Tech Support banner

Status
Not open for further replies.
1 - 15 of 15 Posts

·
Registered
Joined
·
21 Posts
Discussion Starter #1
Hi,

Hope that somebody can assist me with this frustrating issue which continues to perplex my limited computer system knowledge :xolconfus

To offer brief background this system was my primary PC until Jan 2010 prior to replacement with a custom built i7 system. Prior to this new setup I was having to leave the system powered 24/7 as it had become a struggle to reboot successfully from cold However, once booted the system would function flawlessly and happily reboot/restart at will until switched off and a cold restart attempted.

Following a Motherboard failure resulting from a blown PSU on an unrelated Gigabyte spare system I deceided to utilise and upspec memory and PSU on the AA8XE system as a replacement. NB. At this stage the system has no OS or HDD's installed just CPU, memory and floppy/IDE optical drives.

Once again the intermittent cold boot issue has arisen always resulting in a post code fail at 9.0 (Complete µGuru initial process AWARD BIOS take over booting job) with no display. Several Power Off attempts/restarts later and I can successfully boot to FF and Disk Boot Failure (no HDD's installed remember!). Subsequent reboots/restarts are a breeze........until next 'cold' start of course.

I have noticed that when the cold boot 9.0 fail occurs the Graphics Fan does not spin up at boot however when the full boot sequence initiates correctly the Graphics fan comes immediately to life as soon as the power button is pressed. NB. I have tried reseating the graphics card but it has made no difference.

Also the power button will not power off the system when at 9.0 fail......though I suspect this is because the system boot sequence has not fully initiated?

I have also removed all memory and still have the same issue excepting when successful restart fails at C1. Memory sticks replaced and swapped one at a time to eliminate bad memory and no issues found...problem persists.

Once booting normally CPU is recognised all memory (including above test configurations) recognised correctly and IDE optical drives recognised.......even boots off an OS cd!

Have already tried CMOS reset and I am running BIOS defaults at this stage other than 'graphics initialise PCIx' rather than default 'PCI' ......the settings are not currently overclocked.

Seem to recall that RAM voltages at AA8XE default were wrong (.....only a distant memory so may not be correct on that) if so might this cause a 9.0 and this issue??

As the issue is intermittent and relates only to a start from cold I suspect that it is related to the motherboard and wonder if it is possibly a problem with the BIOS chip?

Would welcome any advice to resolve this issue. :smile:
 

·
retired
Joined
·
1,966 Posts
this seems to be a well known issue with this board and the common answers are replace bios chip or rma the board.

on a warm boot (reboot) the GURU Power On Sequence isn't used because it's assumed the voltages are already stable (the psu is already in use) so only the AWARD sequence is processed and it does no voltage checking -- if there were a power problem though it would still be evident due to the system crashing.

so it would appear the problem is confined to the GURU Power On Sequence code of the post code (which is processed for the cold boot only).

your observation that the graphics fan is unpowered when there is a 90 failure would indicate there is a problem when initializing the PCIe power controller IC resulting in no PCIe voltage which would lead to a failed voltage ready state and thus the Power On Sequence fails.

if it were a code error then a bios update could have corrected it, so it's likely a circuit design error or component failure.

the odd thing about this is there are two solutions -- suggesting two problems -- one being there is a problem with (some) bios chips and the other being (some) boards have circuit (and/or signal) issues.

your solution, contact Abit and inform them you have the GURU 90 problem and would like the replacement bios chip to see if it resolves the cold boot issues.
this way you have established dialogue and requested the cheapest fix from them so if the board has to be rma'd you have already proven the chip was not the issue.

the other possibility is the unidentified psu(s) used were underpowered and had difficulty with the cold boot surge power requirements.
 

·
Registered
Joined
·
21 Posts
Discussion Starter #3
Hi Stu,

Many thanks for the comprehensive reply.....I should have added that the AA8XE system was originaly put together by me in 2005 and had no issues at all until 2008/2009....so it was a developed failure rather than always present.

In its original build the PSU utilised was the Enermax Noise Taker II 600W....this has now been replaced with a BeQuiet Dark Power Pro 650W unit.....which was already in a working system. Thus I am fairly confident it is not a PSU issue....unless anybody can suggest something I have missed?

Is it likely the BIOS chip problem could have 'developed'??? ......is component failure the more likely explanation.....

Is there any further information out there relating to the 'Guru 90 problem'??

Many thanks again for your timely response.

PC Pilot
 

·
retired
Joined
·
1,966 Posts
Download CPUZ 1.56 setup, english and run it , click the About tab then Save Report (TXT) and attach the text file to your post.

that board should have Rubycon capacitors so it's not a "bad caps" board.

How to load default for µGuru
µGuru load default Settings
Turn ON your PC with our µGuru™ motherboard and then turn OFF right away as one single action; repeat this action continuously for 3 times.

Now you have set µGuru to a load default setting!
BIOS settings:
Power Cycle Statistics: ABIT EQ
ABIT EQ Beep Control: enabled

CMOS Features Setup Menu
Halt On: No Errors

Advanced BIOS Features
Quick Power On Self Test: Enabled
Delay IDE Initial (Secs): 2
Disable Unused PCI Clock: No
Full Screen LOGO Show: Disabled

Advanced Chipset Features
(need to see cpuz report)
 

·
Registered
Joined
·
21 Posts
Discussion Starter #5
Hi Stu,

Many thanks again for your assistance.......very much appreciated I can tell you :smile: !!

In regard to the uGuru reset, the first problem is getting it to boot in the first place. To explain when the switch on results in the post code 90 error the PC's power switch does not function seemingly as the switch itself requires some action from the Award bios who's sequence at this stage is inhibbited!

Considering this further, I guess that once I am able to boot to the Award BIOS (the many random attemts pulling power from PSU :rolleyes:) and run the system for a while till warm I can try this step. Not sure if I am on the right lines here but...if the reset is successful the Power Cycle Stats will reset to 0 ??? NB. Was very lucky just.... got sucessful boot on only third power pull cycle.......will run for 10 minutes or so and advise later when reset actions carried out!

I have adjusted BIOS settings in accordance with those suggested......for the record before adjustment they were:

Power Cycle Statistics: Press Enter ---- *Runs to OC Guru No further options
ABIT EQ Beep Control: enabled

CMOS Features Setup Menu
Halt On: All but keyboard

Advanced BIOS Features
Quick Power On Self Test: Enabled
Delay IDE Initial (Secs): 1
Disable Unused PCI Clock: Yes
Full Screen LOGO Show: Disabled

* This was from the uGuru Utility menu within the Award BIOS....am I looking in the right place?

Already have latest CPUZ but remember I have no operating system installed as yet, in fact currently no HDD connected. Do you suggest I should install Windows 7 at this stage and run CPUZ?

Would welcom your thoughts on the following received from another board that this same issue was posted to....

Code 9.0 relates to a CPU fault. By that I don't mean that your CPU is faulty but when the uGuru chip passes over to the BIOS the CPU fails to post.

This can be caused by one of the following:
- Faulty CPU. (Try another CPU, if have one)
- Faulty PSU. (Intermittent 12v power to the CPU would cause this problem, try another PSU)
- Faulty motherboard. (Maybe a few caps on the motherboard starting to fail or failed. Hard to fix unless you can see any blown caps on the motherboard)

I doubt it is anything to do with the BIOS. If the BIOS was faulty you wouldn't be able to post at all.
Have now tried the uGuru reset actions and have the following observations:

1. Power Button does not cycle on to off as one single action:confused:
2. Power off not achieved until 'keyboard lock' led's (briefly) illuminate
3. Cycled system 3 times in this manner...closest I coud achieve to 1 action
4. Power cycle stats still held.....not sure if I am on the right track here in any case:confused:

Any further thoughts on all of this??

PC Pilot
 

·
retired
Joined
·
1,966 Posts
In regard to the uGuru reset, the first problem is getting it to boot in the first place. To explain when the switch on results in the post code 90 error the PC's power switch does not function seemingly as the switch itself requires some action from the Award bios who's sequence at this stage is inhibbited!
i believe they are referring to the psu's AC power switch being toggled off/on 3 times, not the case soft_on button.


I have adjusted BIOS settings in accordance with those suggested......for the record before adjustment they were:

Power Cycle Statistics: Press Enter ---- *Runs to OC Guru No further options
ABIT EQ Beep Control: enabled

CMOS Features Setup Menu
Halt On: All but keyboard

Advanced BIOS Features
Quick Power On Self Test: Enabled
Delay IDE Initial (Secs): 1
Disable Unused PCI Clock: Yes
Full Screen LOGO Show: Disabled

* This was from the uGuru Utility menu within the Award BIOS....am I looking in the right place?
yes, thats it. the two i highlight above were the salient ones, these two settings can often hang a system for inconsequential errors. certainly don't need the added confusion from some unimportant event tossing its error into the mix.

Already have latest CPUZ but remember I have no operating system installed as yet, in fact currently no HDD connected. Do you suggest I should install Windows 7 at this stage and run CPUZ?
better to get Hiren's BootCD 13.0 it's excellent and a must have for the serious tech toolkit.
boot to the miniXP -> HBCD->programs->system info->cpuz then save to a floppy or usb drive.

just wanted to make sure there isn't an anomaly with the hardware or how the system is using/misusing it.
Would welcom your thoughts on the following received from another board that this same issue was posted to....

Quote:Code 9.0 relates to a CPU fault. By that I don't mean that your CPU is faulty but when the uGuru chip passes over to the BIOS the CPU fails to post.

This can be caused by one of the following:
- Faulty CPU. (Try another CPU, if have one)
- Faulty PSU. (Intermittent 12v power to the CPU would cause this problem, try another PSU)
- Faulty motherboard. (Maybe a few caps on the motherboard starting to fail or failed. Hard to fix unless you can see any blown caps on the motherboard)

I doubt it is anything to do with the BIOS. If the BIOS was faulty you wouldn't be able to post at all.

my understanding of this 9.0 code is that it shows what is being performed next (ie: finished uguru tests and next action is to send the reset signal to CPU).

the test point (and hopefully the code number shown) just prior to the 9.0 display should be the actual error point.

a little background...
a mobo is managed by the chipset, or to put it another way, the mobo is designed around the chipset--to make use of the features of that chipset. everything you can plug into a mobo (cpu, ram, drives, etc) is subject to the chipset.

when you power_on the system it's the chipset, (and in this case abit's uguru specialty chip), that takes control. it will test to make sure system voltages and cpu fan are ready for use then it sends a reset signal to the CPU which will then start processing whatever is in it's registers. On a cold boot there is nothing in the registers so it checks the boot/reboot bit and goes to it's default memory address (bios chip) and processes the code found there--which happens to be some part of the bios post sequence depending whether booting/rebooting.
(for an OS that suspends to ram: the CPU receives the reset signal and simply processes the code that is already in it's registers.)

i have no idea why abit lists swapping the bios chip as a possible solution, it's their product so they're the experts on its details.

occassionally, a bios chip doesn't make full contact with the socket and can cause intermittent errors--it's very difficult to diagnose an intermittent fault, often it requires a 'force failure technique' to identify the fault point--and one has to keep in mind that anywhere there is a mechanical connection used for electrical paths (ie: cpu, ram, bios, cable connectors, pci slots) is a potential intermittent fail point.

as for board components, semiconducors will usually fail within 3 months if its defective else will last for decades, resistors rarely fail on their own, and capacitors degrade over time so their electrical characteristics alter and is the most likely reason for board failures to eventually occur.
of course a shock to the system (ie: line spike) could degrade any of these components and a system would start having faults thereafter.

financially it's simply not cost effective to spend a lot of time trying to discern the problem. normally, if it can't be easily diagnosed and corrected then swap parts and get it back into production.

nowadays only the hobbyist has the time and interest, (and skill), in troubleshooting to the component level.

as for the board itself a hairline break in a circuit path or a cold solder spot on one of the component connections can mimic the problem your having. often the tell-tale is the very fact the board is difficult to cold start but has no problem with restarts after it has warmed up.

to troubleshoot this you would need to pull the board and do a close visual inspection to see if it's a visible fault, next try booting while still out-of-case as there is no stress on board from mounting screws, next let board set for half hour then use a portable blow dryer on low setting to heat a section of board for 10~15 seconds then boot and repeat until have done whole board or found fault area. beyond that would use a multimeter to test point-to-point while applying a slight finger pressure to components to check if bad solder point.
 

·
Registered
Joined
·
21 Posts
Discussion Starter #7
Hi Stu,

Thank you again for the latest comprehensive reply:Wow1:

as for the board itself a hairline break in a circuit path or a cold solder spot on one of the component connections can mimic the problem your having. often the tell-tale is the very fact the board is difficult to cold start but has no problem with restarts after it has warmed up.

to troubleshoot this you would need to pull the board and do a close visual inspection to see if it's a visible fault, next try booting while still out-of-case as there is no stress on board from mounting screws, next let board set for half hour then use a portable blow dryer on low setting to heat a section of board for 10~15 seconds then boot and repeat until have done whole board or found fault area. beyond that would use a multimeter to test point-to-point while applying a slight finger pressure to components to check if bad solder point.
So time for an update on my progress:

1. Board now removed from chassis and chassis M/B Tray
2. Visual inspection reveals no obvious defects
3. Board connected (out of chassis) to Graphics card, Chassis switches, CD's
4. CPU swapped with P4 3.4 GHz spare
5. Boot attempt fails at 9.0 as before (with 3.73 EE CPU)
6. Eventual boot as before after numerous attempts
7. Board rested 30 mins
8. Dryer played upon board in quarters - results 4 X random boot as before
9. Guru chip reset repeated from PSU on/off cycle results as reported before

my understanding of this 9.0 code is that it shows what is being performed next (ie: finished uguru tests and next action is to send the reset signal to CPU).

The test point (and hopefully the code number shown) just prior to the 9.0 display should be the actual error point.
10. Error code preceeding 9.0 fail is 8.9

better to get Hiren's BootCD 13.0 it's excellent and a must have for the serious tech toolkit.
boot to the miniXP -> HBCD->programs->system info->cpuz then save to a floppy or usb drive.

just wanted to make sure there isn't an anomaly with the hardware or how the system is using/misusing it.
11. Hirens Boot CD obtained (excellent tip.....thanks)
12. CPU-Z run and generated report attached

I propose to repeat with the 3.73 EE CPU in situ

Should I remove/clean the BIOS chip at this stage??

You continued advice very much appreciated!!!

PC Pilot
 

Attachments

·
Registered
Joined
·
21 Posts
Discussion Starter #8
Hi again Stu,

Further progress report..............

I have now swapped the P4 3.73 GHz EE CPU and repeated the boot to CPU-Z this file is now attached.

Some interesting, perhaps related? :confused:, observations:

1. Temperature output reading from CPU shows on Guru Clock & Abit EQ with 3.4 Ghz but not with 3.73 GHz ???
2. Initial restart attempts posted to 9.0 fail but with constant cycling beep alarm?
3. After approx half a dozen attempts (following system connection recheck - no errors found) posted to 9.0 fail with no alarm???
4. As per usual repeated attempts result in normal boot

Any thoughts here????

Cheers!!

PC Pilot
 

Attachments

·
Registered
Joined
·
21 Posts
Discussion Starter #9
Furthe quick update.....CPU fan speed apears to be making random adjustments, sounds like its revving almost like hunting for a comfortable speed....have checked out connections etc no obvious defects????
 

·
Registered
Joined
·
21 Posts
Discussion Starter #10
Sorry for the rapid progress updates....maybe onto something here.....

Whilst seeking to address the reving CPU fan applied pressure to the corner of the fan/heatsing assembly (thus slight board flex) fan immediately speeds from 1440 rpm to 3240 rpm and guess what we have a temperature readout of 46 degrees C for the cpu.

Relax pressure and rpm reduces plus temp readout returns to zero degrees C!!

Now to see if we can identify what is being flexed.......

Will keep you posted

PC Pilot
 

·
retired
Joined
·
1,966 Posts
check between cpu and socket for a hair or debri.

there are a couple of oddities in the print outs that should be investigated.
i suggest using the 3.4Ghz (550 cpu) and set bios to fail safe settings.
then remove the two Micron sticks completely (from slots 3.4), and move the 2nd Crucial stick to slot 3 and make cpuz crucial report.

next, remove Crucial sticks and put the Microns in slots 1 and 3 for cpuz micron report.
-------------------------------------
both reports show the UGuru monitor is getting a consistent undervoltage reading from the ATX +3.3V DUAL regulator

that would correspond to failed on uGuru POST 8.C. 3VDUAL voltage ready
 

·
Registered
Joined
·
21 Posts
Discussion Starter #12
Hi Stu,

check between cpu and socket for a hair or debri.
I have inspected the CPU contacts and the LGA775 pin array for signs of debris and 'airdusted' to ensure they are clear of any contaminents. I also checked the LGA775 latch tension just incase the CPU was not seating firmly enough.

i suggest using the 3.4Ghz (550 cpu) and set bios to fail safe settings.
Interestingly I had already swapped back to the 3.4Ghz CPU before receiving your informative response.

OK so now to update you on the current position.....

Since making the processor swaps and identifying the board flex issue I have kind of hit the buffers. I can no longer get past the 9.0 fail with EITHER CPU :sad:

The CPU temp reading however is now absent on the guru clock (cannot now get to BIOS so cant give an Abit EQ observation off the Uguru menu) on BOTH CPU's but records with slight board flex between the CPU Fan socket and its nearest CPU Heatsink mounting screw (if you are visualy familiar withh the AA8XE) as noted in earlier post.

Interestingly the CPU fan hunting has also disappeared but it is now running permanently at the lower speed of 1440 ish rpm which I think is too slow. The board flex no longer varies this :confused:

On cold boot the PCI Express fan invariably fails to start, however as before it can be started after many random boot attempts but this no longer coincides with the initiation of boot sequence. NB. Once the board has started this Graphics card fan it will easilly start on subsequent (now warm?) boot attempts but still cannot get past the 9.0 fail......any thoughts?

I am interested in your observations concerning the undervoltage readings....
both reports show the UGuru monitor is getting a consistent undervoltage reading from the ATX +3.3V DUAL regulator

that would correspond to failed on uGuru POST 8.C. 3VDUAL voltage ready
as noted earlier I feel the CPU Fan Speed is incorrect and suspect this might possibly be related to these same voltage issues and wonder if this is also controlled by the uGuru chip???

then remove the two Micron sticks completely (from slots 3.4), and move the 2nd Crucial stick to slot 3 and make cpuz crucial report.

next, remove Crucial sticks and put the Microns in slots 1 and 3 for cpuz micron report.
Given the (now apparently permanent?) 9.0 fail is there any benefit from performing the 'memory swap' seeing as I cannot boot to Mini WindowsXP at this point and thus cannot run CPUz! I can however note the voltage readings registered on the '3rd eye' Guru clock if this helps.....

Awaiting your thoughts once more :4-cheers:

PC Pilot
 

·
Registered
Joined
·
21 Posts
Discussion Starter #13
......Forgot to mention, I also no longer have a CPU Frequency reading on the Guru clock with either CPU......both were visible before even to a 'failed at 9.0' start.

Not sure if this holds any significance????
 

·
Registered
Joined
·
21 Posts
Discussion Starter #14
Calling Stu_computer Re: Abit AA8XE 3rd Eye - Intermittent Cold Boot - Post Code 9.0

Hi again Stu.....and everybody

Sorry to trouble you again, still trying to resolve this M/B issue :confused: with my AA8XE 3rd Eye.....been a while since I posted my replies guess you have all been busy.

Hope you can point me in the right direction!!

Thanks again
 

·
retired
Joined
·
1,966 Posts
i stop by once awhile so don't expect prompt replies from me. :grin:

i see there is nothing new (since your last post). i assume you cleared cmos, tried a different graphics card, etc...

from what i read here with board flex creating different faults i'm inclined to think the socket/tracings have become unreliable.

the p4 cpu throws a lot of heat and a fair amount is dissipated down into the socket/mobo which may prematurely (and eventually will) weaken those connections--with multilayer boards there is no way to repair this type of failure.

sometimes a psu with a high ampere rating (30A or more) on the 12volts line may get it up and running (brute force), if you have one available i suggest using your lower wattage cpu only. eventually that will fail as well though if cycled off/on (in other words if you intend to use it then it must always be left on).
but, i'm not suggesting this would be an acceptable solution because something as simple as a vibration (from the CD drive loading for example) could cause a reboot.

for what it's worth i suggest it's time to send that board to the bin and get a replacement so you can make good use of those other parts.
 
1 - 15 of 15 Posts
Status
Not open for further replies.
Top