PC Won't Start


B

barrowhill

Having switched of in normal fashion, son's PC will now not start. No
activity what so ever.

Motherboard is Foxconn 945G7MD/945P7MD with 3GB RAM, 160GBHDD, ATI Radeon
4850 Graphics Card and twin DVD RW.

When case is opened the ATX power LED situated in front of 20+4 pin molex
connector is flashing.

The PSU has been tested and is OK (pins 14 & 15 on 20pin molex shorted).

I've removed grahics card and disconnected DVD's etc. still same.

Currently I've got connected 1GB RAM, 8pin ATX 12v power connecter and 4pin
(3/5/12v), all fans (Case & CPU) but have disconnected 20pin ATX power and
shorted pins 14 & 15. PSU runs when connected to supply

CPU fan and case fan running and mobo Power LED on constant.

If I connect 20 pin to motherboard, no activity, LED flashing.

Can't see any signs of burning etc on mobo.

Is there anything else I should try or look at before concluding mobo faulty
?????
 
Ad

Advertisements

S

SC Tom

barrowhill said:
Having switched of in normal fashion, son's PC will now not start. No
activity what so ever.

Motherboard is Foxconn 945G7MD/945P7MD with 3GB RAM, 160GBHDD, ATI Radeon
4850 Graphics Card and twin DVD RW.

When case is opened the ATX power LED situated in front of 20+4 pin molex
connector is flashing.

The PSU has been tested and is OK (pins 14 & 15 on 20pin molex shorted).

I've removed grahics card and disconnected DVD's etc. still same.

Currently I've got connected 1GB RAM, 8pin ATX 12v power connecter and
4pin (3/5/12v), all fans (Case & CPU) but have disconnected 20pin ATX
power and shorted pins 14 & 15. PSU runs when connected to supply

CPU fan and case fan running and mobo Power LED on constant.

If I connect 20 pin to motherboard, no activity, LED flashing.

Can't see any signs of burning etc on mobo.

Is there anything else I should try or look at before concluding mobo
faulty ?????
If the LED is solid, then the PS is more than likely good. If the LED is
flashing, then the PS is probably bad. I would go with that first. If you
have access to a known good one, put in play. Power supplies are not meant
to be fixed, although with the proper test equipment, I guess it could be
done. But with the cost of them being fairly low, and the possibility of a
bad PS taking out the MB when it goes, I'd just as soon replace it as try to
repair it.
 
P

Paul

barrowhill said:
Having switched of in normal fashion, son's PC will now not start. No
activity what so ever.

Motherboard is Foxconn 945G7MD/945P7MD with 3GB RAM, 160GBHDD, ATI Radeon
4850 Graphics Card and twin DVD RW.

When case is opened the ATX power LED situated in front of 20+4 pin molex
connector is flashing.

The PSU has been tested and is OK (pins 14 & 15 on 20pin molex shorted).

I've removed grahics card and disconnected DVD's etc. still same.

Currently I've got connected 1GB RAM, 8pin ATX 12v power connecter and 4pin
(3/5/12v), all fans (Case & CPU) but have disconnected 20pin ATX power and
shorted pins 14 & 15. PSU runs when connected to supply

CPU fan and case fan running and mobo Power LED on constant.

If I connect 20 pin to motherboard, no activity, LED flashing.

Can't see any signs of burning etc on mobo.

Is there anything else I should try or look at before concluding mobo faulty
?????

I'm most interested in:

"the ATX power LED situated in front of 20+4 pin molex connector is flashing"

On an Asus brand motherboard, the single LED on the motherboard is tied
directly to the +5VSB rail (it's not switched with a transistor or
anything). And as such, it monitors the voltage coming from the power supply.
Such a LED is not allowed to blink. It should light up, as soon as the switch
on the back of the ATX supply is switched on. The +5VSB supply, is used for things
like powering RAM contents, while the computer sleeps. On an Asus board,
that LED basically tells you not to pull any RAM sticks, because
the system is still powered. Once the supply is powered off, the
LED should die about 30 seconds later, indicating it is safe to
add or remove components from the computer. It's a safety LED, to
help avoid damage to components.

I'll assume the Foxconn implementation is similar, although I don't
remember too many other brands doing that. Asus is pretty consistent
in putting that LED on their boards, and I think it's worth an
extra 10 cents, to have it there.

That +5VSB rail on the ATX supply, has limited current output. Depending on
supply, it delivers between 2 and 3 amps, on typical supplies. The
motherboard might need about 1 amp, so there should be some margin.

The ATX power supply fan does not spin while the computer is sleeping.
A little bit of heat is dissipated by the supply, while it is making
the +5VSB, but not enough heat to require the fan.

Now, on modern motherboards, the clever designers have decided it would be
fun, to power all the USB ports off +5VSB. That allows a USB keyboard,
to be used to wake the computer. But a downside of doing that, is any
"dumb" electrical loads on USB, can overpower the supply. In marginal
cases, the +5VSB overheats and shuts off. When it cools, it can come
back on. And that can lead to a "blinking" monitor LED.

An alternative theory, is the +5VSB supply inside the ATX supply, is
weak. The motherboard loading of 1 ampere, isn't excessive, but if
the supply is weak, you can get the "blinking LED" problem.

For this fine theory to be true, you'd have to know whether that
is the function of the Foxconn LED you're looking at. I can't find mention
of a LED in the manual. Can you provide a URL to the manual
you're using, and the page that shows the LED. While I have a
(crappy) picture of the motherboard, the resolution isn't fine
enough to locate the LED you describe.

In any case, you can try disconnecting all the USB loads on the
computer, and see if the blinking stops. Or, try another
power supply, and see if the +5VSB holds up better on the
alternate supply. Yes, the motherboard could easily overload
that rail. It wouldn't take much of a fault to do it.

My bet is, this is a weak supply problem.

Paul
 
B

barrowhill

Paul,

I believe LED operation same as ASUS. All testing has been done without any
load. No USB inc KB and Mouse, no DVD's, no Monitor and no RAM. I have
replaced PSU with another but with 20 pin molex connected to mobo nothing
apparent happens; no fans running inc PSU

Both PSU's tested using standard test method - Pin 14/Green [Power Supply
On] connected to ground/Black without any load at all. All power
connectors - +3.3v, +/-5v, +/- 12v - when measured for output are OK.

Mobo Manual available here

http://www.foxconnsupport.com/drive...2H/manual/English/945P(G)7MD-En-V1.0-1010.pdf

but mobo layout not detailed enough to show LED or give any explanation

Regards
 
P

Paul

barrowhill said:
Paul,

I believe LED operation same as ASUS. All testing has been done without any
load. No USB inc KB and Mouse, no DVD's, no Monitor and no RAM. I have
replaced PSU with another but with 20 pin molex connected to mobo nothing
apparent happens; no fans running inc PSU

Both PSU's tested using standard test method - Pin 14/Green [Power Supply
On] connected to ground/Black without any load at all. All power
connectors - +3.3v, +/-5v, +/- 12v - when measured for output are OK.

Mobo Manual available here

http://www.foxconnsupport.com/drive...2H/manual/English/945P(G)7MD-En-V1.0-1010.pdf

but mobo layout not detailed enough to show LED or give any explanation

Regards

So for the second supply, the motherboard isn't able to operate PS_ON# ?

On the first supply, perhaps the PS_ON# signal was being pulled to a voltage
in the middle of the range.

The PS_ON# signal, from the motherboard to the power supply, is open collector.
To turn on the power supply, the motherboard pulls the PS_ON# signal to ground.
Ideally, somewhere in the 0.4V to 0.8V region, is where you'd expect to
see the signal level, when the motherboard turns on the supply.

When the computing session is finished, the motherboard releases the PS_ON# signal,
and a pullup returns it to a 5V level. The valid signal range, to turn off the
supply, would be any level above about 2.4V or so.

Voltages between those two values, are undefined. At around 1.3V, for example,
the power supply will be confused. It might come on, it might not. Perhaps it
would even oscillate (go on then off).

You described "jamming" the supply on, by grounding PS_ON#. Perhaps you can
ground PS_ON# while the 20 pin is connected to the motherboard ? As long
as the motherboard is using open collector drive, it is safe to ground
PS_ON#. To ground PS_ON#, you can access exposed metal, where the wires
enter the nylon shell of the main connector.

If that works, and the computer starts, it would tell you that the
motherboard can no longer drive PS_ON# properly for some reason. Operating
the computer with a jumper installed on PS_ON#, is not practical. It
also has some danger involved - if you jumper PS_ON# to be permanently
ON, then if the processor overheats, the motherboard cannot protect
itself by turning the power off. So using the jumper, also disables
the THERMTRIP protection feature. But for a quick test, it is probably OK.

If that test works, then you'd have to figure out why the open collector
driver on the motherboard, isn't working any more.

Ten years ago, the method used would look like the circuit on page 18
in the upper right hand corner. The 74F07 is an open collector
driver, and it is running from +5VSB. Pin 2 is the output of the
driver. If this motherboard broke, you'd replace IC U36 and put
a new one in. Modern motherboards will be harder to fix - the
drive could actually be coming from the chipset (Southbridge) itself.
The 74F07 is pretty powerful, and not as likely to break. Using
the 74F07 is termed "buffering" or improving the drive strength.

http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF

The 74F07 here, lists an IOL max of 64 milliamps, which is much
higher than the few milliamps needed to switch on the ATX supply.
So that reference schematic, shouldn't suffer from any PS_ON#
problems.

http://www.nxp.com/documents/data_sheet/74F07.pdf

Paul
 
Ad

Advertisements

B

barrowhill

Paul,

Both PSU's when connected to mobo do not respond to PS ON. Both PSU's test
OK when PS ON is "forced" which leads me to the conclusion that the problem
lies with the mobo.

As I understand it PS ON is pulled up to +5 by the PSU and must be driven
low to turn PSU on; hence, when I test PSU operation by connecting pin 13
(PS ON) to common rail the PSU turns ON. What I'd liked to understand is
what conditions must exist for PS ON to be driven low (and hence turn on
PSU) when 20 pin molex is connected to mobo

Your suggested test, whilst impractical for long term solution, would
probably confirm mobo component fault. If so, replacement mobo will be
required. I'll probably give this test a go.

Regards


Paul said:
barrowhill said:
Paul,

I believe LED operation same as ASUS. All testing has been done without
any
load. No USB inc KB and Mouse, no DVD's, no Monitor and no RAM. I have
replaced PSU with another but with 20 pin molex connected to mobo nothing
apparent happens; no fans running inc PSU

Both PSU's tested using standard test method - Pin 14/Green [Power Supply
On] connected to ground/Black without any load at all. All power
connectors - +3.3v, +/-5v, +/- 12v - when measured for output are OK.

Mobo Manual available here

http://www.foxconnsupport.com/drive...2H/manual/English/945P(G)7MD-En-V1.0-1010.pdf

but mobo layout not detailed enough to show LED or give any explanation

Regards

So for the second supply, the motherboard isn't able to operate PS_ON# ?

On the first supply, perhaps the PS_ON# signal was being pulled to a
voltage
in the middle of the range.

The PS_ON# signal, from the motherboard to the power supply, is open
collector.
To turn on the power supply, the motherboard pulls the PS_ON# signal to
ground.
Ideally, somewhere in the 0.4V to 0.8V region, is where you'd expect to
see the signal level, when the motherboard turns on the supply.

When the computing session is finished, the motherboard releases the
PS_ON# signal,
and a pullup returns it to a 5V level. The valid signal range, to turn off
the
supply, would be any level above about 2.4V or so.

Voltages between those two values, are undefined. At around 1.3V, for
example,
the power supply will be confused. It might come on, it might not. Perhaps
it
would even oscillate (go on then off).

You described "jamming" the supply on, by grounding PS_ON#. Perhaps you
can
ground PS_ON# while the 20 pin is connected to the motherboard ? As long
as the motherboard is using open collector drive, it is safe to ground
PS_ON#. To ground PS_ON#, you can access exposed metal, where the wires
enter the nylon shell of the main connector.

If that works, and the computer starts, it would tell you that the
motherboard can no longer drive PS_ON# properly for some reason. Operating
the computer with a jumper installed on PS_ON#, is not practical. It
also has some danger involved - if you jumper PS_ON# to be permanently
ON, then if the processor overheats, the motherboard cannot protect
itself by turning the power off. So using the jumper, also disables
the THERMTRIP protection feature. But for a quick test, it is probably OK.

If that test works, then you'd have to figure out why the open collector
driver on the motherboard, isn't working any more.

Ten years ago, the method used would look like the circuit on page 18
in the upper right hand corner. The 74F07 is an open collector
driver, and it is running from +5VSB. Pin 2 is the output of the
driver. If this motherboard broke, you'd replace IC U36 and put
a new one in. Modern motherboards will be harder to fix - the
drive could actually be coming from the chipset (Southbridge) itself.
The 74F07 is pretty powerful, and not as likely to break. Using
the 74F07 is termed "buffering" or improving the drive strength.

http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF

The 74F07 here, lists an IOL max of 64 milliamps, which is much
higher than the few milliamps needed to switch on the ATX supply.
So that reference schematic, shouldn't suffer from any PS_ON#
problems.

http://www.nxp.com/documents/data_sheet/74F07.pdf

Paul
 
Ad

Advertisements

B

barrowhill

Paul,

Just to update............Up and working !!!

Took for granted that PS ON signal pulled to ground via operation of On/OFF
switch. Thought I check this out and test operation and connected cabling.

Also as I'd got mobo out on desktop tool CPU fan/heatsink off to clean
(really dirty). Took out chip at same time and surprised to see quite a bit
of "fluffy" dust particles between chip and mobo. Cleaned out. Noticed
also that there was very little thermal paste applied so purchased some and
applied before putting back.

When powered up; lo and behold, kicks into life..........mobo LED
permanently on prior to pushing on/off switch and goes off when on/off
switch pressed and PSU kicks into life.

Not sure what made it work or what really was wrong but thanks for your help

Regards

Paul said:
barrowhill said:
Paul,

I believe LED operation same as ASUS. All testing has been done without
any
load. No USB inc KB and Mouse, no DVD's, no Monitor and no RAM. I have
replaced PSU with another but with 20 pin molex connected to mobo nothing
apparent happens; no fans running inc PSU

Both PSU's tested using standard test method - Pin 14/Green [Power Supply
On] connected to ground/Black without any load at all. All power
connectors - +3.3v, +/-5v, +/- 12v - when measured for output are OK.

Mobo Manual available here

http://www.foxconnsupport.com/drive...2H/manual/English/945P(G)7MD-En-V1.0-1010.pdf

but mobo layout not detailed enough to show LED or give any explanation

Regards

So for the second supply, the motherboard isn't able to operate PS_ON# ?

On the first supply, perhaps the PS_ON# signal was being pulled to a
voltage
in the middle of the range.

The PS_ON# signal, from the motherboard to the power supply, is open
collector.
To turn on the power supply, the motherboard pulls the PS_ON# signal to
ground.
Ideally, somewhere in the 0.4V to 0.8V region, is where you'd expect to
see the signal level, when the motherboard turns on the supply.

When the computing session is finished, the motherboard releases the
PS_ON# signal,
and a pullup returns it to a 5V level. The valid signal range, to turn off
the
supply, would be any level above about 2.4V or so.

Voltages between those two values, are undefined. At around 1.3V, for
example,
the power supply will be confused. It might come on, it might not. Perhaps
it
would even oscillate (go on then off).

You described "jamming" the supply on, by grounding PS_ON#. Perhaps you
can
ground PS_ON# while the 20 pin is connected to the motherboard ? As long
as the motherboard is using open collector drive, it is safe to ground
PS_ON#. To ground PS_ON#, you can access exposed metal, where the wires
enter the nylon shell of the main connector.

If that works, and the computer starts, it would tell you that the
motherboard can no longer drive PS_ON# properly for some reason. Operating
the computer with a jumper installed on PS_ON#, is not practical. It
also has some danger involved - if you jumper PS_ON# to be permanently
ON, then if the processor overheats, the motherboard cannot protect
itself by turning the power off. So using the jumper, also disables
the THERMTRIP protection feature. But for a quick test, it is probably OK.

If that test works, then you'd have to figure out why the open collector
driver on the motherboard, isn't working any more.

Ten years ago, the method used would look like the circuit on page 18
in the upper right hand corner. The 74F07 is an open collector
driver, and it is running from +5VSB. Pin 2 is the output of the
driver. If this motherboard broke, you'd replace IC U36 and put
a new one in. Modern motherboards will be harder to fix - the
drive could actually be coming from the chipset (Southbridge) itself.
The 74F07 is pretty powerful, and not as likely to break. Using
the 74F07 is termed "buffering" or improving the drive strength.

http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF

The 74F07 here, lists an IOL max of 64 milliamps, which is much
higher than the few milliamps needed to switch on the ATX supply.
So that reference schematic, shouldn't suffer from any PS_ON#
problems.

http://www.nxp.com/documents/data_sheet/74F07.pdf

Paul
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Similar Threads


Top