E6600 and D975XBX2 overclock ... can anyone share their settings?

[slappynutz]

After getting loads of help identifying the correct Bad Axe 2 bios settings I needed to disable in order to overclock, I've been running my E6600 completely stable at 3.2ghz for a couple weeks without tweaking the voltages at all. Pretty happy with the performance setting the frequency at 356mhz.

But I've been getting the urge to push it further and was just wondering what people using the same configuration have their voltage settings at on the Bad Axe 2.

At the default bios setting of 1.3250V I get about 1.27V measured in Vista with Speedfan. I've tried bumping the voltage up to 1.38 but have been unable to get it stable at even 366mhz (for 3.3ghz overall). Kind of uncomfortable cranking the Vcore up all the way to 1.4 and figure the problem might be the FSB voltage (which is also default). Since it's stable at 1.27 measured in Vista, I find it hard to believe I'd need to crank the Vcore by 0.7V just to eke out another 100mhz.

I've got my memory running at stock settings, 4-4-4-15 at 2.04V in the bios without any issues. The northbridge does tend to run a bit hot at 47c idle, 51c load (despite the CPU temp being much lower) and so I bought an Antec Spot Cool that I'll aim directly at the NB heatsink later today.

Just wondering if anyone with the same board/chip wouldn't mind sharing their voltage settings so I can compare. Thanks a lot.

 

[Xilikon]

As a fellow BX2 owner, here are some things :

-If you push the voltage and cannot get stable at 3.3 GHz, it could be the heat causing issues. Try to check with CoreTemp and TAT to make sure it stay below 65C.

-It could also be because you are using the wrong memory settings, causing the RAM to be overclocked. Use http://www.peakin.com/xbx2/calculator.html to get the most optimal settings. As example, if you use DDR2-800, make sure you are using 1:1 ratio so set the reference memory frequency at 200 and memory frequency at 400. Use cpuz, in the memory tab to make sure it is close or below 800 (400).

-Bump up the vMCH and vFSB one notch or two to cope with the extra stress overclocking bring.

Mine was stable all the way up to 3.3 GHz and I didn't go higher since temperature get close to 70-72C, which is a bit too much for my taste. I did those things as well, using cheap shit DDR2-400 RAM.

 

[slappynutz]

Thanks for the suggestions.

The memory is OCZ 6400 Platinum Rev2 (4 sticks) which runs up to 400mhz rated and I've set the memory frequency at 533 and the reference frequency at 266, which is 1:1 for a DDR2 equivalent of 357, well under the spec. At 2.04V and 4-4-4-15 I'm not thinking that I've got a ram problem, but I could be wrong.

When I set the overclock to 3.3, up from 3.2, it will boot into windows and fail OCCT at about the 15 minute mark regardless of whether I leave the Vcore at 1.3250 in Bios or bump it up to 1.3850. So I tried bumping the vMCH and vFSB up two notches each ... regardless, OCCT fails every time at 3.3ghz.

Temps don't seem to be an issue. I have a Scythe Ninja, now running with the supplied 120mm fan, surrounded by two 120mm exhaust fans at the top of my Antec P180. Leaving Speedfan open during stress testing my CPU temp never goes over 48c load (in a warm apartment), but the Motherboard (northbridge I'm guessing) temp spikes up to 52c. These temps are exactly the same in testing when I run the E6600 at 3.0, 3.15, 3.2 and 3.3.

 

[Xilikon]

Then, try to bump to 1.4v or even 1.425v since it's at this voltage that my 3.3 oc is tested stable.

My piece of advice is to first try to bump the voltage high enough to get you stable at everything. When you foudn the stable OC, bring down the voltage a few notches, testing each time till it become unstable then back up one or two notch then you got the stable oc settings

 

[slappynutz]

I'll try that again tonight. I remember upping my voltage in the Bios then not seeing it move from 1.27/1.28 measured in Vista. Will try to push everything more once I get the extra fans in there.