Stability Of S3 Stand By - Sleep On 5::6 W/ OC?
O.K. I admit I am getting better stability with my returned RMA. But I still get occassional failures to resume from S3 Sleep (Stabd By) on my 5:6 divider on a modest OC to 4750FSB and above (havent test daily mild 450FSB on 5:6 yet).More vdimm or vcore and a tad more NB did not matter. I went as high as 1.488v, (the 2nd Mobo in my sig @ 480x8) PLL: 1.648v, FSBT: 1.504v, NB: 1.54v, and CPU+NB GTL_Ref = 63x/67x respectively. Even tested RAM to 1.288v and still no go even though it passed with colors and temps were always good. Passed several hrs of all modes of Prime 95 v25.5 each time. Yet on a given day after many functions of S3, sometimes into S3 and back again, it just fails either in the morning after hrs of sleep or middle of the day when only in Sleep for few hrs.
Anybody find a butter zone for this? Note, S4 (Hibernate) always workes in any of the test. And S3 always works on a 1:1 even over 500FSB and with lower voltages all accross the board. So what gives?
Q6600 @ 4GHz in 1:1, DDR1000
Quote:
Originally Posted by
cyrixMII300
i need help everyone. could someone who has done 4.0GHz on air with a Q6600 please post your bios setting and also volts? I am at 3.6GHz now and cannot possibly go over even a MHz but i know that the baby can do it. so please could someone post your settings?
in the case that it has already been done, please let me know which thread and i will find it. thanks a gazzillion...
MAXIMUX FORMULA SE BIOS v0903 (new return)
[set as] = reported
------------------------
Extreme Tweaker Settings
Ai Overclock: [Manual]
OC From CPU Level Up: [Auto]
CPU Ratio Control: [Manial]
Ratio CMOS Setting: [8]
FSB Frequency: [500]
FSB Strap to North Bridge: [Auto]
PCI-E Frequency: [115]
DRAM Frequency: [Auto] (can try divider later if stable)
DRAM Command Rate: [2T]
DRAM Timing Control: [Manual]
Primary Info : 5-5-5-15-3-42-6-3 (try 4-4-4-13 @ 2.24v when stable)
Secondary Info : 8-3-5-4-6-4-6
DRAM Static Read Control: [Enabled] (helps with FSB on these core and mobo
Ai Clock Twister: [Light] (moderate if stable later)
Transaction Booster: [Disabled] (Auto later if stable)
Relax Level [0] 0-7 for more stability or Auto, if stable)
CPU Voltage ----------------- [1.58xxv] = 1.58v (some need 1.60v for stability @ 4GHz)
CPU PLL Voltage ------------- [1.54v] ==== 1.648v (may need upto 1.712v, not more then 1.79v) EDIE: I used 1.712 afterall)
North Bridge Voltage -------- [1.49v] ==== 1.520v
DRAM Voltage ---------------- [2.12v] === 2.224v (or whaterver you need)
FSB Termination Voltage ----- [1.44v] ==== 1.504v (shouldnt need more then 1.54v here)
South Bridge Voltage -------- [1.075v] === 1.104v (1.114 for RAID set stability)
Loadline Calibration -------- [Enabled]
CPU GTL Refference ---------- [0.63%] --
North Bridge GTL Reference -- [0.67%] --
DDR2 Channel A REF Voltage -- [Auto] ---
DDR2 Channel B REF Voltage -- [Auto] ---
DDR2 Controller REF Voltage - [Auto] =====
SB 1.5v Voltage ------------- [1.55v] ==== 1.600v
CPU + PCI-E Spread Spectrum --------- [Disabled]
This for example of 4GHz @ 1:1 for DDR1000, most stable for mine, but Quads are already hard to keep this high so try dividers at your desired risk. I did get some to hold. 450x9 as well, much easier but less RAM in 1:1 so try dividers here if you can get away with it.
Quote:
Originally Posted by
Ric2L
How are you guys keeping your N/B temps down on your non SE boards. If I try to run 400 fsb my N/B temps go up over 52 deg. while stressing with Prime95. From what I read at the start of this thread you should not let it go over 47? I have a thermal sensor on mine ran to the front panel. With a 3.2Ghz overclock on my Q6600 (356x9) it runs at 45 deg while stressing, and 40 idle. I'd like to try 400x8, or maybe 3.6ghz once my ultra 120 extreme gets here, but the N/B temps worry me. Don't want to fry this thing.
minimal voltages and more airflow for SE users on air. I found for SE on air, leaving barbs open but "stove piping" them to outside or one in line of fresh air and other in draft of exhaulst adds a tad of cooling effort, as well as good draft of air from cool sorse blowing over. The tiny fans just buckled directly onto it didntr seam to halp as much, and only made more noise.