Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 2 of 7 FirstFirst 1234 ... LastLast
Results 11 to 20 of 61

Thread: GA-X48-DQ6 Won't OC Anymore




  1. #11
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Sorry! Leave alone, on one of the above settings, Auto or 1-31 or 0.

    It makes the static tRead Value minus what you set this at certain times, so messes with things really.. Just leave unset, I put the 3 choices because some get different choices if you have already changed it once before

  2. #12
    Join Date
    May 2008
    Posts
    22

    Default Re: GA-X48-DQ6 Won't OC Anymore

    OK, memtest86+ v. 2.01 is complete on the Corsair 3 GB of DDR2-800 RAM. Test result---100% pass with 0 errors . It's cool to watch it test. So that's what all of those letter combinations are. Memory addresses! Also, while running Prime95 and core temps at 38c, I can keep my finger on the NB fins and top plate INDEFINITELY. As in all day long. Remember I have the Antec 900 and I installed the side fan that blows across the video card (9800 GTX). The air blast also hits the NB heatsink head on. Unless I'm supposed to put my finger somewhere else...

    On to the Subtimings. When you say change just one Subtiming at-a-time, then test, do you mean change just one, and then setup the oc to see if it will boot successfully? Or do you mean test for booting the change without an oc setup? I hope I'm making sense.

    Here is what the subtimings are set to RED FONT (just left of the [Auto] brackets). To the right of the RED FONT are your suggestions.

    4-4-4-12 of course, then >>
    ACT to ACT Delay (tRRD)______3 3-4
    Rank Write to READ Delay______3 4 Or 9-11
    Write to Precharge Delay______6 3-6
    Refresh to ACT Delay__________4222-38/42 (Start with 42)
    Read to Precharge Delay_______3 3-5
    Static tRead Value____________6 7-10 (Do not try 6 no matter what you read!!!)
    Static tRead Phase Adjust______0 Just leave it alone with what it reads!

    What do think about starting with "Static tRead Value" set to 10?
    Last edited by michaelboss; 05-10-2008 at 02:31 PM. Reason: correction + additional comment

  3. #13
    Join Date
    May 2008
    Posts
    22

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Maybe this will help a lot. It is a screen shot from Everest Ultimate of the memory timings readout. It is a zip file. Maybe you will see an obvious culprit to modify.

    Everest Mem Timings Screen Shot.bmp.zip

  4. #14
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Nope, sounds like your NB is good to go!

    I mean, change them all to the higher settings, and fully test the ram for stability. Once you are sure it is stable, go in and lower ONE setting and then do the memory tests again, when you are sure that one setting change was stable, go in and lower the next one.

    So you know, just to the left of AUTO is NOT what AUTO sets things at, it is what is running in the BIOS, JUST AT THAT TIME. IT changes OFTEN and is not nothing to go by or even look at.

    I stand by what I suggested, just start with those settings, nothing on auto except the last one.

    I know it takes some time to change one thing at a time and test each time, but that is the better way to do things, because if you change 2 or 3 then it fails you are unsure which caused what.

    You can Likely start Static tRead (Which is Performance Level and a major speed defining setting) on 7, if it dont boot set to 8.

  5. #15
    Join Date
    May 2008
    Location
    Canada
    Posts
    5

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Quote Originally Posted by michaelboss View Post
    My new build has been running for 2 weeks now. Has an E8500 CPU. The board accepted a FSB 400 MHz oc just fine. CPUZ verified 400 MHz FSB and 9.5 multiplier to reach 3800 MHz. Prime95 ran for 2 hrs with no errors. All temps were normal (cpu max was 43 c). Then I noticed Everest Ultimate + CPUZ showed the oc was gone! I have tried everything to get back the oc, but the board just keeps restarting a couple of times, then loads the original settings (except it keeps my ram timings). Read some of the other similar threads, but nothing is working. Help-I'm lost!

    Here are some sys specs and attemted BIOS info:
    MIT
    CPU Clock Ratio 9 X
    Fine CPU Clock Ratio +0.5
    CPU Host Clock Control Enabled
    CPU Host Frequency 400
    PCI Express Frequency 100
    C.I.A.2 Disabled
    Performance Enhance Standard
    System Memory Multiplier 2.00D
    DRAM Time Selectable Manual
    CAS Latency Time 4
    DRAM RAS# to CAS# Delay 4
    DRAM RAS# Precharge 4
    Precharge delay(tRAS) 12
    Advanced Timing Controls All set to Auto
    CPU/PCIEX Clock Driving Control 800mv
    CPU Clock Skew Control Normal
    (G)MCH Clock Skew Control Normal
    System Voltage Control Manual
    DDR2 OverVoltage Control +0.40V
    PCI-E OverVoltage Control +0.10V, also none
    FSB OverVoltage Control +0.10V, also none
    (G)MCH OverVoltage Control +0.100, also none
    Loadline Calibration Enable
    CPU Voltage Control 1.3125V, as well as default volts
    board revision 1.1, F6 BIOS

    E8500 cpu + Zalman 9700nt cooler + arctic silver 5
    Enermax Infinity 720 wt psu
    Antec 900 case with the side fan installed (lots of air flow!)
    Crucial Ballistix PC-6400 3 GB (2x 1GB + 2x 512 MB) running @ 4-4-4-12 @ 2.2v (yes it's on the compatible memory list)

    Michaelboss
    This also happened to me with my P35- DS4 .. Pushed my E6850 to 4ghz locked up on orthos , when it rebooted , locked at 3ghz even when i loaded presets for OC .. ie : 3.6 or 3.8 ghz would aways boot 3.0 ghz
    To rectify either reflash bios or clear cmos ( short jumper pins ) easier way to do it ..
    Hope this helps :)

  6. #16
    Join Date
    May 2008
    Posts
    22

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Well guys, I think I am going to abandon my attempt at a traditional overclock, i.e. by increasing the FSB, and then adjusting the memory divider to make things match up. The absolute mind blower is that the very first attempt worked and CPU-Z verified it.

    I have become very frustrated because literally everything I have tried has failed. I was able to get all of Lsdmeasap's memory subtimings to boot. I followed your plan by taking all of them off of auto and setting them all to their highest suggested setting. Then, one-by-one lowering them all to their lowest setting. Everything worked. But then no matter how little or how much I increased the FSB, it would never boot without those two or three restarts, then it would boot at the FSB and CPU speed defaults. Get this, I even tried lowering the CPU multiplier, but increasing the FSB, thereby just achieving the exact same default final CPU speed, and even that would not boot. I even tried lowering the CPU speed, which is underclocking, and that would not work. ANY CHANGE to the FSB would not work.

    I tried clearing the CMOS as kissparo suggested. I found that it cleared some settings, but not all. Things like USB keyboard and USB mouse support settings were not affected. Then I tried the same sequence of oc attempts after clearing the CMOS, and none of those worked either!

    BUT THEN VIOLA, I read in the mobo manual about C.I.A.2 (CPU Intelligent Accelerator 2). It's supposed to adjust the FSB according to the CPU load and has 5 presets. I chose "Sports" -and what do you know- CPU-Z showed the FSB increase and corresponding CPU speed increase. CPU-Z also showed the memory correctly as well. I ran Prime95 for 30 minutes without a glitch. Take the laod away by stopping Prime95 and CPU-Z shows everything at the defaults.

    So I then went all out and changed C.I.A.2 to the highest setting of "Full Thrust". Prime95 showed a hardware error on the 4th test on both cores. I reset C.I.A.2 to the next highest "Turbo" and ran Prime95 again. Once more on the 4th test there was an error. This time I went into MIT and raised the timings up one notch that Lsdmeasap gave me, as well as raising the voltage of the RAM to +0.30 (I know it's supposed to be 2.2v total, but he convinced me that a little less was safer and would work--He was right!). This time Prime95 ran without an error for 1-1/2 hours. Everest shows a 19% increase in CPU clock. CPU-Z had a 397 FSB and CPU speed of 3771.3 MHz. Memory had 476.3 MHz DRAM clock with the correct 5:6 ratio. Turn off Prime95 and all values fall to their defaults. Turn on Prim95 and all values overclock by 19%!!! Etc.

    If you think about it, isn't this the ideal situation? It's like not oc'ing and using C1E (CPU Enhanced Halt), but in reverse--you're normal until you need a big boost, and then you get it. When you don't need the boost any more, you return to normal.

    Maybe C.I.A.2 gives some memory and CPU's trouble, but not mine. I am very happy now. I can't wait for the OCZ 1066 RAM to arrive on Monday. I might be able to get this puppy going on "Full Thrust" close to 4 GHz and the RAM handle it as well.

    The down side is not knowing why the manual oc of the FSB would not take after the original first try did. But, all has ended well...so far.

    Thanks Lsdmeasap very much for your help. I am closer to geek status thanks to you and am no longer afraid of the dreaded subtimings zone. Also thanks for the memtest link and teaching me about burning an ISO image so I could use it. When the new RAM gets here, the very first thing I will do is test it. Thanks again...

    michaelboss

  7. #17
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Well ya know, you should have left all the memory subtimings high, then overclock your CPU to what you wanted. THEN bring down the memory timings after you have your CPU where you wanted it. And of course to go very high with 1:1 on the memory you would need 555 for the main timings most likely, but yeah you need the subtiming loose while you up the FSB until you get it to where you want. Then do the bring down one at a time and test on the memory

    And you know most 6400 ram will do 1066 just with loose timings such as 555. And 1066 ram as you will see in CPU-z is really 6400 ram anyway, just tested to run at 555 timings officially

    So, you got memtest ISO burnt then eh? Thats cool!


    Well I am glad to see you are getting somewhere with my suggestions, if you feel you have the time, turn off that C.I.A.2. And EIST, TM2, And C1E and do the memory subtimings back up to the upper area and then do the FSB increases. With you lowering the memory subtimings first to a lowest point then increasing the FSB that is why it fails. The memory subtimings were good LOW testing them because you had the FSB set low, you went in the wrong order so that is why it failed is all


    Turn off all those features, Raise Memory timings and subtimings, raise FSB to desired level until unstable ten bring back down until it is at max stable point. THEN start with the lowering and testing of the ram.

    Good luck, and of course post back if you need some more help! Good to see you getting somewhere and not afraid to jump in and try things. It is all pretty easy to find the right settings, it just takes so long testing and trying things that people make it sound harder then it really is. Just start with everything loose on the ram, raise FSB until to high, then slowly lower and test everything.

    All time consuming is the bad thing, but in the end it will pay off.
    Last edited by Lsdmeasap; 12-01-2008 at 01:04 PM.

  8. #18
    Join Date
    May 2008
    Posts
    22

    Default Re: GA-X48-DQ6 Won't OC Anymore

    OK, I'll go through it again. I did not have TM2 off, although I had thought of it. I had tried the FSB oc with all of the sumtimings on their high settings. What I meant above was I verified they would all work on lowest without changing the FSB, but failed to mention that I had returned them to their high settings to begin the oc.

    Well so far the Corsair has passed memtest. The way the Corsair went to 476.3 MHz from 400 MHz in "Turbo" mode just proves what you just said-that it will oc up in MHz. But, mine did it to 476.3 MHz at the 4-4-4-12 timings. But I'm going to put the OCZ in anyway.

    I guess I also failed to mention that when I tried to underclock, the subtimings were in the high setting as well. The underclock failed. Lsdmeasap, it seems that any change to the FSB fails, unless the C.I.A.2 does it.

    But I will slow down even more and try again. You're right, it takes a million years and that's why a lot of people give up or don't want to try. Everest has the number of my startups at quite a bit more than double from the time I started trying to oc.

    I'll get back to you.

    OK, I'm Back. Lsdmeasap, I'm telling you the facts, once again absolutely nothing works. Memory at 5-5-5-18 with all the subtimings at their highest settings, all those features are off, all voltages are where you said. Again it appears to me that the FSB simply will not accept a manual change of any kind-up or down. Well, what is your prognosis doctor?

    I take it that you've never encountered this situation before.
    Last edited by michaelboss; 05-12-2008 at 03:42 AM. Reason: additional content

  9. #19
    Lsdmeasap's Avatar
    Lsdmeasap is offline GIGABYTE Guru
    Join Date
    Jan 2008
    Location
    PCB Island
    Posts
    25,940

    Default Re: GA-X48-DQ6 Won't OC Anymore

    Ahh ok! Well ya got me confused them!

    Yes, I have seen this before. A few other things in the bios to check before my last sugestion. Do you have USB mouse and Keyboard support enabled? IF so, disable those, even if you use them they will work fine anyway. Also find and disable Legacy USB storage Detect

    Then try again

    If that does not help, download the latest NON beta BIOS for your board, put it on your floppy or USB and get ready to flash. But first find a Hard drive Jumper, open your case and put a jumper on the Clear CMos Pins and leave it there. Then Unplug your Power Supply from the wall and let your computer sit for 10 minutes, with PSU unplugged and Clear CMOS pins with the jumper on them. Then after 10 Minutes remove the jumper and plug your PSU back in and start the computer up and enter Qflash. Once there flash the BIOS to the newest BIOS you just downloaded.

    Then reboot and go into the bios and Choose "Load Optimized Defaults" Which is also F7, then save and apply and reboot F10, then come back into the BIOS and make your initial changes for hard drives and all the rest of the things you use or dont use. Then.....yeap start your tests again.

    Most people do not load optimized defaults and save and apply/reboot when they update the bios, this causes errors and partial problems from the previous bios to still be in the CMOS memory.

    I suggest my way of clearing the CMOS because I have done all the others and sometimes they work and sometimes not, That way is FOR SURE, and easier as you do not take out the battery or anything

  10. #20
    Join Date
    May 2008
    Posts
    22

    Default Re: GA-X48-DQ6 Won't OC Anymore

    USB and Keyboard support as well as Legacy USB Storage Detect were disabled for the last 30-40 attempts. There is no other NON Beta BIOS. My Ga-X48-DQ6 is a fairly new board and the F6 BIOS is the original BIOS. At their website there is a Beta F7b that is somewhat recent, but it is a Beta (F7b 2008/04/03 beta BIOS update CPU ID To support the same CPU but with new stepping).

Page 2 of 7 FirstFirst 1234 ... LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

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