LinuxQuestions.org
Share your knowledge at the LQ Wiki.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Hardware
User Name
Password
Linux - Hardware This forum is for Hardware issues.
Having trouble installing a piece of hardware? Want to know if that peripheral is compatible with Linux?

Notices


Reply
  Search this Thread
Old 01-08-2014, 12:35 PM   #1
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Rep: Reputation: 23
[Solved] Instability Intel i5 750 Random Shutdown and Boot Loop


System Spec
Intel i5 750 2.66GHz
Motherboard GA-P55A-UD3
BIOS Award F11
Fujitsu 500G Sata II HD
ATI HD5770 PCi Express 1G DDR3
Memory Crucial Ballistix 4G DDR3 Kit 1333 (2 pieces 2G each stick)


I posted previously on an unstable system, random shutdowns, and also randomly starting:
http://www.linuxquestions.org/questi...ty-4175487140/


I had another computer same hardware and swapped out all my components, originally thought it was Motherboard so ordered another on Ebay. All was well for 2 days, then randomly my system powered down.

I went through BIOS (Award F5 though new board was F11).
The only difference was the CPU Clock Drive Voltage (from
MIT BIOS page, sub menu Advanced Frequency).

On my friends PC his BIOS voltages for clock drive were:
PCI Clock Drive 800mV
PCI Express Clock Drive 900mV

On my BIOS and the new motherboard both were set at 700mV.
I changed the values in BIOS to 800 and 900mV respectively,
and all was well for 9 days.


Yesterday at 530am I heard a single BIOS beep and PC powered on (it was shut down not suspend or hibernation). I should have copied the exact BIOS message but was words to the effect of system was overclocked.
When I tried to restart this went into a "reboot loop" on for 1/2 second then off again never reaching BIOS. I am used to this and leave it powered off for a number of hours and then reset BIOS to defaults.

I have downloaded the datasheet for Intel i5 750 CPU:
http://download.intel.com/design/pro...hts/322164.pdf


but cannot find anything directly related to value for PCI clock drive.

The BIOS allows clock drive voltage from 700mV to 1V all within the core voltage range of the i5 750. Currently I've set Clock drive now to 900mV and will see how long system is stable.

Previously system was stable for 3 years, so something is ageing or out of tolerance, its such a sporadic problem and
difficult to fix.

Currently back using original motherboard and flashed BIOS to F11. One thing is the changes in BIOS have made this stable for a few days.
Anyone have experience in faults like this or know the voltage limits for clock signals on Intel i5.


Sorry for long post.
Thanks in advance for any help.

Last edited by hal8000b; 01-18-2014 at 05:18 PM.
 
Old 01-09-2014, 09:01 PM   #2
jailbait
LQ Guru
 
Registered: Feb 2003
Location: Virginia, USA
Distribution: Debian 12
Posts: 8,334

Rep: Reputation: 547Reputation: 547Reputation: 547Reputation: 547Reputation: 547Reputation: 547
Last summer I had similar problems caused by an aging power supply. The power supply intermittently could not meet the power surge needed to start all of my hard drives and fans when I powered on. I replaced the old 75W power supply with a new 100W power supply and the problem disappeared.

My Dell desktop computer has small green diagnostic lights on the front panel. Each light has a number. By recording the sequence of diagnostic numbers on good power ups and failed power ups I was able to find an explanation on the Internet of what my problem was.

If you have another computer the same as the failing computer I suggest that you swap the power supplies and see what happens.

------------------------
Steve Stites
 
Old 01-10-2014, 11:33 AM   #3
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,982

Rep: Reputation: 492Reputation: 492Reputation: 492Reputation: 492Reputation: 492
Do you hear any high-pitched noises like squealing or squeaking ? If so, try to pinpoint the source.

What do the fans do during the shutdowns ?

Did you run memtest86 ?

Is shutdown more likely during high system load ?

I'm leaning towards a PSU issue ATM. Check the mobo for swollen capacitors or high-pitched noises.
 
Old 01-11-2014, 07:01 AM   #4
cascade9
Senior Member
 
Registered: Mar 2011
Location: Brisneyland
Distribution: Debian, aptosid
Posts: 3,753

Rep: Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935
I'd really doubt that increasing the clock drive voltage setting is going to make the system more stable-

Quote:
CPU/PCI-E Clock Driving Control

The default setting is 800mv, with a range of voltage control offered between 700mv-1000mv. As this is a differential amplifier circuit, increasing voltage may actually decrease the clock signal accuracy due to increased power supply noise. Differential circuits are used in preference to single ended circuits because of their noise rejection and low voltage operating capabilities. Increasing voltage to these circuits in turn increases "nasties" such as overshoot and output clock signal jitter. This in turn counteracts the benefits of using a differential amplifier in the first place. We did experiment with various levels of overvoltage and found no gains in stability whatsoever, further cementing our beliefs that more is not always better.
http://www.anandtech.com/show/2410/13

It could be PSU related. Of course, it might not be the PSU either, but its more likely than the PCIe clock drive voltage IMO.
 
Old 01-11-2014, 03:49 PM   #5
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Original Poster
Rep: Reputation: 23
Thank you all for your help.

Still have intermittent problems. Shutdown after several hours, can be 1 hour or 9 hours. Monitored using conky
all CPU cores running normal 30 - 40C, CPU core varies between 0.86 and 1.1V

Second problem is when switched off (not from sleep or standby) will turn on sometimes after 3 hours or 6 hours
later. The big clue is the message on BIOS:

"The system has experienced boot failures because of overclocking or changes of voltages. The last settings on
this page may not coincide with current hardware settings".


The BIOS is not overclocked but using default settings.
The above message is what appears when the machine has been in off state, so is it possible that the motherboard
voltage regulator is drifting when switched off? WOL and wake on ring are disabled in BIOS

I have since deliberately underclocked the CPU running at 2.60GHz and system shut down after 6 hours.
Have also tried fixing voltage core at 1.10V and still had system shutdown after 4 hours.

I have previously tried the PSU on a friends computer and it was ok for 6 days. The voltage regulator
on the gigabyte motherboards do have a bolt-on heatsink, not sure if this is overheating or not.

I do have another motherboard which I bought on Ebay which worked well for 2 days. With this motherboard
system powered itself down and only difference I found was clock and PCIe voltage were set to 700mV, as
stated in my first post.

I think I'll have to put the ebay motherboard back in and see if system is more stable.
All worked well for 3 years, so not sure if this is a problem due to PSU ageing or motherboard voltage
regulator drifting; something has altered, just cannot identify the problem at the moment.
 
Old 01-11-2014, 04:31 PM   #6
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,982

Rep: Reputation: 492Reputation: 492Reputation: 492Reputation: 492Reputation: 492
An important note:
Do NOT mess with the BIOS voltage settings. Leave everything on default. If it doesn't work that way, then you may have a major hardware issue.

I'm now quite positive that the PSU is the root cause. You tried two different motherboards with the same result.
 
Old 01-11-2014, 04:53 PM   #7
jailbait
LQ Guru
 
Registered: Feb 2003
Location: Virginia, USA
Distribution: Debian 12
Posts: 8,334

Rep: Reputation: 547Reputation: 547Reputation: 547Reputation: 547Reputation: 547Reputation: 547
Quote:
Originally Posted by hal8000b View Post
"The system has experienced boot failures because of overclocking or changes of voltages. The last settings on
this page may not coincide with current hardware settings".

.
This problem is much more likely to be a flaky power supply than a flaky motherboard voltage regulator.

-----------------------
Steve Stites
 
Old 01-14-2014, 07:11 AM   #8
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Original Poster
Rep: Reputation: 23
Update:
I again borrowed the exact same PSU from a friend and installed it on my computer.
So far results are stable... and as already stated by jailbait and metaschima a new
motherboard had same problems.

What has made this difficult to pinpoint is the sporadic behaviour of the fault,
once unplugged and left for a while the faulty power supply would work for a few hours
or few days.

Antec have 3 year warranty but unfortunately fall outside the warranty, will test for
another couple days and then mark this thread as solved.

Thanks to everyone for your help.
 
Old 01-14-2014, 09:32 AM   #9
cascade9
Senior Member
 
Registered: Mar 2011
Location: Brisneyland
Distribution: Debian, aptosid
Posts: 3,753

Rep: Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935
Quote:
Originally Posted by hal8000b View Post
Second problem is when switched off (not from sleep or standby) will turn on sometimes after 3 hours or 6 hours
later. The big clue is the message on BIOS:

"The system has experienced boot failures because of overclocking or changes of voltages. The last settings on
this page may not coincide with current hardware settings".


The BIOS is not overclocked but using default settings.
The above message is what appears when the machine has been in off state, so is it possible that the motherboard
voltage regulator is drifting when switched off? WOL and wake on ring are disabled in BIOS
I've got this msg with flaky motherboards/BIOSes many times before. You can also get this msg if you have CPU damage. Its unlikely to be anything to do with voltage regualators 'drifting' when shut down.

Quote:
Originally Posted by hal8000b View Post
I have since deliberately underclocked the CPU running at 2.60GHz and system shut down after 6 hours.
Have also tried fixing voltage core at 1.10V and still had system shutdown after 4 hours.
If you want to test your system, use the defaults. Chagning random settings in the hope it might help can lead to all sorts of issues, and make any problems worse.
 
Old 01-14-2014, 11:48 AM   #10
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Original Poster
Rep: Reputation: 23
Just when I thought all was well.....

System shut down again about 1 hour 1630GMT ago.
This was with alternate power supply. All worked well since about 2100GMT so was stable for about 19 1/2 hours.

I have a spare motherboard that I got from Ebay.
What is making this problem so difficult is the length of time between failures.

It could well be CPU, back in December I did try all my components ( CPU, memory PSU, and graphics card) on friends computer (same hardware) and motheroard, and all worked well for about 5 days so bought another GA-P55A-UD3 from Ebay.

Two days later with new motherboard in my machine PC shutdown suddenly. Comparing against my friends PC only anomaly was difference in clock voltage in BIOS settings. Once I changed this all
appeared stable, then I went back to original motherboard.
( I cant remember how long I tested new motherboard but think it was about 6 hours before going back to original)

I suppose its possible I have more than 1 fault.
Going to try my other motherboard from ebay and see how that works out.

Open to other suggestions though, thanks for support.
 
Old 01-14-2014, 11:56 AM   #11
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,982

Rep: Reputation: 492Reputation: 492Reputation: 492Reputation: 492Reputation: 492
So you don't hear any squealing / squeaking from any component ? Can you run it on the onboard graphics and see if it still does it, just to eliminate the graphics card ?
 
Old 01-15-2014, 02:50 AM   #12
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Original Poster
Rep: Reputation: 23
Quote:
Originally Posted by metaschima View Post
So you don't hear any squealing / squeaking from any component ? Can you run it on the onboard graphics and see if it still does it, just to eliminate the graphics card ?
No squealing/squeaking noises. When I first switch on from standby the noise could be described as a tiny squeak. but think this is fan just building up to speed.
Normal noise is just fan no other high pitched squeal.

Motherboard GA-P55A-UD3 has no onboard graphics, just slot 1156, Intel P55A chipset, onboard sound, SATA2/3 and USB2/3 support and onboard ethernet.

Situation now is I have an identical (working computer). I have put my PSU in the identical computer and all has been well overnight.

On my system the frequency of this fault is becoming more apparent. First time it ran 10 days, then 9 days then 6 days. Each time its failing faster. Yesterday failed in 19 1/2 hours. If something failed completely it is easier to fault.


What is the best way to test for instability? Run memtest or run prime 95
http://www.mersenne.org/freesoft/

or other tests?

One other important issue. Yesterday during the 19 hour period that systems was stable, I did hear 3 beeps from BIOS PC speaker whilst I was
browsing internet using KDE. Looking these codes up:

https://en.wikipedia.org/wiki/BIOS_b...IOS_beep_codes

This was post BIOS and does say base memory error. Previously I had tried my memory in the other computer but not for 9 days, only about 3 days.

Last edited by hal8000b; 01-15-2014 at 02:59 AM. Reason: Update
 
Old 01-15-2014, 09:31 AM   #13
cascade9
Senior Member
 
Registered: Mar 2011
Location: Brisneyland
Distribution: Debian, aptosid
Posts: 3,753

Rep: Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935Reputation: 935
Quote:
Originally Posted by hal8000b View Post
One other important issue. Yesterday during the 19 hour period that systems was stable, I did hear 3 beeps from BIOS PC speaker whilst I was
browsing internet using KDE. Looking these codes up:

https://en.wikipedia.org/wiki/BIOS_b...IOS_beep_codes

This was post BIOS and does say base memory error. Previously I had tried my memory in the other computer but not for 9 days, only about 3 days.
Umm.....a bit wrong on a few counts. The GA-P55A-UD3 uses an Award BIOS, not AMI-

http://www.gigabyte.com/products/pro...px?pid=3242#sp

The link you've supplied is to AMI BIOS POST beep codes (POST = 'Power On Self Test').-

POST codes are the beep(s) you hear when the machine boots. Hearing beeps while running can be caused by various issues, e.g. temp warnings, CPU fan failure alarm, applications, even too many keys on the keyboard being hit at once.....


Quote:
Originally Posted by hal8000b View Post
On my system the frequency of this fault is becoming more apparent. First time it ran 10 days, then 9 days then 6 days. Each time its failing faster. Yesterday failed in 19 1/2 hours. If something failed completely it is easier to fault.
Most of the systems I've seen with a "The system has experienced boot failures because of overclocking or changes of voltages..." msg have been ones I've bought to me for fixing. The one system I've had of my own where this problem occured it started very occasionally, then got more and more frequent until the motherboard finally gave up.

Yes, I know you've changed boards and used a different PSU. It could still be a board/PSU issue, but I'm thinking you may have CPU damage, or an overheating problem.

Quote:
Originally Posted by hal8000b View Post
What is the best way to test for instability? Run memtest or run prime 95
http://www.mersenne.org/freesoft/

or other tests
Its not a stability problem, its an intermittent fault.

Quote:
Originally Posted by hal8000b View Post
Motherboard GA-P55A-UD3 has no onboard graphics, just slot 1156, Intel P55A chipset, onboard sound, SATA2/3 and USB2/3 support and onboard ethernet.
Just for your info, dont use the SATAIII ports, even if you do get a SATAIII device. Its a marvell 9128 controler, and they are junk, even with windows.

Last edited by cascade9; 01-15-2014 at 09:33 AM.
 
Old 01-15-2014, 11:37 AM   #14
metaschima
Senior Member
 
Registered: Dec 2013
Distribution: Slackware
Posts: 1,982

Rep: Reputation: 492Reputation: 492Reputation: 492Reputation: 492Reputation: 492
Both memtest and prime95 are good for testing RAM and CPU and also system stability. Try running them and seeing if they provide any clues. Run prime95 in mode 1 to try and exclude the CPU.
 
Old 01-15-2014, 02:21 PM   #15
hal8000b
Member
 
Registered: Mar 2001
Location: UK
Distribution: Mint, Arch, Debian7
Posts: 194

Original Poster
Rep: Reputation: 23
Had a few late nights on this fault but have friends PC with exact same hardware so can swap components to prove the fault.

The problem now is how long to test for. The time between failures is not constant, although yesterday went 19 1/2 hours before system shutdown.

I've gone back to spare motherboard ftrom Ebay will test with default BIOS settings and see how things go.

Thanks for your help.
@Cascade 9 Will disable SATA III controller, I dont have any SATA III devices only SATA II hard drive connected to SATA II port, thanks for this advice.
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
GUI program to check cpu temperature, fun speed and voltage cccc Debian 5 06-30-2012 06:25 AM
Cpu voltage control babai Linux - Hardware 3 03-22-2010 09:25 PM
How to get the cpu core voltage without using lm_sensors ? gauthamk Linux - Software 10 02-27-2008 10:01 AM
setting CPU voltage with cpufreq/cpufreqd pafrica Linux - Laptop and Netbook 2 09-03-2005 03:04 AM

LinuxQuestions.org > Forums > Linux Forums > Linux - Hardware

All times are GMT -5. The time now is 04:09 PM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration