Results 1 to 9 of 9

Thread: Computer behaves strangely when cold. Please help :(

  1. #1
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183

    Angry Computer behaves strangely when cold. Please help :(

    Hi people,

    You may have noticed in this thread http://www.xtremesystems.org/forums/...d.php?t=254657
    that I was having lots of trouble overclocking.

    I will try and explain this and keep it short.
    Basically my settings that I found stable are the following:

    285x13 = 3705MHz
    CPU-NB ratio 9 = 2565MHz
    Ram ratio 1:2.66 = 1520 (all timings on auto)
    HT link ratio 7 = 1995MHz
    All voltage auto except:
    CPU voltage = 1.3V (shows as 1.39 in CPUID hardware monitor & in bios)
    CPU-NB voltage = 1.25V
    DRAM voltage = 1.621(shows as 1.65V in CPUID monitro)

    I tested these settings on OCCT for 7 hours on CPU-Linpack and 7 hours CPU-OCCT. Linpack was set to 90% mem usage 64bit and the CPU mode was on medium data set with normal priority.

    This morning I turned on my computer and while trying to boot windows I got a BSOD, Bad Pool Header.
    I then returned all my bios settings to stock and restarted. The computer would get to just after the windows 7 loading screen and reboot with no error.

    I then disabled C1E in the bios and windows shows my desktop for a second before rebooting. I re-enabled C1E and rebooted. Same problem, just see desktop for a second.
    I rebooted again and then it went into windows with no issues.

    I found it also restarts in a similar fashion with Ubuntu, but once is has gone into windows once, no problem.

    The C1E has given me strange issues when running stock settings and I have always disabled it when overclocking.

    Also as soon as I turn my RAM voltage to 1.651 in the bios the computer fails Linpack after about 3-30mins. In Hardware monitor its reported as 1.7V when set to 1.651 in bios.

    Im not sure if these errors are do to the computer being cold. Its about 12-15 degrees celcius in the mornings.

    Any ideas? Im tempted to give my supplier my whole system and let them check it out, but their building is probably heated and may not find any errors.



    MSI 890FXA-GD70 Bios 1.6
    AMD Phenom II X6 1055T
    Apacer Giant II 2x 2GB DDR3 240-Pin DIMM, 1.65V [PC17600] 10-10-10-30-2T
    MSI 8800GT OC 512Mb
    Corsair H50 in push pull config
    CM Storm case

  2. #2
    Xtreme Member
    Join Date
    Aug 2009
    Location
    Perth, Western Australia
    Posts
    123
    Tried manually setting the ram timings (the main ones) to what your ram is rated to?

    I had some issues with the Kingston Hyperx's when running higher than rated (10-10-10-28) timings and would have lot less issues running it at it's rated timings.

    Although i did have a dud set of hyper's....
    PII x6 1090T @ 4.03ghz @ 37C Load | MSI 890FXA-GD70 | Asus EAH5870VT | Corsair Dominator GT DDR3 1600C7 | Western Digital Raptor 36gb x2 RAID 0 | Coolermaster HAF-932 | Swiftech MCP650, EK Supreme-HF, MCPRES-4 and 1/2" tubing

  3. #3
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183
    That's the thing, I highly doubt it will pass so many hours of OCCT with problematic timings.
    I have a feeling it's the RAM, or perhaps some bad soldering on the motherboard.
    I am really out of ideas for testing this thing now.

  4. #4
    Xtreme Member
    Join Date
    Apr 2010
    Location
    In a world beyong borders
    Posts
    221
    Hi.

    Can u reset the BIOS? and test the RAM in other RAM slots.
    SAINT19

    Two things are infinite: the universe and human stupidity; and I'm not sure about the universe.". Albert Einstein.
    Max overclock archived, 1090T @ 6.5GHz


    Phenom II X6 1090T BE @ 3.8GHz and NB @ 3000MHz both with 1.325V on BIOS
    Gigabyte 890FXA-UD5 Rev. 2.0 with F4 BIOS
    Crucial Ballistix Tracer 4GB (2x2GB) DDR3 1600 8-8-8-21-1T
    MSI GTX 680 Lightning 2GB
    Corsair Force GT 120GB SSD + Hitachi 2x500GB HDD
    EK Supreme HF Full Nickel + MCP655-D5 + MCR320-QP

  5. #5
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183
    I have just reset the bios and am going to try test one stick at a time on memtest.
    I tested this ram last week, both sticks in for 10 passes of memtest without errors. The previous day I had received something around 6000errors by 17%.
    So it's a bit intermittent...

  6. #6
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183
    Just ran memtest on each module for 1 pass quickly. No errors. I am going to leave everything stock for a couple of days and see how it behaves...
    There is one or two other people who have had similar problems with different components, but no solutions.
    The rebooting with no errors as it enters windows it a bit weird. Does the same with ubuntu which sort of indicates a hardware fault.

  7. #7
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183
    Ok, just to let you guys know.
    I returned all the components and they found one or both of the RAM slots closest to the CPU to be "Loose".
    I'm not sure if this will solve the voltage setting problem, but I will be getting it back today.

  8. #8
    Xtreme Member
    Join Date
    Apr 2010
    Location
    In a world beyong borders
    Posts
    221
    That're good news that should be the source of ur problem. Let me know how the RMA works.
    SAINT19

    Two things are infinite: the universe and human stupidity; and I'm not sure about the universe.". Albert Einstein.
    Max overclock archived, 1090T @ 6.5GHz


    Phenom II X6 1090T BE @ 3.8GHz and NB @ 3000MHz both with 1.325V on BIOS
    Gigabyte 890FXA-UD5 Rev. 2.0 with F4 BIOS
    Crucial Ballistix Tracer 4GB (2x2GB) DDR3 1600 8-8-8-21-1T
    MSI GTX 680 Lightning 2GB
    Corsair Force GT 120GB SSD + Hitachi 2x500GB HDD
    EK Supreme HF Full Nickel + MCP655-D5 + MCR320-QP

  9. #9
    Xtreme Member
    Join Date
    Jun 2008
    Location
    South Africa
    Posts
    183
    The new board seems to be working fine, it's actually much much faster.
    Even when it was passing memtest, the benchmark in everest was kinda low, but now this thing is flying.
    Gave a bash at 4GHz last night, wasn't really stable at 1.45V so I gave up.
    But I plan to run the same settings as my original post just with the NB ratio higher.
    I will perhaps push it a bit this weekend and see if my mobo blows up like everyone elses...

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •