POSTing. Usually
you'll find a CR2032 coin cell on the motherboard, and it should be delivering
3.0 volts or a bit more. If it is much below 2.3 volts, that might cause
problems. Sometimes, it drops all the way to zero, and some motherboards
don't seem to be able to start when that happens. I don't really know the
reason why that happens, because as far as I know, all the important
circuitry should be getting power from the power supply.

You should also do a visual inspection for bad capacitors. In the pictures
here, you can see the area around the pressure relief seams is starting
to bulge, which means pressure is building up inside. The rubber seal on the
bottom of the electrolytic capacitor can also push out of the metal can, causing
the cap to tilt to one side. When a cap bursts, orange-brown liquid will dry
near the seams or spill onto the motherboard surface. For things like caps in
the Vcore circuit, that will eventually lead to a failure to POST, or instability
and crashing while in Windows.

http://www.badcaps.net/images/caps/kt7/image004.png

I've even had those bad capacitors, inside my Antec power supply. That
caused the supply to "sizzle" at startup, and eventually resulted in the
processor crashing in the BIOS.

Paul
From: Justin Shifflett on
On 11 July, 22:44, Paul <nos...(a)needed.com> wrote:
> Justin Shifflett wrote:
> > I was told I should post here.
>
> > I don't want to have to replace the motherboard if I don't have to.
> > But I think I do.
>
> > My PC will turn on. The fans will run. But the PC won't detect the
> > keyboard. It will however detect the mouse (it's optical, and glows
> > red). Nothing will happen on the monitor, and the monitor light will
> > turn yellow like nothing is on-- but the monitor works fine, as it's
> > being used on the current PC I'm on.
>
> > I called HP (The company that made my computer). They went through
> > the steps with me-- getting Static Electricity out of the system,
> > reseating the Ram, listening for beeps (there weren't any as far as I
> > could tell.) ANd they said they thought it was a Motherboard issue.
>
> > I don't know what else to do. I've tried just about everything I can
> > think of. It's weird how the mouse will be read, but nothing else
> > will. I dont' know if it's the motherboard, the CPU, or what.
>
> > Any tips?
>
> I like to start with the model number. Like "Pavilion 1234" or the like.
>
> Using the model number, I like to check to see whether that computer
> has any "favorite" failure modes. Knowing the approximate age also
> helps, as a particular power supply failure, ruins the motherboard and
> things like the hard drive.
>
> One test is to:
>
> 1) Remove all AC power to the computer. Unplug it to be safe.
> 2) Remove the RAM sticks and place them in an antistatic bag for
> safe keeping.
> 3) Turn on the computer and listen for beeps.
> 4) If you hear beeps, it means the CPU is working. The CPU is what
> helped make the beeps.
> 5) If turning off all power, adding one stick of RAM back, and powering
> up yields no beep, after step (3) gave beeps, it could be a bad stick of
> RAM. If the RAM is bad in low memory locations below 640K, sometimes
> that causes the BIOS to die before there are beeps. (That has happened
> to me on an Nforce2 motherboard.)
>
> A bad CMOS battery has been known to prevent a PC from POSTing. Usually
> you'll find a CR2032 coin cell on the motherboard, and it should be delivering
> 3.0 volts or a bit more. If it is much below 2.3 volts, that might cause
> problems. Sometimes, it drops all the way to zero, and some motherboards
> don't seem to be able to start when that happens. I don't really know the
> reason why that happens, because as far as I know, all the important
> circuitry should be getting power from the power supply.
>
> You should also do a visual inspection for bad capacitors. In the pictures
> here, you can see the area around the pressure relief seams is starting
> to bulge, which means pressure is building up inside. The rubber seal on the
> bottom of the electrolytic capacitor can also push out of the metal can, causing
> the cap to tilt to one side. When a cap bursts, orange-brown liquid will dry
> near the seams or spill onto the motherboard surface. For things like caps in
> the Vcore circuit, that will eventually lead to a failure to POST, or instability
> and crashing while in Windows.
>
> http://www.badcaps.net/images/caps/kt7/image004.png
>
> I've even had those bad capacitors, inside my Antec power supply. That
> caused the supply to "sizzle" at startup, and eventually resulted in the
> processor crashing in the BIOS.
>
> Paul

The model is HP Pavilion a600f It's almost 2 years old (Got in August
of 08)

With all ram removed, the computer makes absolutely no beeps.

I've been having BSOD's recently, mostly on Cold Boot. Then things
would be fine.

My motherboard, if it matters, is a Foxconn MCP73M02H1

The capacitors look fine.

I really appreciate the help, and you really seem to know your stuff.
From: Paul on
Justin Shifflett wrote:
> On 11 July, 22:44, Paul <nos...(a)needed.com> wrote:
>> Justin Shifflett wrote:
>>> I was told I should post here.
>>> I don't want to have to replace the motherboard if I don't have to.
>>> But I think I do.
>>> My PC will turn on. The fans will run. But the PC won't detect the
>>> keyboard. It will however detect the mouse (it's optical, and glows
>>> red). Nothing will happen on the monitor, and the monitor light will
>>> turn yellow like nothing is on-- but the monitor works fine, as it's
>>> being used on the current PC I'm on.
>>> I called HP (The company that made my computer). They went through
>>> the steps with me-- getting Static Electricity out of the system,
>>> reseating the Ram, listening for beeps (there weren't any as far as I
>>> could tell.) ANd they said they thought it was a Motherboard issue.
>>> I don't know what else to do. I've tried just about everything I can
>>> think of. It's weird how the mouse will be read, but nothing else
>>> will. I dont' know if it's the motherboard, the CPU, or what.
>>> Any tips?
>> I like to start with the model number. Like "Pavilion 1234" or the like.
>>
>> Using the model number, I like to check to see whether that computer
>> has any "favorite" failure modes. Knowing the approximate age also
>> helps, as a particular power supply failure, ruins the motherboard and
>> things like the hard drive.
>>
>> One test is to:
>>
>> 1) Remove all AC power to the computer. Unplug it to be safe.
>> 2) Remove the RAM sticks and place them in an antistatic bag for
>> safe keeping.
>> 3) Turn on the computer and listen for beeps.
>> 4) If you hear beeps, it means the CPU is working. The CPU is what
>> helped make the beeps.
>> 5) If turning off all power, adding one stick of RAM back, and powering
>> up yields no beep, after step (3) gave beeps, it could be a bad stick of
>> RAM. If the RAM is bad in low memory locations below 640K, sometimes
>> that causes the BIOS to die before there are beeps. (That has happened
>> to me on an Nforce2 motherboard.)
>>
>> A bad CMOS battery has been known to prevent a PC from POSTing. Usually
>> you'll find a CR2032 coin cell on the motherboard, and it should be delivering
>> 3.0 volts or a bit more. If it is much below 2.3 volts, that might cause
>> problems. Sometimes, it drops all the way to zero, and some motherboards
>> don't seem to be able to start when that happens. I don't really know the
>> reason why that happens, because as far as I know, all the important
>> circuitry should be getting power from the power supply.
>>
>> You should also do a visual inspection for bad capacitors. In the pictures
>> here, you can see the area around the pressure relief seams is starting
>> to bulge, which means pressure is building up inside. The rubber seal on the
>> bottom of the electrolytic capacitor can also push out of the metal can, causing
>> the cap to tilt to one side. When a cap bursts, orange-brown liquid will dry
>> near the seams or spill onto the motherboard surface. For things like caps in
>> the Vcore circuit, that will eventually lead to a failure to POST, or instability
>> and crashing while in Windows.
>>
>> http://www.badcaps.net/images/caps/kt7/image004.png
>>
>> I've even had those bad capacitors, inside my Antec power supply. That
>> caused the supply to "sizzle" at startup, and eventually resulted in the
>> processor crashing in the BIOS.
>>
>> Paul
>
> The model is HP Pavilion a600f It's almost 2 years old (Got in August
> of 08)
>
> With all ram removed, the computer makes absolutely no beeps.
>
> I've been having BSOD's recently, mostly on Cold Boot. Then things
> would be fine.
>
> My motherboard, if it matters, is a Foxconn MCP73M02H1
>
> The capacitors look fine.
>
> I really appreciate the help, and you really seem to know your stuff.

Do you have any warranty left on the thing ? You don't want
to pay for stuff you don't have to, if there is warranty
left on it.

Based on your "having BSOD's recently, mostly on Cold Boot, then things
would be fine", I'd say this is a power supply issue. It sounds almost
exactly like my Antec power supply symptoms. I'd probably
start by checking supply voltages with a multimeter, and verify the
Power_Good signal on the main connector, was being asserted after
you start the thing. The power supply sends the logic signal Power_Good,
after all the voltages have stabilized out of the power supply. The
motherboard comes out of reset, once Power_Good is there, and
any "local Power_Good" from the motherboard itself, is true.
The motherboard has its own onboard regulators, and some of those
have a status signal indicating they're ready. When all power
regulating devices are happy, that is when the BIOS POST starts.

You can find copies of the ATX spec, if you're curious about how it works,
or what color wire does what function.

http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf

Depending on what is appearing in the BSOD, that may help refine
the problem. If BSODs crash in random pieces of code every time,
then you'd assume it is some outside force like the power supply
failing. If the same routine is named every time, like something
beginning in "nv", it could be a graphics driver running into
problems with a failing GPU inside your Northbridge (under the
heatsink). The predecessor to the 7100, the 6100/6150, had
a few problems with poorly tested built-in GPUs. Some people
saw colored blocks from their built-in graphics.

http://h10025.www1.hp.com/ewfrf/wc/document?docname=c01357119&lc=en&cc=us&dlc=en&product=3644689

If you don't have a multimeter, and aren't interested in probing
stuff like that, you can start by swapping the power supply. Or
have a local shop do the debug for you, and identify to the nearest
failed component, what is busted ("motherboard" or "power supply").
Processors hardly ever fail, your RAM test didn't conclusively
indicate a problem, so at the moment, I'd have to assume motherboard
or PSU as a starting point. And PSU is the easiest one to swap first.
(You can even use a 20 pin supply in the 24 pin connector, as the
extra four pins are redundant and other wires already carry that
power.)

(20 pin goes up against one end of the 24 pin connector, latch side to latch side)
http://www.playtool.com/pages/psuconnectors/20in24.jpg

If I'm mistaken, and you've actually had a video card stuffed in
the video card slot, you could also pull that out, and try running
the computer from the built-in video. My suggestions above are
based on assuming you're using the 7100 built-in graphics (rear
video connectors in the I/O plate area). Some Nvidia chipset
motherboards in the past, have had excessively sensitive PCI Express
slots, such that a slot would blow out without much provocation.

This is a really silly idea, but every once in a while, a poster
has a problem like this, and the problem turns out to be a loose
power cord. They work loose, where they plug into the supply on the
back of the computer. That sometimes simulates a "weak supply" syndrome.
Pretty low odds, but I have to mention it.

In addition to the BSODs, sometimes the supply will give additional
hints -

1) A puff of smoke at startup - almost looks like dust.
Perhaps you can also smell something a bit different than normal.

2) Weird noises. Sizzling. Coil noise. Whining. Anything that
might indicate a component under stress.

3) When computers have fixed speed fans, after a while, you get used
to the "stability level" of the fans. The human ear is sensitive
to pitch. On two power supplies here, just before they failed,
the fan speed on the fixed fans started to "wander" a bit. That's
another thing I look for, is a change in the stable tone of fixed
speed fans. If your fans are speed controlled, and constantly
varying, then you can't use this criterion. HPs and Dells are likely
to have variable speed cooling fans, so this test probably won't
help you.

HTH,
Paul
From: Justin Shifflett on
On 11 July, 23:21, Paul <nos...(a)needed.com> wrote:
> Justin Shifflett wrote:
> > On 11 July, 22:44, Paul <nos...(a)needed.com> wrote:
> >> Justin Shifflett wrote:
> >>> I was told I should post here.
> >>> I don't want to have to replace the motherboard if I don't have to.
> >>> But I think I do.
> >>> My PC will turn on. The fans will run. But the PC won't detect the
> >>> keyboard. It will however detect the mouse (it's optical, and glows
> >>> red). Nothing will happen on the monitor, and the monitor light will
> >>> turn yellow like nothing is on-- but the monitor works fine, as it's
> >>> being used on the current PC I'm on.
> >>> I called HP (The company that made my computer). They went through
> >>> the steps with me-- getting Static Electricity out of the system,
> >>> reseating the Ram, listening for beeps (there weren't any as far as I
> >>> could tell.) ANd they said they thought it was a Motherboard issue.
> >>> I don't know what else to do. I've tried just about everything I can
> >>> think of. It's weird how the mouse will be read, but nothing else
> >>> will. I dont' know if it's the motherboard, the CPU, or what.
> >>> Any tips?
> >> I like to start with the model number. Like "Pavilion 1234" or the like.
>
> >> Using the model number, I like to check to see whether that computer
> >> has any "favorite" failure modes. Knowing the approximate age also
> >> helps, as a particular power supply failure, ruins the motherboard and
> >> things like the hard drive.
>
> >> One test is to:
>
> >> 1) Remove all AC power to the computer. Unplug it to be safe.
> >> 2) Remove the RAM sticks and place them in an antistatic bag for
> >> safe keeping.
> >> 3) Turn on the computer and listen for beeps.
> >> 4) If you hear beeps, it means the CPU is working. The CPU is what
> >> helped make the beeps.
> >> 5) If turning off all power, adding one stick of RAM back, and powering
> >> up yields no beep, after step (3) gave beeps, it could be a bad stick of
> >> RAM. If the RAM is bad in low memory locations below 640K, sometimes
> >> that causes the BIOS to die before there are beeps. (That has happened
> >> to me on an Nforce2 motherboard.)
>
> >> A bad CMOS battery has been known to prevent a PC from POSTing. Usually
> >> you'll find a CR2032 coin cell on the motherboard, and it should be delivering
> >> 3.0 volts or a bit more. If it is much below 2.3 volts, that might cause
> >> problems. Sometimes, it drops all the way to zero, and some motherboards
> >> don't seem to be able to start when that happens. I don't really know the
> >> reason why that happens, because as far as I know, all the important
> >> circuitry should be getting power from the power supply.
>
> >> You should also do a visual inspection for bad capacitors. In the pictures
> >> here, you can see the area around the pressure relief seams is starting
> >> to bulge, which means pressure is building up inside. The rubber seal on the
> >> bottom of the electrolytic capacitor can also push out of the metal can, causing
> >> the cap to tilt to one side. When a cap bursts, orange-brown liquid will dry
> >> near the seams or spill onto the motherboard surface. For things like caps in
> >> the Vcore circuit, that will eventually lead to a failure to POST, or instability
> >> and crashing while in Windows.
>
> >>http://www.badcaps.net/images/caps/kt7/image004.png
>
> >> I've even had those bad capacitors, inside my Antec power supply. That
> >> caused the supply to "sizzle" at startup, and eventually resulted in the
> >> processor crashing in the BIOS.
>
> >> Paul
>
> > The model is HP Pavilion a600f It's almost 2 years old (Got in August
> > of 08)
>
> > With all ram removed, the computer makes absolutely no beeps.
>
> > I've been having BSOD's recently, mostly on Cold Boot. Then things
> > would be fine.
>
> > My motherboard, if it matters, is a Foxconn MCP73M02H1
>
> > The capacitors look fine.
>
> > I really appreciate the help, and you really seem to know your stuff.
>
> Do you have any warranty left on the thing ? You don't want
> to pay for stuff you don't have to, if there is warranty
> left on it.
>
> Based on your "having BSOD's recently, mostly on Cold Boot, then things
> would be fine", I'd say this is a power supply issue. It sounds almost
> exactly like my Antec power supply symptoms. I'd probably
> start by checking supply voltages with a multimeter, and verify the
> Power_Good signal on the main connector, was being asserted after
> you start the thing. The power supply sends the logic signal Power_Good,
> after all the voltages have stabilized out of the power supply. The
> motherboard comes out of reset, once Power_Good is there, and
> any "local Power_Good" from the motherboard itself, is true.
> The motherboard has its own onboard regulators, and some of those
> have a status signal indicating they're ready. When all power
> regulating devices are happy, that is when the BIOS POST starts.
>
> You can find copies of the ATX spec, if you're curious about how it works,
> or what color wire does what function.
>
> http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2...
>
> Depending on what is appearing in the BSOD, that may help refine
> the problem. If BSODs crash in random pieces of code every time,
> then you'd assume it is some outside force like the power supply
> failing. If the same routine is named every time, like something
> beginning in "nv", it could be a graphics driver running into
> problems with a failing GPU inside your Northbridge (under the
> heatsink). The predecessor to the 7100, the 6100/6150, had
> a few problems with poorly tested built-in GPUs. Some people
> saw colored blocks from their built-in graphics.
>
> http://h10025.www1.hp.com/ewfrf/wc/document?docname=c01357119&lc=en&c...
>
> If you don't have a multimeter, and aren't interested in probing
> stuff like that, you can start by swapping the power supply. Or
> have a local shop do the debug for you, and identify to the nearest
> failed component, what is busted ("motherboard" or "power supply").
> Processors hardly ever fail, your RAM test didn't conclusively
> indicate a problem, so at the moment, I'd have to assume motherboard
> or PSU as a starting point. And PSU is the easiest one to swap first.
> (You can even use a 20 pin supply in the 24 pin connector, as the
> extra four pins are redundant and other wires already carry that
> power.)
>
> (20 pin goes up against one end of the 24 pin connector, latch side to latch side)http://www.playtool.com/pages/psuconnectors/20in24.jpg
>
> If I'm mistaken, and you've actually had a video card stuffed in
> the video card slot, you could also pull that out, and try running
> the computer from the built-in video. My suggestions above are
> based on assuming you're using the 7100 built-in graphics (rear
> video connectors in the I/O plate area). Some Nvidia chipset
> motherboards in the past, have had excessively sensitive PCI Express
> slots, such that a slot would blow out without much provocation.
>
> This is a really silly idea, but every once in a while, a poster
> has a problem like this, and the problem turns out to be a loose
> power cord. They work loose, where they plug into the supply on the
> back of the computer. That sometimes simulates a "weak supply" syndrome.
> Pretty low odds, but I have to mention it.
>
> In addition to the BSODs, sometimes the supply will give additional
> hints -
>
> 1) A puff of smoke at startup - almost looks like dust.
> Perhaps you can also smell something a bit different than normal.
>
> 2) Weird noises. Sizzling. Coil noise. Whining. Anything that
> might indicate a component under stress.
>
> 3) When computers have fixed speed fans, after a while, you get used
> to the "stability level" of the fans. The human ear is sensitive
> to pitch. On two power supplies here, just before they failed,
> the fan speed on the fixed fans started to "wander" a bit. That's
> another thing I look for, is a change in the stable tone of fixed
> speed fans. If your fans are speed controlled, and constantly
> varying, then you can't use this criterion. HPs and Dells are likely
> to have variable speed cooling fans, so this test probably won't
> help you.
>
> HTH,
> Paul

The fans come on, the power light comes on. And the mouse is read.
Shouldn't that mean that it's not a PSU Issue? I went through a flow
chart, and it moved me to motherboard issue or CPU issue.

My bsod's cited:
svchost.exe
WerFault.exe
services.exe
ntkrnlmp.exe ( nt! ?? ::FNODOBFM::`string'+1829 )

The last one cited Firefox.