ASROCK problem

I'm having trouble getting my desktop computer to boot up today.

I'd left it on continuously for two or three days and finally shut it
down when I went to bed last night. When I tried to boot it up this
morning, I got basically nothing.

The fans start immediately but I don't even get the single beep. I'm not
seeing a Windows logo or even the first screen which tells me what do
press to get into the BIOS. Pressing F2 does not get me to the BIOS when
I start pressing it repeatedly from the moment I powered up.

I thought it might simply be that the video cable had gotten dislodged
but I checked it where it connects to the monitor, where it goes into
its extension and where it goes into the case and all connections seem
tight. I also opened the case and looked for anything that might not be
seated properly, particularly hard drives, but everything seems to be
properly seated.

I'm not very good with this kind of hardware issue so I need some advice
on what to check next.

The desktop has been working fine since I reseated the SATA cables (see
my post in September for details). I'm not sure what caused this latest
problem. I have two cats and it's always possible they bumped something
as they walked around the computer but that's about the only vague
glimmer of an idea I have.


--
Rhino

Rhino wrote:
I'm having trouble getting my desktop computer to boot up today.

I'd left it on continuously for two or three days and finally shut it
down when I went to bed last night. When I tried to boot it up this
morning, I got basically nothing.

The fans start immediately but I don't even get the single beep. I'm not
seeing a Windows logo or even the first screen which tells me what do
press to get into the BIOS. Pressing F2 does not get me to the BIOS when
I start pressing it repeatedly from the moment I powered up.

I thought it might simply be that the video cable had gotten dislodged
but I checked it where it connects to the monitor, where it goes into
its extension and where it goes into the case and all connections seem
tight. I also opened the case and looked for anything that might not be
seated properly, particularly hard drives, but everything seems to be
properly seated.

I'm not very good with this kind of hardware issue so I need some advice
on what to check next.

The desktop has been working fine since I reseated the SATA cables (see
my post in September for details). I'm not sure what caused this latest
problem. I have two cats and it's always possible they bumped something
as they walked around the computer but that's about the only vague
glimmer of an idea I have.



When you're having problems, the best direction is to
"simplify" the system and retest.

For example, you can fully power off the system, pull the
plug, then remove the DIMMs and store them in an antistatic
bag. If the system has a video card (non-integrated video),
you can remove the video card as well.

Now, power up the system, and listen for a repeating beep code.
That will be an error beep code. The BIOS firmware and the
processor running it, make that repeating code. That tells
you the processor is working. If you add the RAM back in,
the beep code pattern should change. Finally, adding in the
video card, should again remove all error reasons, and then you'll
be back to just the one beep at startup.

It's possible for the reset button to getting stuck. The
board stays in reset and gives your symptoms.

Check that the ATX12V cable is fully seated. A good
board isolates ATX12V from 12V on the main cable, and
the processor won't get any 12V if the ATX12V isn't
plugged in. The cables have latches to prevent them
from working loose.

Clearing CMOS can sometimes help, if something got
corrupted there. (Again, that's a procedure you
do with the power cord unplugged, to prevent damage.
BIOS settings must be reloaded later if you decide
to try that one, like enabling AHCI or whatever.)

If you have a multimeter, you can connect the black
lead of the multimeter to the chassis, then touch
the red probe to the top of the CMOS CR2032 coin cell
and check that it's around 3V or so. Some motherboards,
they refuse to start if the battery is flat. Since your
motherboard isn't that old, it would be hard to understand
the battery being flat already, but it's something to
check.

That's all that comes to mind right now. Remove stuff,
listen for a beep code. If you can't get a response
under any circumstances, next you'd do a check
on the power supply. With the multimeter, you'd check
for operating voltages on the thing. You'd also make
sure the fan is spinning on the supply (as proof
that the motherboard sent PS_ON# a logic low or
grounded zero volt signal). If the power supply
is running, there is a kind of "POWER_GOOD" signal
on the main cable, as well as all the main voltages
to check.

(Pinout and details, here.)

http://www.formfactors.org/developer...public_br2.pdf

So when it's simplified, you have power supply,
motherboard, processor. You can stick the meter on
the power supply, to eliminate that (all voltages within 5%,
the power status signal OK). And then that
leaves motherboard or processor. Processors hardly
ever fail (you can reseat the processor to verify
it's not a contact problem). That leaves bad motherboard.

I probe the back of the ATX main cable where the wires
go in, to get voltage readings. But you need a strong
light to be able to see that, and some motherboards
don't make access to the main cable, all that easy.
The nylon shell has just enough room
to put the voltage probe red lead down in the
shell, until it comes in contact with the metal
of the pin in there.

You check 3.3V, 5V, 12V, -12V, +5VSB are within
5% of nominal. Check PS_ON# is low (close to zero
volts). The power supply fan wouldn't run without
that being true. That's what turns on the supply.
And whatever signal indicates POWER_GOOD, that one
is probably a logic 1, high value, and would be
around 5V.

When I say 5% of nominal, on 12V that means the
voltage level will be between 11.4V and 12.6V. The
power supply isn't infinitely accurate or anything,
and will always be off a bit. The power supply
typically uses "bulk regulation" and on the primary
side there is only one control that says "make more
voltage or make less voltage". So the individual voltages
are only established by turns ratio, and are not
controlled individually as such. The most heavily
loaded rail ends up on the low side, because of
how the bulk regulation works. But it should
still stay within the 5% range.

There is a picture of a power supply schematic here,
if you like pretty pictures. While this won't
make any sense the first time you look at it,
it grows on you. Modern supplies have more
circuitry than this, and this much circuit
"just makes the voltages". Some will have
a bit more monitoring. And a few modern ones,
the architecture is entirely different (two
stage regulation).

http://www.pavouk.org/hw/en_atxps.html

Paul

On 2013-10-31 8:36 PM, Paul wrote:
Rhino wrote:
I'm having trouble getting my desktop computer to boot up today.

I'd left it on continuously for two or three days and finally shut it
down when I went to bed last night. When I tried to boot it up this
morning, I got basically nothing.

The fans start immediately but I don't even get the single beep. I'm
not seeing a Windows logo or even the first screen which tells me what
do press to get into the BIOS. Pressing F2 does not get me to the BIOS
when I start pressing it repeatedly from the moment I powered up.

I thought it might simply be that the video cable had gotten dislodged
but I checked it where it connects to the monitor, where it goes into
its extension and where it goes into the case and all connections seem
tight. I also opened the case and looked for anything that might not
be seated properly, particularly hard drives, but everything seems to
be properly seated.

I'm not very good with this kind of hardware issue so I need some
advice on what to check next.

The desktop has been working fine since I reseated the SATA cables
(see my post in September for details). I'm not sure what caused this
latest problem. I have two cats and it's always possible they bumped
something as they walked around the computer but that's about the only
vague glimmer of an idea I have.



When you're having problems, the best direction is to
"simplify" the system and retest.

For example, you can fully power off the system, pull the
plug, then remove the DIMMs and store them in an antistatic
bag. If the system has a video card (non-integrated video),
you can remove the video card as well.

Now, power up the system, and listen for a repeating beep code.
That will be an error beep code. The BIOS firmware and the
processor running it, make that repeating code. That tells
you the processor is working. If you add the RAM back in,
the beep code pattern should change. Finally, adding in the
video card, should again remove all error reasons, and then you'll
be back to just the one beep at startup.

It's possible for the reset button to getting stuck. The
board stays in reset and gives your symptoms.

Check that the ATX12V cable is fully seated. A good
board isolates ATX12V from 12V on the main cable, and
the processor won't get any 12V if the ATX12V isn't
plugged in. The cables have latches to prevent them
from working loose.

Clearing CMOS can sometimes help, if something got
corrupted there. (Again, that's a procedure you
do with the power cord unplugged, to prevent damage.
BIOS settings must be reloaded later if you decide
to try that one, like enabling AHCI or whatever.)

If you have a multimeter, you can connect the black
lead of the multimeter to the chassis, then touch
the red probe to the top of the CMOS CR2032 coin cell
and check that it's around 3V or so. Some motherboards,
they refuse to start if the battery is flat. Since your
motherboard isn't that old, it would be hard to understand
the battery being flat already, but it's something to
check.

That's all that comes to mind right now. Remove stuff,
listen for a beep code. If you can't get a response
under any circumstances, next you'd do a check
on the power supply. With the multimeter, you'd check
for operating voltages on the thing. You'd also make
sure the fan is spinning on the supply (as proof
that the motherboard sent PS_ON# a logic low or
grounded zero volt signal). If the power supply
is running, there is a kind of "POWER_GOOD" signal
on the main cable, as well as all the main voltages
to check.

(Pinout and details, here.)

http://www.formfactors.org/developer...public_br2.pdf

So when it's simplified, you have power supply,
motherboard, processor. You can stick the meter on
the power supply, to eliminate that (all voltages within 5%,
the power status signal OK). And then that
leaves motherboard or processor. Processors hardly
ever fail (you can reseat the processor to verify
it's not a contact problem). That leaves bad motherboard.

I probe the back of the ATX main cable where the wires
go in, to get voltage readings. But you need a strong
light to be able to see that, and some motherboards
don't make access to the main cable, all that easy.
The nylon shell has just enough room
to put the voltage probe red lead down in the
shell, until it comes in contact with the metal
of the pin in there.

You check 3.3V, 5V, 12V, -12V, +5VSB are within
5% of nominal. Check PS_ON# is low (close to zero
volts). The power supply fan wouldn't run without
that being true. That's what turns on the supply.
And whatever signal indicates POWER_GOOD, that one
is probably a logic 1, high value, and would be
around 5V.

When I say 5% of nominal, on 12V that means the
voltage level will be between 11.4V and 12.6V. The
power supply isn't infinitely accurate or anything,
and will always be off a bit. The power supply
typically uses "bulk regulation" and on the primary
side there is only one control that says "make more
voltage or make less voltage". So the individual voltages
are only established by turns ratio, and are not
controlled individually as such. The most heavily
loaded rail ends up on the low side, because of
how the bulk regulation works. But it should
still stay within the 5% range.

There is a picture of a power supply schematic here,
if you like pretty pictures. While this won't
make any sense the first time you look at it,
it grows on you. Modern supplies have more
circuitry than this, and this much circuit
"just makes the voltages". Some will have
a bit more monitoring. And a few modern ones,
the architecture is entirely different (two
stage regulation).

http://www.pavouk.org/hw/en_atxps.html

Paul
Thanks for all the info, Paul. I have some more for you; I hope it makes
sense.

I think you've already dug up my motherboard model number but here it is
again: ASRock N68C-GS FX.

I hadn't seen your reply yet - I don't have internet at my new place yet
so have to go to the library or a friend's place to go online - so I
went about the problem somewhat differently. I tried booting from my
Unbuntu 12.04 LTS CD but had the same issues as when I booted from XP on
the hard drive. I also tried booting from an old Knoppix CD that I had
lying around; same issues again.

The monitor is a Samsung T260 and it is clearly getting power since it
cycles through the HDMI/Digital/Analog options. (I'm not sure how to put
this correctly but when I power up the computer and if the monitor is
on, it displays HDMI in the upper left corner for several seconds, then
Digital, then Analog. I assume this is because there are connections for
HDMI, DVI and RGB at the back and it is not sure where the signal will
come from so it checks all three.) Anyway, I noticed that after it
cycles through all three options two or three times, it displays an
error message: Check signal cable.

That made me believe the cable was faulty or improperly seated so I had
a look at it. It seemed to be seated just fine. There was an RGB male to
RGB female cable running out of the monitor; then, there was an RGB male
to RGB male connected to the end of that, serving to extend the cable;
then the RGB male end went into the back of the tower. But everything
seemed snug so I wasn't sure what to do next. I unplugged the RGB cable
where it went into the monitor and tried reseat it. It was a bit of a
struggle to get the orientation right so I looked at the end of the
cable and saw that two of the middle row of pins looked wrong: one of
the pins seemed to be missing entirely, and the other seemed to be
pushed halfway back into the plug. That looked like exactly the kind of
thing that could cause my problem so I removed that cable entirely and
just ran the RGB male to RGB male cable directly from the monitor to the
tower; all of its pins on both ends seemed fine. But I still get the
"check signal cable" message from the monitor so that obviously DIDN'T
solve the problem. So I'm back to being confused about what is wrong.

For what it's worth, I tried connecting the RGB cable from my monitor to
my laptop and then used F8 to cycle between the various options like "PC
Screen only", "Extend", "Duplicate" and "Second screen only". This made
it clear that the monitor has no problem displaying the desktop of the
laptop so the monitor itself (and apparently its RGB port) would appear
to be fine.

Could the RBG port on the mobo be failing? Maybe I just need a "gender
bender" to convert the monitor cable to female at the tower end so that
I can use the male RGB port which is right above the female RGB port.

I'm guessing that there's nothing else wrong. Does that make any sense?
Like I said, I'm lousy at this hardware stuff ;-)

Unfortunately, I had to put my voltmeter in storage (out of town) and
don't have easy access to another one. I also downsized dramatically
during my move and got rid of various old monitors, cables, and such
things that would come in really handy right now for diagnostic
purposes. That, and the lack of internet access at home, make this
problem a lot harder to solve than it otherwise would be....

I'm not sure if I'll see your reply to this before I go offline again
for the day so, if I don't, I'll try the other stuff you suggested when
I get home (or at least as much as I can without a voltmeter). But I'm
thinking the "check signal cable" message is the Big Clue we need to
sort this out. I wish I had thought to mention that when I first posted
but I didn't see this symptom until today....



--
Rhino

Rhino wrote:
On 2013-10-31 8:36 PM, Paul wrote:
Rhino wrote:
I'm having trouble getting my desktop computer to boot up today.

I'd left it on continuously for two or three days and finally shut it
down when I went to bed last night. When I tried to boot it up this
morning, I got basically nothing.

The fans start immediately but I don't even get the single beep. I'm
not seeing a Windows logo or even the first screen which tells me what
do press to get into the BIOS. Pressing F2 does not get me to the BIOS
when I start pressing it repeatedly from the moment I powered up.

I thought it might simply be that the video cable had gotten dislodged
but I checked it where it connects to the monitor, where it goes into
its extension and where it goes into the case and all connections seem
tight. I also opened the case and looked for anything that might not
be seated properly, particularly hard drives, but everything seems to
be properly seated.

I'm not very good with this kind of hardware issue so I need some
advice on what to check next.

The desktop has been working fine since I reseated the SATA cables
(see my post in September for details). I'm not sure what caused this
latest problem. I have two cats and it's always possible they bumped
something as they walked around the computer but that's about the only
vague glimmer of an idea I have.



When you're having problems, the best direction is to
"simplify" the system and retest.

For example, you can fully power off the system, pull the
plug, then remove the DIMMs and store them in an antistatic
bag. If the system has a video card (non-integrated video),
you can remove the video card as well.

Now, power up the system, and listen for a repeating beep code.
That will be an error beep code. The BIOS firmware and the
processor running it, make that repeating code. That tells
you the processor is working. If you add the RAM back in,
the beep code pattern should change. Finally, adding in the
video card, should again remove all error reasons, and then you'll
be back to just the one beep at startup.

It's possible for the reset button to getting stuck. The
board stays in reset and gives your symptoms.

Check that the ATX12V cable is fully seated. A good
board isolates ATX12V from 12V on the main cable, and
the processor won't get any 12V if the ATX12V isn't
plugged in. The cables have latches to prevent them
from working loose.

Clearing CMOS can sometimes help, if something got
corrupted there. (Again, that's a procedure you
do with the power cord unplugged, to prevent damage.
BIOS settings must be reloaded later if you decide
to try that one, like enabling AHCI or whatever.)

If you have a multimeter, you can connect the black
lead of the multimeter to the chassis, then touch
the red probe to the top of the CMOS CR2032 coin cell
and check that it's around 3V or so. Some motherboards,
they refuse to start if the battery is flat. Since your
motherboard isn't that old, it would be hard to understand
the battery being flat already, but it's something to
check.

That's all that comes to mind right now. Remove stuff,
listen for a beep code. If you can't get a response
under any circumstances, next you'd do a check
on the power supply. With the multimeter, you'd check
for operating voltages on the thing. You'd also make
sure the fan is spinning on the supply (as proof
that the motherboard sent PS_ON# a logic low or
grounded zero volt signal). If the power supply
is running, there is a kind of "POWER_GOOD" signal
on the main cable, as well as all the main voltages
to check.

(Pinout and details, here.)

http://www.formfactors.org/developer...public_br2.pdf

So when it's simplified, you have power supply,
motherboard, processor. You can stick the meter on
the power supply, to eliminate that (all voltages within 5%,
the power status signal OK). And then that
leaves motherboard or processor. Processors hardly
ever fail (you can reseat the processor to verify
it's not a contact problem). That leaves bad motherboard.

I probe the back of the ATX main cable where the wires
go in, to get voltage readings. But you need a strong
light to be able to see that, and some motherboards
don't make access to the main cable, all that easy.
The nylon shell has just enough room
to put the voltage probe red lead down in the
shell, until it comes in contact with the metal
of the pin in there.

You check 3.3V, 5V, 12V, -12V, +5VSB are within
5% of nominal. Check PS_ON# is low (close to zero
volts). The power supply fan wouldn't run without
that being true. That's what turns on the supply.
And whatever signal indicates POWER_GOOD, that one
is probably a logic 1, high value, and would be
around 5V.

When I say 5% of nominal, on 12V that means the
voltage level will be between 11.4V and 12.6V. The
power supply isn't infinitely accurate or anything,
and will always be off a bit. The power supply
typically uses "bulk regulation" and on the primary
side there is only one control that says "make more
voltage or make less voltage". So the individual voltages
are only established by turns ratio, and are not
controlled individually as such. The most heavily
loaded rail ends up on the low side, because of
how the bulk regulation works. But it should
still stay within the 5% range.

There is a picture of a power supply schematic here,
if you like pretty pictures. While this won't
make any sense the first time you look at it,
it grows on you. Modern supplies have more
circuitry than this, and this much circuit
"just makes the voltages". Some will have
a bit more monitoring. And a few modern ones,
the architecture is entirely different (two
stage regulation).

http://www.pavouk.org/hw/en_atxps.html

Paul
Thanks for all the info, Paul. I have some more for you; I hope it makes
sense.

I think you've already dug up my motherboard model number but here it is
again: ASRock N68C-GS FX.

I hadn't seen your reply yet - I don't have internet at my new place yet
so have to go to the library or a friend's place to go online - so I
went about the problem somewhat differently. I tried booting from my
Unbuntu 12.04 LTS CD but had the same issues as when I booted from XP on
the hard drive. I also tried booting from an old Knoppix CD that I had
lying around; same issues again.

The monitor is a Samsung T260 and it is clearly getting power since it
cycles through the HDMI/Digital/Analog options. (I'm not sure how to put
this correctly but when I power up the computer and if the monitor is
on, it displays HDMI in the upper left corner for several seconds, then
Digital, then Analog. I assume this is because there are connections for
HDMI, DVI and RGB at the back and it is not sure where the signal will
come from so it checks all three.) Anyway, I noticed that after it
cycles through all three options two or three times, it displays an
error message: Check signal cable.

That made me believe the cable was faulty or improperly seated so I had
a look at it. It seemed to be seated just fine. There was an RGB male to
RGB female cable running out of the monitor; then, there was an RGB male
to RGB male connected to the end of that, serving to extend the cable;
then the RGB male end went into the back of the tower. But everything
seemed snug so I wasn't sure what to do next. I unplugged the RGB cable
where it went into the monitor and tried reseat it. It was a bit of a
struggle to get the orientation right so I looked at the end of the
cable and saw that two of the middle row of pins looked wrong: one of
the pins seemed to be missing entirely, and the other seemed to be
pushed halfway back into the plug. That looked like exactly the kind of
thing that could cause my problem so I removed that cable entirely and
just ran the RGB male to RGB male cable directly from the monitor to the
tower; all of its pins on both ends seemed fine. But I still get the
"check signal cable" message from the monitor so that obviously DIDN'T
solve the problem. So I'm back to being confused about what is wrong.

For what it's worth, I tried connecting the RGB cable from my monitor to
my laptop and then used F8 to cycle between the various options like "PC
Screen only", "Extend", "Duplicate" and "Second screen only". This made
it clear that the monitor has no problem displaying the desktop of the
laptop so the monitor itself (and apparently its RGB port) would appear
to be fine.

Could the RBG port on the mobo be failing? Maybe I just need a "gender
bender" to convert the monitor cable to female at the tower end so that
I can use the male RGB port which is right above the female RGB port.

I'm guessing that there's nothing else wrong. Does that make any sense?
Like I said, I'm lousy at this hardware stuff ;-)

Unfortunately, I had to put my voltmeter in storage (out of town) and
don't have easy access to another one. I also downsized dramatically
during my move and got rid of various old monitors, cables, and such
things that would come in really handy right now for diagnostic
purposes. That, and the lack of internet access at home, make this
problem a lot harder to solve than it otherwise would be....

I'm not sure if I'll see your reply to this before I go offline again
for the day so, if I don't, I'll try the other stuff you suggested when
I get home (or at least as much as I can without a voltmeter). But I'm
thinking the "check signal cable" message is the Big Clue we need to
sort this out. I wish I had thought to mention that when I first posted
but I didn't see this symptom until today....

But the PC beep is all important.

You said the fans are spinning, but there is no beep.
Fans means the power supply is delivering some +12V
on the main cable. We know that much.

If the computer case speaker doesn't beep once at powerup,
then the CPU probably isn't executing the BIOS firmware.
It's either crashed, or it's not getting power (ATX12V
2x2 connector).

If you are getting the standard single beep at power up,
we can move on to something else.

Based on "no beep", it's either motherboard, CPU, or power
supply.

Paul

On 2013-11-01 6:54 PM, Paul wrote:
Rhino wrote:
On 2013-10-31 8:36 PM, Paul wrote:
Rhino wrote:
I'm having trouble getting my desktop computer to boot up today.

I'd left it on continuously for two or three days and finally shut it
down when I went to bed last night. When I tried to boot it up this
morning, I got basically nothing.

The fans start immediately but I don't even get the single beep. I'm
not seeing a Windows logo or even the first screen which tells me what
do press to get into the BIOS. Pressing F2 does not get me to the BIOS
when I start pressing it repeatedly from the moment I powered up.

I thought it might simply be that the video cable had gotten dislodged
but I checked it where it connects to the monitor, where it goes into
its extension and where it goes into the case and all connections seem
tight. I also opened the case and looked for anything that might not
be seated properly, particularly hard drives, but everything seems to
be properly seated.

I'm not very good with this kind of hardware issue so I need some
advice on what to check next.

The desktop has been working fine since I reseated the SATA cables
(see my post in September for details). I'm not sure what caused this
latest problem. I have two cats and it's always possible they bumped
something as they walked around the computer but that's about the only
vague glimmer of an idea I have.



When you're having problems, the best direction is to
"simplify" the system and retest.

For example, you can fully power off the system, pull the
plug, then remove the DIMMs and store them in an antistatic
bag. If the system has a video card (non-integrated video),
you can remove the video card as well.

Now, power up the system, and listen for a repeating beep code.
That will be an error beep code. The BIOS firmware and the
processor running it, make that repeating code. That tells
you the processor is working. If you add the RAM back in,
the beep code pattern should change. Finally, adding in the
video card, should again remove all error reasons, and then you'll
be back to just the one beep at startup.

It's possible for the reset button to getting stuck. The
board stays in reset and gives your symptoms.

Check that the ATX12V cable is fully seated. A good
board isolates ATX12V from 12V on the main cable, and
the processor won't get any 12V if the ATX12V isn't
plugged in. The cables have latches to prevent them
from working loose.

Clearing CMOS can sometimes help, if something got
corrupted there. (Again, that's a procedure you
do with the power cord unplugged, to prevent damage.
BIOS settings must be reloaded later if you decide
to try that one, like enabling AHCI or whatever.)

If you have a multimeter, you can connect the black
lead of the multimeter to the chassis, then touch
the red probe to the top of the CMOS CR2032 coin cell
and check that it's around 3V or so. Some motherboards,
they refuse to start if the battery is flat. Since your
motherboard isn't that old, it would be hard to understand
the battery being flat already, but it's something to
check.

That's all that comes to mind right now. Remove stuff,
listen for a beep code. If you can't get a response
under any circumstances, next you'd do a check
on the power supply. With the multimeter, you'd check
for operating voltages on the thing. You'd also make
sure the fan is spinning on the supply (as proof
that the motherboard sent PS_ON# a logic low or
grounded zero volt signal). If the power supply
is running, there is a kind of "POWER_GOOD" signal
on the main cable, as well as all the main voltages
to check.

(Pinout and details, here.)

http://www.formfactors.org/developer...public_br2.pdf


So when it's simplified, you have power supply,
motherboard, processor. You can stick the meter on
the power supply, to eliminate that (all voltages within 5%,
the power status signal OK). And then that
leaves motherboard or processor. Processors hardly
ever fail (you can reseat the processor to verify
it's not a contact problem). That leaves bad motherboard.

I probe the back of the ATX main cable where the wires
go in, to get voltage readings. But you need a strong
light to be able to see that, and some motherboards
don't make access to the main cable, all that easy.
The nylon shell has just enough room
to put the voltage probe red lead down in the
shell, until it comes in contact with the metal
of the pin in there.

You check 3.3V, 5V, 12V, -12V, +5VSB are within
5% of nominal. Check PS_ON# is low (close to zero
volts). The power supply fan wouldn't run without
that being true. That's what turns on the supply.
And whatever signal indicates POWER_GOOD, that one
is probably a logic 1, high value, and would be
around 5V.

When I say 5% of nominal, on 12V that means the
voltage level will be between 11.4V and 12.6V. The
power supply isn't infinitely accurate or anything,
and will always be off a bit. The power supply
typically uses "bulk regulation" and on the primary
side there is only one control that says "make more
voltage or make less voltage". So the individual voltages
are only established by turns ratio, and are not
controlled individually as such. The most heavily
loaded rail ends up on the low side, because of
how the bulk regulation works. But it should
still stay within the 5% range.

There is a picture of a power supply schematic here,
if you like pretty pictures. While this won't
make any sense the first time you look at it,
it grows on you. Modern supplies have more
circuitry than this, and this much circuit
"just makes the voltages". Some will have
a bit more monitoring. And a few modern ones,
the architecture is entirely different (two
stage regulation).

http://www.pavouk.org/hw/en_atxps.html

Paul
Thanks for all the info, Paul. I have some more for you; I hope it
makes sense.

I think you've already dug up my motherboard model number but here it
is again: ASRock N68C-GS FX.

I hadn't seen your reply yet - I don't have internet at my new place
yet so have to go to the library or a friend's place to go online - so
I went about the problem somewhat differently. I tried booting from my
Unbuntu 12.04 LTS CD but had the same issues as when I booted from XP
on the hard drive. I also tried booting from an old Knoppix CD that I
had lying around; same issues again.

The monitor is a Samsung T260 and it is clearly getting power since it
cycles through the HDMI/Digital/Analog options. (I'm not sure how to
put this correctly but when I power up the computer and if the monitor
is on, it displays HDMI in the upper left corner for several seconds,
then Digital, then Analog. I assume this is because there are
connections for HDMI, DVI and RGB at the back and it is not sure where
the signal will come from so it checks all three.) Anyway, I noticed
that after it cycles through all three options two or three times, it
displays an error message: Check signal cable.

That made me believe the cable was faulty or improperly seated so I
had a look at it. It seemed to be seated just fine. There was an RGB
male to RGB female cable running out of the monitor; then, there was
an RGB male to RGB male connected to the end of that, serving to
extend the cable; then the RGB male end went into the back of the
tower. But everything seemed snug so I wasn't sure what to do next. I
unplugged the RGB cable where it went into the monitor and tried
reseat it. It was a bit of a struggle to get the orientation right so
I looked at the end of the cable and saw that two of the middle row of
pins looked wrong: one of the pins seemed to be missing entirely, and
the other seemed to be pushed halfway back into the plug. That looked
like exactly the kind of thing that could cause my problem so I
removed that cable entirely and just ran the RGB male to RGB male
cable directly from the monitor to the tower; all of its pins on both
ends seemed fine. But I still get the "check signal cable" message
from the monitor so that obviously DIDN'T solve the problem. So I'm
back to being confused about what is wrong.

For what it's worth, I tried connecting the RGB cable from my monitor
to my laptop and then used F8 to cycle between the various options
like "PC Screen only", "Extend", "Duplicate" and "Second screen only".
This made it clear that the monitor has no problem displaying the
desktop of the laptop so the monitor itself (and apparently its RGB
port) would appear to be fine.

Could the RBG port on the mobo be failing? Maybe I just need a "gender
bender" to convert the monitor cable to female at the tower end so
that I can use the male RGB port which is right above the female RGB
port.

I'm guessing that there's nothing else wrong. Does that make any
sense? Like I said, I'm lousy at this hardware stuff ;-)

Unfortunately, I had to put my voltmeter in storage (out of town) and
don't have easy access to another one. I also downsized dramatically
during my move and got rid of various old monitors, cables, and such
things that would come in really handy right now for diagnostic
purposes. That, and the lack of internet access at home, make this
problem a lot harder to solve than it otherwise would be....

I'm not sure if I'll see your reply to this before I go offline again
for the day so, if I don't, I'll try the other stuff you suggested
when I get home (or at least as much as I can without a voltmeter).
But I'm thinking the "check signal cable" message is the Big Clue we
need to sort this out. I wish I had thought to mention that when I
first posted but I didn't see this symptom until today....

But the PC beep is all important.

You said the fans are spinning, but there is no beep.
Fans means the power supply is delivering some +12V
on the main cable. We know that much.

If the computer case speaker doesn't beep once at powerup,
then the CPU probably isn't executing the BIOS firmware.
It's either crashed, or it's not getting power (ATX12V
2x2 connector).

If you are getting the standard single beep at power up,
we can move on to something else.

Based on "no beep", it's either motherboard, CPU, or power
supply.



Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money is
NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....

--
Rhino

Rhino wrote:


Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money is
NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....


Turn off the power, pull the DIMMs, put them in an
antistatic bag so they don't get damaged. Turn on
the machine, and listen for beeps. If there is no
RAM, the processor can still run, and the code
it runs controls the beep pattern. Hearing it
beep in such a situation, is proof the processor
still works.

If that processor did not have the square 2x2 ATX12V
connected, then it could not beep. Your system was
working, and only if somehow that connector worked
itself loose, would that theory be practical.

Checking the power supply is going to be a more
"techy" procedure. If you have a multimeter, you can
check the voltages. Otherwise, you can replace the
supply (if you think that is cheaper than going to
a shop and paying a diagnosis fee). And there's still
no guarantee it's not the motherboard.

Is the power supply brand new ? Or was it
moved from an older system ? Does the supply
have a history of problems ?

My worst experience here, with diagnosing things,
is I ended up buying duplicates of practically
everything in the computer. And then the
"home repair" idea isn't as attractive. If you
get lucky on the "swap and test" thing, sometimes
you come out ahead.

When my RAM suddenly started throwing errors here,
just on a whim I adjusted Vnb (Northbridge) by
a step on the controls, and the errors stopped.
And that saved me a few dollars. I've had a fair
number of sticks of RAM that outright failed,
and when it happens, it's in the 1.5 to 2 year
timeframe. The RAM in that case, was "generic"
and didn't have a famous brand name on the sticks.
I bought the RAM locally. It "seemed like a deal",
and now, I don't buy crap like that any more.
My trick now, is to read the reviews for products,
before I buy them.

If RAM fails, and low memory locations cannot
hold valid contents, that sometimes makes a
motherboard "fail silently". My Nforce2 board
did that one day. A stick of Ballistix RAM
decided to have one of the chips completely
die on it, and the motherboard couldn't even
beep.

But that probably doesn't happen all that
often. When sticks of RAM get errors, usually
it's just a few bits on the DIMM with stuck-at
failures. Entire chips croaking, isn't as common.
And I didn't even overvolt those DIMMs. It
wasn't a "stress failure".

Pulling the DIMMs, I could get a "beep" from
the motherboard, so I knew my motherboard,
power supply and processor were OK. Since I'd
just pulled the RAM, I then knew there was somethine
seriously wrong with the RAM. Further testing
(in single channel mode), revealed the dead chip.

Paul

On 2013-11-02 5:20 PM, Paul wrote:
Rhino wrote:


Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money
is NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....


Turn off the power, pull the DIMMs, put them in an
antistatic bag so they don't get damaged. Turn on
the machine, and listen for beeps. If there is no
RAM, the processor can still run, and the code
it runs controls the beep pattern. Hearing it
beep in such a situation, is proof the processor
still works.

If that processor did not have the square 2x2 ATX12V
connected, then it could not beep. Your system was
working, and only if somehow that connector worked
itself loose, would that theory be practical.

Checking the power supply is going to be a more
"techy" procedure. If you have a multimeter, you can
check the voltages. Otherwise, you can replace the
supply (if you think that is cheaper than going to
a shop and paying a diagnosis fee). And there's still
no guarantee it's not the motherboard.

Is the power supply brand new ? Or was it
moved from an older system ? Does the supply
have a history of problems ?

My worst experience here, with diagnosing things,
is I ended up buying duplicates of practically
everything in the computer. And then the
"home repair" idea isn't as attractive. If you
get lucky on the "swap and test" thing, sometimes
you come out ahead.

When my RAM suddenly started throwing errors here,
just on a whim I adjusted Vnb (Northbridge) by
a step on the controls, and the errors stopped.
And that saved me a few dollars. I've had a fair
number of sticks of RAM that outright failed,
and when it happens, it's in the 1.5 to 2 year
timeframe. The RAM in that case, was "generic"
and didn't have a famous brand name on the sticks.
I bought the RAM locally. It "seemed like a deal",
and now, I don't buy crap like that any more.
My trick now, is to read the reviews for products,
before I buy them.

If RAM fails, and low memory locations cannot
hold valid contents, that sometimes makes a
motherboard "fail silently". My Nforce2 board
did that one day. A stick of Ballistix RAM
decided to have one of the chips completely
die on it, and the motherboard couldn't even
beep.

But that probably doesn't happen all that
often. When sticks of RAM get errors, usually
it's just a few bits on the DIMM with stuck-at
failures. Entire chips croaking, isn't as common.
And I didn't even overvolt those DIMMs. It
wasn't a "stress failure".

Pulling the DIMMs, I could get a "beep" from
the motherboard, so I knew my motherboard,
power supply and processor were OK. Since I'd
just pulled the RAM, I then knew there was somethine
seriously wrong with the RAM. Further testing
(in single channel mode), revealed the dead chip.



Thanks, Paul. I will try what you've suggested. In fact, I had planned
to try it this past weekend but forgot to save your suggestions to a
file before going offline. But I'm going to save them now and then try
it when I get home. I can certainly remove the RAM and see if I get a
beep. Let's see what that reveals and go from there....




--
Rhino

On 04/11/2013 14:47, Rhino wrote:
On 2013-11-02 5:20 PM, Paul wrote:
Rhino wrote:


Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money
is NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....


Turn off the power, pull the DIMMs, put them in an
antistatic bag so they don't get damaged. Turn on
the machine, and listen for beeps. If there is no
RAM, the processor can still run, and the code
it runs controls the beep pattern. Hearing it
beep in such a situation, is proof the processor
still works.

If that processor did not have the square 2x2 ATX12V
connected, then it could not beep. Your system was
working, and only if somehow that connector worked
itself loose, would that theory be practical.

Checking the power supply is going to be a more
"techy" procedure. If you have a multimeter, you can
check the voltages. Otherwise, you can replace the
supply (if you think that is cheaper than going to
a shop and paying a diagnosis fee). And there's still
no guarantee it's not the motherboard.

Is the power supply brand new ? Or was it
moved from an older system ? Does the supply
have a history of problems ?

My worst experience here, with diagnosing things,
is I ended up buying duplicates of practically
everything in the computer. And then the
"home repair" idea isn't as attractive. If you
get lucky on the "swap and test" thing, sometimes
you come out ahead.

When my RAM suddenly started throwing errors here,
just on a whim I adjusted Vnb (Northbridge) by
a step on the controls, and the errors stopped.
And that saved me a few dollars. I've had a fair
number of sticks of RAM that outright failed,
and when it happens, it's in the 1.5 to 2 year
timeframe. The RAM in that case, was "generic"
and didn't have a famous brand name on the sticks.
I bought the RAM locally. It "seemed like a deal",
and now, I don't buy crap like that any more.
My trick now, is to read the reviews for products,
before I buy them.

If RAM fails, and low memory locations cannot
hold valid contents, that sometimes makes a
motherboard "fail silently". My Nforce2 board
did that one day. A stick of Ballistix RAM
decided to have one of the chips completely
die on it, and the motherboard couldn't even
beep.

But that probably doesn't happen all that
often. When sticks of RAM get errors, usually
it's just a few bits on the DIMM with stuck-at
failures. Entire chips croaking, isn't as common.
And I didn't even overvolt those DIMMs. It
wasn't a "stress failure".

Pulling the DIMMs, I could get a "beep" from
the motherboard, so I knew my motherboard,
power supply and processor were OK. Since I'd
just pulled the RAM, I then knew there was somethine
seriously wrong with the RAM. Further testing
(in single channel mode), revealed the dead chip.



Thanks, Paul. I will try what you've suggested. In fact, I had planned
to try it this past weekend but forgot to save your suggestions to a
file before going offline. But I'm going to save them now and then try
it when I get home. I can certainly remove the RAM and see if I get a
beep. Let's see what that reveals and go from there....

If still no beeps and your system has a separate graphics card,
remove the graphics card and try powering up again.
If you now get beeps, it's a dead graphics card (although
sometimes removing and re-inserting such cards can 'fix'
the problem.)
If the graphics are on the motherboard, this is of no help of
course.
HTH
--
Rob

On 2013-11-02 5:20 PM, Paul wrote:
Rhino wrote:


Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money
is NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....


Turn off the power, pull the DIMMs, put them in an
antistatic bag so they don't get damaged. Turn on
the machine, and listen for beeps. If there is no
RAM, the processor can still run, and the code
it runs controls the beep pattern. Hearing it
beep in such a situation, is proof the processor
still works.

If that processor did not have the square 2x2 ATX12V
connected, then it could not beep. Your system was
working, and only if somehow that connector worked
itself loose, would that theory be practical.

Checking the power supply is going to be a more
"techy" procedure. If you have a multimeter, you can
check the voltages. Otherwise, you can replace the
supply (if you think that is cheaper than going to
a shop and paying a diagnosis fee). And there's still
no guarantee it's not the motherboard.

Is the power supply brand new ? Or was it
moved from an older system ? Does the supply
have a history of problems ?

My worst experience here, with diagnosing things,
is I ended up buying duplicates of practically
everything in the computer. And then the
"home repair" idea isn't as attractive. If you
get lucky on the "swap and test" thing, sometimes
you come out ahead.

When my RAM suddenly started throwing errors here,
just on a whim I adjusted Vnb (Northbridge) by
a step on the controls, and the errors stopped.
And that saved me a few dollars. I've had a fair
number of sticks of RAM that outright failed,
and when it happens, it's in the 1.5 to 2 year
timeframe. The RAM in that case, was "generic"
and didn't have a famous brand name on the sticks.
I bought the RAM locally. It "seemed like a deal",
and now, I don't buy crap like that any more.
My trick now, is to read the reviews for products,
before I buy them.

If RAM fails, and low memory locations cannot
hold valid contents, that sometimes makes a
motherboard "fail silently". My Nforce2 board
did that one day. A stick of Ballistix RAM
decided to have one of the chips completely
die on it, and the motherboard couldn't even
beep.

But that probably doesn't happen all that
often. When sticks of RAM get errors, usually
it's just a few bits on the DIMM with stuck-at
failures. Entire chips croaking, isn't as common.
And I didn't even overvolt those DIMMs. It
wasn't a "stress failure".

Pulling the DIMMs, I could get a "beep" from
the motherboard, so I knew my motherboard,
power supply and processor were OK. Since I'd
just pulled the RAM, I then knew there was somethine
seriously wrong with the RAM. Further testing
(in single channel mode), revealed the dead chip.



Sorry for the delay in following up; I got sidetracked by other matters.

I took the DIMMs out this morning (with the power disconnected) and then
powered up: I got three long beeps, repeated several times with a short
pause between each group of three. I powered off, put the DIMMs back in
and powered up again, expecting to see the same issue: no beeps at all
and reboot not proceeding. But instead, I got a single short beep, some
pops from the speakers, and then Ubuntu booted up (I still had the
Ubuntu 12.04 CD in the tray). Ubuntu behaved normally so I shut down and
removed the CD and rebooted to XP. XP also came up fine. I'm a little
puzzled about why removing the DIMMs and putting them back in fixed
anything but I'm not inclined to look a gift horse in the mouth ;-)

The thing that bothers me is that I get two or three loud pops when I
boot up (either Ubuntu or XP); more importantly, the audio is distorted
when I play an audio file in XP (didn't try it in Ubuntu) it is pretty
badly distorted. I'm not sure what to make of that. Audio was fine
before this incident began. I thought that I was past my audio problems
with the USB gizmo. I'd appreciate any tips on how to get the sound
working better....


--
Rhino

Rhino wrote:
On 2013-11-02 5:20 PM, Paul wrote:
Rhino wrote:


Darn. I was hoping it was something as simple as a bad cable. This
sounds like it is going to be expensive to diagnose and fix and money
is NOT something I have on hand these days.

I'll figure out what I can on my own (and with your help)....


Turn off the power, pull the DIMMs, put them in an
antistatic bag so they don't get damaged. Turn on
the machine, and listen for beeps. If there is no
RAM, the processor can still run, and the code
it runs controls the beep pattern. Hearing it
beep in such a situation, is proof the processor
still works.

If that processor did not have the square 2x2 ATX12V
connected, then it could not beep. Your system was
working, and only if somehow that connector worked
itself loose, would that theory be practical.

Checking the power supply is going to be a more
"techy" procedure. If you have a multimeter, you can
check the voltages. Otherwise, you can replace the
supply (if you think that is cheaper than going to
a shop and paying a diagnosis fee). And there's still
no guarantee it's not the motherboard.

Is the power supply brand new ? Or was it
moved from an older system ? Does the supply
have a history of problems ?

My worst experience here, with diagnosing things,
is I ended up buying duplicates of practically
everything in the computer. And then the
"home repair" idea isn't as attractive. If you
get lucky on the "swap and test" thing, sometimes
you come out ahead.

When my RAM suddenly started throwing errors here,
just on a whim I adjusted Vnb (Northbridge) by
a step on the controls, and the errors stopped.
And that saved me a few dollars. I've had a fair
number of sticks of RAM that outright failed,
and when it happens, it's in the 1.5 to 2 year
timeframe. The RAM in that case, was "generic"
and didn't have a famous brand name on the sticks.
I bought the RAM locally. It "seemed like a deal",
and now, I don't buy crap like that any more.
My trick now, is to read the reviews for products,
before I buy them.

If RAM fails, and low memory locations cannot
hold valid contents, that sometimes makes a
motherboard "fail silently". My Nforce2 board
did that one day. A stick of Ballistix RAM
decided to have one of the chips completely
die on it, and the motherboard couldn't even
beep.

But that probably doesn't happen all that
often. When sticks of RAM get errors, usually
it's just a few bits on the DIMM with stuck-at
failures. Entire chips croaking, isn't as common.
And I didn't even overvolt those DIMMs. It
wasn't a "stress failure".

Pulling the DIMMs, I could get a "beep" from
the motherboard, so I knew my motherboard,
power supply and processor were OK. Since I'd
just pulled the RAM, I then knew there was somethine
seriously wrong with the RAM. Further testing
(in single channel mode), revealed the dead chip.



Sorry for the delay in following up; I got sidetracked by other matters.

I took the DIMMs out this morning (with the power disconnected) and then
powered up: I got three long beeps, repeated several times with a short
pause between each group of three. I powered off, put the DIMMs back in
and powered up again, expecting to see the same issue: no beeps at all
and reboot not proceeding. But instead, I got a single short beep, some
pops from the speakers, and then Ubuntu booted up (I still had the
Ubuntu 12.04 CD in the tray). Ubuntu behaved normally so I shut down and
removed the CD and rebooted to XP. XP also came up fine. I'm a little
puzzled about why removing the DIMMs and putting them back in fixed
anything but I'm not inclined to look a gift horse in the mouth ;-)

The thing that bothers me is that I get two or three loud pops when I
boot up (either Ubuntu or XP); more importantly, the audio is distorted
when I play an audio file in XP (didn't try it in Ubuntu) it is pretty
badly distorted. I'm not sure what to make of that. Audio was fine
before this incident began. I thought that I was past my audio problems
with the USB gizmo. I'd appreciate any tips on how to get the sound
working better....



Sounds like the DIMM wasn't making good contact.

This is actually one of the reasons for pulling the DIMM :-)
You get the wiping action of a fresh insertion.

*******

With regard to a USB audio solution, pops could happen
if there are transients on +5VSB. Either the power supply
is not completely happy on the +5VSB output rail, or some
hardware in your computer is placing a significant load
on +5VSB.

At one time, the user could control the source of the
USB bus voltage, by means of motherboard jumpers.
On an old system, you could change power sources,
and do that change on a "USB stack" basis. Two USB
ports would share a single jumper. Some motherboards
would have four or five jumpers to play with.

You'd leave the USB keyboard stack set to +5VSB (so you
can do "wake via keyboard). And leave any other ports
powered by the +5VSB rail.

*******

OK. Good news. Your motherboard has two 1x3 power
headers with a jumper on each. So as a user, you
do have some control over the USB Vbus source.

PS2_USB_PWR1 PS/2 stack, as well as USB0,USB1,USB2,USB3 ports

USB_PWR2 USB4 .. USB9 (six ports total)

In the interests of science, turn off the computer,
unplug it, wait 30 seconds, move both jumpers to
the 1_2 position for +5V operation. The +5V rail
in your computer, is a lot stronger than the +5VSB
rail.

Power up, listen for pops. Tell us whether it's fixed.

If it is fixed, the issue could be too much current flow
through the USB_PWR type headers. Asrock loves to save
a few cents, by running too many ports through a USB_PWR jumper.
On an older Asus motherboard, it was two ports only, per
each 1x3 header.

The action of moving the jumper, will help from a second
perspective. It'll be a fresh wiping action, in case
the jumper wasn't making good contact.

One of the 1x3 headers is on the upper left hand corner of
the motherboard. The other 1x3 header is on the lower right.
Each one shares one inrush capacitor, over six or so ports.
Again, not that wonderful. The capacitor near a USB connector,
is to "hold up" or maintain the Vbus voltage, when a new
USB device is plugged in. USB devices are hard on the
Vbus voltage, as any bypass capacitor inside a USB peripheral,
draws an inrush of current. The capacitor on the motherboard,
near the header, is there to battle with that.

If you leave both jumpers in the 1_2 position, the only
side effect would be losing "wake on keyboard" capability.
Just as an example of why the ports were running from
+5VSB in the first place.

Paul