No POST, CPU dead or voltage problem?

[yocky]

I'm attempting a new build with an Athlon 64 3000+ on an Epox 9npa+
Ultra board. The mobo has 3 Ledion indicators which supposedly show
whether there is a problem with the Chipset, DIMMs or the CPU.
On powering up with RAM (2x512 PC3200), CPU and graphics (Gainward 6600
PCI-e) the first two LEDs are on but the CPU one isn't. I get power to
three fans (CPU, Chipset and Graphics) and the POST LED is on but it
just shows FF. I get no further.

I assumed the CPU was dud but thought I'd check the PSU voltages. It's
an Antec 450W PSU and it supplies one 24 pin and one 4 pin ATX
connector on the motherboard.
All the voltages seem fine apart from at the Grey PWROK connection
which shows less than 1V (about 0.7V from memory). I'm led to believe
than this should be measuring 3-5V.

Could this be the root of my problem, or am I overcomplicating things
and it's just that the CPU or motherboard is dud.

I've tried reseating everything, clearing the CMOS, messing about with
the RAM configuration, tried w/o the graphics card, but always get the
same results.

 

[JAD]

http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/23811.pdf

Description of Power-Up Sequence

The power-up sequence is as follows:

1. The power supply ramps and the startup VID value becomes
stable.

2. At this time, the processor core regulator begins ramping
VCC_CORE. In addition, the SRAM regulator begins
ramping VCC_SRAM based on the value of VCC2SEL and
the system clock generator begins to power up.

3. A resistor network powered by VCC_SRAM begins to
provide VCCA.

4. The power supply asserts PWROK as much as 500ms after
the power supply begins to provide stable 5V and 3.3V. (This
is not the same signal as the PowerGood signal to the
Southbridge input).

5. PWROK is provided to the processor once the processor
core voltage regulator produces a stable voltage based on
VID and 5V from the power supply. In addition, pullup
resistors on the open-drain system clocks from the clock
generator become stable and SYSCLK/SYSCLK# begins
transitioning.

6. PowerGood is provided to the Southbridge once 5V is valid
from the power supply and VCC_CORE is valid.
(PowerGood to the Southbridge is based on processor
VCC_CORE, not PWROK). Before PowerGood is asserted,
the Southbridge drives CPURESET# and NBRESET#
asserted.

7. When the Southbridge sees the assertion of PowerGood, it
waits 1.8ms and then deasserts NBRESET#. It then waits
1.5µs before deasserting CPURESET#.

 

[yocky]

I've got to admit I'm winging this a little and interpreting that info
is beyond my capabilities. Either my CPU is dead and the voltage
reading is indicating that, or the PSU or motherboard is causing
insufficient power to be delivered to the CPU and that's the reason for
the lack of CPU LED action.

Beyond that diagnosis I'm lost.

 

[yocky]

I've got to admit I'm winging this a little and interpreting that info
is beyond my capabilities. Either my CPU is dead and the voltage
reading is indicating that, or the PSU or motherboard is causing
insufficient power to be delivered to the CPU and that's the reason for
the lack of CPU LED action.

Beyond that diagnosis I'm lost.

 

[Peter]

I'm attempting a new build with an Athlon 64 3000+ on an Epox 9npa+
Ultra board. The mobo has 3 Ledion indicators which supposedly show
whether there is a problem with the Chipset, DIMMs or the CPU.
On powering up with RAM (2x512 PC3200), CPU and graphics (Gainward 6600
PCI-e) the first two LEDs are on but the CPU one isn't. I get power to
three fans (CPU, Chipset and Graphics) and the POST LED is on but it
just shows FF. I get no further.

I assumed the CPU was dud but thought I'd check the PSU voltages. It's
an Antec 450W PSU and it supplies one 24 pin and one 4 pin ATX
connector on the motherboard.
All the voltages seem fine apart from at the Grey PWROK connection
which shows less than 1V (about 0.7V from memory). I'm led to believe
than this should be measuring 3-5V.

Could this be the root of my problem, or am I overcomplicating things
and it's just that the CPU or motherboard is dud.

I've tried reseating everything, clearing the CMOS, messing about with
the RAM configuration, tried w/o the graphics card, but always get the
same results.

Only experience I've had with a problematic Epox mobo giving FF reading
and wouldn't boot, it turned out to be a faulty memory slot. When I
moved the memory to a different slot the computer booted up just fine.
Moved it back to original slot and same FF problem. From your details
you may have already tried this option, though have you tried just
either memory stick in all slots?

 

[Michael Hawes]

I've got to admit I'm winging this a little and interpreting that info
is beyond my capabilities. Either my CPU is dead and the voltage
reading is indicating that, or the PSU or motherboard is causing
insufficient power to be delivered to the CPU and that's the reason for
the lack of CPU LED action.

Beyond that diagnosis I'm lost.

http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/23811.pdf

Description of Power-Up Sequence

The power-up sequence is as follows:

1. The power supply ramps and the startup VID value becomes
stable.

2. At this time, the processor core regulator begins ramping
VCC_CORE. In addition, the SRAM regulator begins
ramping VCC_SRAM based on the value of VCC2SEL and
the system clock generator begins to power up.

3. A resistor network powered by VCC_SRAM begins to
provide VCCA.

4. The power supply asserts PWROK as much as 500ms after
the power supply begins to provide stable 5V and 3.3V. (This
is not the same signal as the PowerGood signal to the
Southbridge input).

5. PWROK is provided to the processor once the processor
core voltage regulator produces a stable voltage based on
VID and 5V from the power supply. In addition, pullup
resistors on the open-drain system clocks from the clock
generator become stable and SYSCLK/SYSCLK# begins
transitioning.

6. PowerGood is provided to the Southbridge once 5V is valid
from the power supply and VCC_CORE is valid.
(PowerGood to the Southbridge is based on processor
VCC_CORE, not PWROK). Before PowerGood is asserted,
the Southbridge drives CPURESET# and NBRESET#
asserted.

7. When the Southbridge sees the assertion of PowerGood, it
waits 1.8ms and then deasserts NBRESET#. It then waits
1.5µs before deasserting CPURESET#.

Try removing memory and all adapters and data cables. Should give beep
code for 'no memoy'. If doesn't, is motherboard, CPU or PSU. Try a known
good PSU.
Mike.

 

[yocky]

I've got to admit I'm winging this a little and interpreting that info
is beyond my capabilities. Either my CPU is dead and the voltage
reading is indicating that, or the PSU or motherboard is causing
insufficient power to be delivered to the CPU and that's the reason for
the lack of CPU LED action.

Beyond that diagnosis I'm lost.


Try removing memory and all adapters and data cables. Should give beep
code for 'no memoy'. If doesn't, is motherboard, CPU or PSU. Try a known
good PSU.
Mike.

I've already tried with no memory and get the same result. I think I
have to get to the bottom of this CPU LED not coming on. It's either a
dead CPU (most likely I think), PSU problem because of the voltage
issue I mentioned (waiting in vain for Antec to get back to me), or
mobo issue (need to find out exacttly what this CPU LED indicates).

 

[yocky]

I'm attempting a new build with an Athlon 64 3000+ on an Epox 9npa+
Ultra board. The mobo has 3 Ledion indicators which supposedly show
whether there is a problem with the Chipset, DIMMs or the CPU.
On powering up with RAM (2x512 PC3200), CPU and graphics (Gainward 6600
PCI-e) the first two LEDs are on but the CPU one isn't. I get power to
three fans (CPU, Chipset and Graphics) and the POST LED is on but it
just shows FF. I get no further.

I assumed the CPU was dud but thought I'd check the PSU voltages. It's
an Antec 450W PSU and it supplies one 24 pin and one 4 pin ATX
connector on the motherboard.
All the voltages seem fine apart from at the Grey PWROK connection
which shows less than 1V (about 0.7V from memory). I'm led to believe
than this should be measuring 3-5V.

Could this be the root of my problem, or am I overcomplicating things
and it's just that the CPU or motherboard is dud.

I've tried reseating everything, clearing the CMOS, messing about with
the RAM configuration, tried w/o the graphics card, but always get the
same results.

It was the PSU.

I would have tried a different one earlier but I didn't think the old
one I've got had a 20/24 pin ATX and a 4 pin one but it does, it fired
everything up straight off.

It just goes to show (I think) what a lottery it is. I assumed that
because the PSU seemed to be working generally that it was working
completely. I was ready to send the CPU back and would have done if it
wasn't for the CPU Ledion indicator on the Epox board combined with my
voltage check.

Quite a learning curve with this stuff isn't it?!

 

[yocky]

Will replacing the new Antec PSU with the same model definitely solve
this issue? Or is there a possibility it's incompatible with my setup?

 

[fj]

Will replacing the new Antec PSU with the same model definitely solve
this issue? Or is there a possibility it's incompatible with my setup?

As long as the power supply has the necessary cables to plug into the power
sockets/receptacles on the motherboard and has enough amps [for each of the
'rails' - i.e., 5v, 3.3v, 12v, etc], there shouldn't be any compatibilty
issues.

 

[yocky]

Will replacing the new Antec PSU with the same model definitely solve
this issue? Or is there a possibility it's incompatible with my setup?

As long as the power supply has the necessary cables to plug into the power
sockets/receptacles on the motherboard and has enough amps [for each of the
'rails' - i.e., 5v, 3.3v, 12v, etc], there shouldn't be any compatibilty
issues.

Well that's what I thought but Antec support have warned me they are
looking into just that possibility. Also when I tried my new PSU on an
old motherboard last night it worked a treat.

I know what you mean but it does appear that I have a working mobo and
a working PSU which aren't playing nicely together.

I've tested the voltages, how do I test the amperage?