No POST problem after installing new CPU HSF

[damc2000]

Hello.

My PC's failed to POST ever since I tried to install a new HSF for my
CPU. It powers on okay but the D-bracket diagnostic LEDs go immediately
to 1-2
red, 3-4 green. No beeps. According to my mobo manual this is related
to the IDE controller initialisation. I've seen a few posts with
similar issues but
with no definitive solution. Here's what I've tried;

Cleaned and reseated CPU, refitted HSF.
Checked all mobo leads etc.
Cleared CMOS, both jumper and battery-removal methods.
Removed mobo from case, removed all cards etc.
Tried known good RAM stick.

All of above produced the same LED error code.

I tested the PSU mobo voltage. All plugs were normal apart from the
-12v plug. It measured at -10.58v. Could this be the cause of the
problem?

If not PSU then I guess it must be either CPU or mobo. When I fitted
the new HSF I might have been a bit sloppy with the thermal grease;
some may have got on the wrong parts of the CPU top. There is no
outward sign of fried CPU though.

I would be grateful for any insight as I don't want to invest in a new
component unless I'm sure that it's the right one to replace. I don't
have a known good PSU, CPU or mobo to check which is faulty.

[MSI KT3 Ultra2, Athlon XP 2500+ (Barton), 1GB Crucial 2700 RAM, 400w
PSU, Speeze SPA04S4-U WhisperRock V Sk A HSF, Realtek Dynamode 10/100
NIC, Nvidia GeForce FX5200 AGPx4, Maxtor Diamondmax Plus9 ATA130 80
Gig]

damc2000

 

[kony]

Hello.

My PC's failed to POST ever since I tried to install a new HSF for my
CPU. It powers on okay but the D-bracket diagnostic LEDs go immediately
to 1-2
red, 3-4 green. No beeps. According to my mobo manual this is related
to the IDE controller initialisation. I've seen a few posts with
similar issues but
with no definitive solution. Here's what I've tried;

I assume power was off while doing this and that you had
grounded yourself to prevent ESD. Beyond that they typical
things should be rechecked and rechecked again- cables,
cards, memory, any of which may have come loose. It can
help to quit and later look at it again, you might then see
things you didn't previously. A strong light might help
too.

Cleaned and reseated CPU, refitted HSF.
Checked all mobo leads etc.
Cleared CMOS, both jumper and battery-removal methods.
Removed mobo from case, removed all cards etc.
Tried known good RAM stick.

All of above produced the same LED error code.

When you wrote "removed mobo from case" did you then try it
outside of case or reinstall it first? If outside of case,
don't run it on an antistatic surface/bag/pad/whatever as
they conduct electricity and have been observed to prevent
proper operation.

I tested the PSU mobo voltage. All plugs were normal apart from the
-12v plug. It measured at -10.58v. Could this be the cause of the
problem?

No the -12V is not used and can be ignored. Check the
battery voltage.

If not PSU then I guess it must be either CPU or mobo. When I fitted
the new HSF I might have been a bit sloppy with the thermal grease;
some may have got on the wrong parts of the CPU top. There is no
outward sign of fried CPU though.

What kind of heatsink compound was it? Arctic Silver or
other silver or copper paste should not be allowed around
any contacts or surface components/bridges/etc but the
ceramic or typical zinc paste should not be a problem.

I'd check the core, just in case it might've been chipped by
the heatsink installation, though I expect you checked that
the last time you pulled the 'sink off.

I would be grateful for any insight as I don't want to invest in a new
component unless I'm sure that it's the right one to replace. I don't
have a known good PSU, CPU or mobo to check which is faulty.

Well I dont' think it's the IDE controller but towards
reducing variables, it'd be good to leave all non-essential
things disconnected till it's POSTing again, leave only CPU,
video, heatsink/fan, 1 memory module for further attempts.
IF it's possible the video card was damaged (like flexing or
contact that discharged electrostatically, you might try
another video card.

[MSI KT3 Ultra2, Athlon XP 2500+ (Barton), 1GB Crucial 2700 RAM, 400w
PSU, Speeze SPA04S4-U WhisperRock V Sk A HSF, Realtek Dynamode 10/100
NIC, Nvidia GeForce FX5200 AGPx4, Maxtor Diamondmax Plus9 ATA130 80
Gig]

What power supply? If it's a generic, I'd suspect it next.
If not generic, hard to guess (and guess is really what it
would be)... motherboard is definitely possible, or a good
name-brand PSU failing, video, cards/memory/cable contacts,
other things I can't think of at the moment.

 

[damc2000]

Many thanks for the detailed reply.

[I assume power was off while doing this and that you had
grounded yourself to prevent ESD. Beyond that they typical
things should be rechecked and rechecked again- cables,
cards, memory, any of which may have come loose. It can
help to quit and later look at it again, you might then see
things you didn't previously. A strong light might help
too.]

I did wear an anti-static strap so hopefully I didn't damage anything
that way.

[When you wrote "removed mobo from case" did you then try it
outside of case or reinstall it first? If outside of case,
don't run it on an antistatic surface/bag/pad/whatever as
they conduct electricity and have been observed to prevent
proper operation.]

Tried it outside case on piece of foam.

[What kind of heatsink compound was it? Arctic Silver or
other silver or copper paste should not be allowed around
any contacts or surface components/bridges/etc but the
ceramic or typical zinc paste should not be a problem.]

Just says white grease. I assume ceramic.

[What power supply? If it's a generic, I'd suspect it next.
If not generic, hard to guess (and guess is really what it
would be)... motherboard is definitely possible, or a good
name-brand PSU failing, video, cards/memory/cable contacts,
other things I can't think of at the moment.]

My PSU is a cheap generic one. I probably should get a better one but I
thought that since the voltage measurements for the mobo socket were ok
then it was still funtional. I think I can discount cards including
video card as the same error code is produced with only CPU/HSF and RAM
left in. I agree mobo is a strong possibility. The onboard sound packed
up on in last month also.

damc2000