A8V and Athlon 3700+ "System Failed CPU Test" ???

AusKerb · 9th August 2006, 12:24 AM

Hi all ...

A8V Deluxe - BIOS - 1018 ( Latest Beta )
ATI 9550 ( AGP - no extra power needed )
GEIL PC3500 - 2 X 512mb
Athlon 3700+ ( "San Diego" (90 nm) 2200 MHz 1024 KiB 1000 MHz 11x 1.35/1.40 V )
Antec 400 WATT - Model - SL400

How did this all start ? Ok ... I upgraded my GF's PC with an old board I had ... the A8v ... I bought her a new processor ... the 3700+ and slapped it together with other stuff I had lying around ...Right ... I got it going after clearing the CMOS and setting up bios, and installed windows ... I then flashed to the latest stable bios ( 1017 ) from ASUS site ... all was well ...
... I shut the PC down for the night and went to bed ... until then the PC had only been restarted and not shut down or powered off ...

... in the morning I start it up to show my GF how she has a new PC and it turns on ... all fans and HDD's but no screen ... and lasts about 5 secs ... and shuts off ... I try again ... same thing ...

I try a 3rd time and it stays running ... but then after 10secs or so and still no screen I get the dreaded ..."System Failed CPU Test" from the speakers ... So I shut the PC off ... try try again ... etc ... The test is repeatable ... if I switch the PSU off ... let the board "power out" and switch it on again ... Power button ... and 5 secs turns off ... and again ... and 3rd time stays on and "System Failed CPU Test" from the speakers

ok ... so hours later and after reading many many posts on this I have tried a few things and this is what I have ...

1 - Checked the CPU fan was in the right header ... yep ...

2 - Take memory out ... same thing ...

3 - Try different vid card ... same thing

4 - Increase VCore ... increased the VCore to 1.4V ( max std. rating ) ... same thing ...

5 - Tried another BIOS ... Installed the Latest beta 1018 ... same thing

6 - Power Supply is running at - 3.23V for the 3.3V and 4.97V for the 5V ... So I tried another power supply thinking it might be a little low on 3.3V ... Same thing ... ( It was only a 250WATT model, but I only had 1 HDD plugged in and no other periphierals to maximise the power to the board )

.... I'm realing starting to suspect the motherboard is dodgey ... especially so after two things ... 1st I was reading about some dodgey leaking capacitor issue with boards around this age and the quirk I found that seemed to be linked to that issue ...

Now there is a real quirk ... I discovered early on when I was hunting the issue that if I reset the CMOS with the jumper to defaults, the PC would boot as normal ... Further, if I left it up and running it would continue into bios and windows ... all works right as rain ... It will even restart ok ...BUT ... If I shut it down for 1 minute or more or switch the switch on the PSU off for more than 10 secs the same old story ... 5 secs off twice then "System Failed CPU Test" ... reset the CMOS and bang it's back up again ...

... also I did further testing ... If I have it running and switch the PSU off for a 1 second and hit the power up button immediatly it will be ok ... but If I switch the PSU switch for more than enough for the lights on the motherboard to go out ( 10 secs or so ) ... same old problem ...

So ... it makes me believe that there is some part of the motherboard, that is "losing" it's charge or something ... and it seems to be affected still when the PSU is on, but the PC is shutdown for a mintue or more ...

Anyone with idea's ?

thx in advance ... sorry for the book ...

Kerb

fo3 · 9th August 2006, 12:38 AM

I'd try the cpu in another board.
If thats fine in another board, well you have the answer.
It seems weird though, if something is going wrong with the mobo they don't get far enough into POST to blame other gear. I've had a few slot 1 mobos slowly dying for caps and other reasons and they never come up with a cpu test fail. Closest was a dead abit one with a high/low cpu temp error, but that was DOA, the others just ran unstable or didn't POST at all until BIOS reset etc.
If the mobo is 'losing charge', it should reset the bios and boot in failsafe mode, no screw something up bad enough that it does something to the cpu so it doesn't POST ok.
I'd check the mobo for shorts (eg to the case), cut tracks, scorches/burn marks, blown transistors. Then try the cpu on another mobo.

edit: V
A 3.2 voltage on the 3.3 rail doesn't cause the behaviour you're experiencing.

AusKerb · 9th August 2006, 12:41 AM

I orginally had a 3500+ out of my old PC in the setup I built for the GF ... this processor was on the way out and happened to die the morning I boot it up after the power down for the GF to see a blank screen ... But it did survive the whole windows install and setup, plus the bios flash ...

... coincidence I believe ... as I have a Water Cooling setup in my PC where it came from and it had been running very hot in there prior to me moving it to the Air cooling H/S/F combo in the GF's new PC ... Also it had been O/C'ed heaps ... I believe that was the last straw for it getting that hot ...

... so I bought the 3700+ for her brank spankers and as I said it works fine when I use the CMOS trick to boot the PC ... I have also had her PC running with my new X2 4200+ CPU, but only to make sure it turned on, when troubleshooting to see if the 3500+ had actually died before the 3700+ was delivered ...

Motherboard is dead then ?

3.3V rail within spec ?

9th August 2006, 12:24 AM	#1
AusKerb Member Join Date: Nov 2003 Posts: 8	A8V and Athlon 3700+ "System Failed CPU Test" ??? Hi all ... A8V Deluxe - BIOS - 1018 ( Latest Beta ) ATI 9550 ( AGP - no extra power needed ) GEIL PC3500 - 2 X 512mb Athlon 3700+ ( "San Diego" (90 nm) 2200 MHz 1024 KiB 1000 MHz 11x 1.35/1.40 V ) Antec 400 WATT - Model - SL400 How did this all start ? Ok ... I upgraded my GF's PC with an old board I had ... the A8v ... I bought her a new processor ... the 3700+ and slapped it together with other stuff I had lying around ...Right ... I got it going after clearing the CMOS and setting up bios, and installed windows ... I then flashed to the latest stable bios ( 1017 ) from ASUS site ... all was well ... ... I shut the PC down for the night and went to bed ... until then the PC had only been restarted and not shut down or powered off ... ... in the morning I start it up to show my GF how she has a new PC and it turns on ... all fans and HDD's but no screen ... and lasts about 5 secs ... and shuts off ... I try again ... same thing ... I try a 3rd time and it stays running ... but then after 10secs or so and still no screen I get the dreaded ..."System Failed CPU Test" from the speakers ... So I shut the PC off ... try try again ... etc ... The test is repeatable ... if I switch the PSU off ... let the board "power out" and switch it on again ... Power button ... and 5 secs turns off ... and again ... and 3rd time stays on and "System Failed CPU Test" from the speakers ok ... so hours later and after reading many many posts on this I have tried a few things and this is what I have ... 1 - Checked the CPU fan was in the right header ... yep ... 2 - Take memory out ... same thing ... 3 - Try different vid card ... same thing 4 - Increase VCore ... increased the VCore to 1.4V ( max std. rating ) ... same thing ... 5 - Tried another BIOS ... Installed the Latest beta 1018 ... same thing 6 - Power Supply is running at - 3.23V for the 3.3V and 4.97V for the 5V ... So I tried another power supply thinking it might be a little low on 3.3V ... Same thing ... ( It was only a 250WATT model, but I only had 1 HDD plugged in and no other periphierals to maximise the power to the board ) .... I'm realing starting to suspect the motherboard is dodgey ... especially so after two things ... 1st I was reading about some dodgey leaking capacitor issue with boards around this age and the quirk I found that seemed to be linked to that issue ... Now there is a real quirk ... I discovered early on when I was hunting the issue that if I reset the CMOS with the jumper to defaults, the PC would boot as normal ... Further, if I left it up and running it would continue into bios and windows ... all works right as rain ... It will even restart ok ...BUT ... If I shut it down for 1 minute or more or switch the switch on the PSU off for more than 10 secs the same old story ... 5 secs off twice then "System Failed CPU Test" ... reset the CMOS and bang it's back up again ... ... also I did further testing ... If I have it running and switch the PSU off for a 1 second and hit the power up button immediatly it will be ok ... but If I switch the PSU switch for more than enough for the lights on the motherboard to go out ( 10 secs or so ) ... same old problem ... So ... it makes me believe that there is some part of the motherboard, that is "losing" it's charge or something ... and it seems to be affected still when the PSU is on, but the PC is shutdown for a mintue or more ... Anyone with idea's ? thx in advance ... sorry for the book ... Kerb

9th August 2006, 12:38 AM	#2
fo3 (Taking a Break) Join Date: Apr 2002 Posts: 2,972	I'd try the cpu in another board. If thats fine in another board, well you have the answer. It seems weird though, if something is going wrong with the mobo they don't get far enough into POST to blame other gear. I've had a few slot 1 mobos slowly dying for caps and other reasons and they never come up with a cpu test fail. Closest was a dead abit one with a high/low cpu temp error, but that was DOA, the others just ran unstable or didn't POST at all until BIOS reset etc. If the mobo is 'losing charge', it should reset the bios and boot in failsafe mode, no screw something up bad enough that it does something to the cpu so it doesn't POST ok. I'd check the mobo for shorts (eg to the case), cut tracks, scorches/burn marks, blown transistors. Then try the cpu on another mobo. edit: V A 3.2 voltage on the 3.3 rail doesn't cause the behaviour you're experiencing. Last edited by fo3; 9th August 2006 at 12:45 AM.

9th August 2006, 12:41 AM	#3
AusKerb Member Join Date: Nov 2003 Posts: 8	I orginally had a 3500+ out of my old PC in the setup I built for the GF ... this processor was on the way out and happened to die the morning I boot it up after the power down for the GF to see a blank screen ... But it did survive the whole windows install and setup, plus the bios flash ... ... coincidence I believe ... as I have a Water Cooling setup in my PC where it came from and it had been running very hot in there prior to me moving it to the Air cooling H/S/F combo in the GF's new PC ... Also it had been O/C'ed heaps ... I believe that was the last straw for it getting that hot ... ... so I bought the 3700+ for her brank spankers and as I said it works fine when I use the CMOS trick to boot the PC ... I have also had her PC running with my new X2 4200+ CPU, but only to make sure it turned on, when troubleshooting to see if the 3500+ had actually died before the 3700+ was delivered ... Motherboard is dead then ? 3.3V rail within spec ?