Tweaking NB & FSB voltages. IP-35pro.

northun_munky

northun_munky

northun_munky

Associate
Joined
1 Mar 2008
Posts
955
Location
Near Longleat, Raaaar!!
Hi Guys,

I've got my q6600 running at 3.4 (426x8, RAM @ 511) prime95 stable overnight by using 1.35v on the CPU under load without having to change either the CPU VTT or the MCH voltages. I'm using 2x2g sticks of ram.

I'm trying to hit 3.6 unsuccesfully with random BSOD under P95 and have increased the vcore to 1.46 (under load) but haven't yet touched NB or FSB v's.

My Question is, in anyones experience, would it help stabalise my overclock by tweaking them? Or should I leave them be an just increase the vcore? I'd really like to hit 3.6 if I can.

Cheers
 
I found with my ip35pro board I had to increase the nb voltage and VTT to get it stable at high overclocks (got mine up to 3.8Ghz)

But you have to keep an eye on the nb and voltage regulator temps though (everest or similar will do this) as if they get high then it can make the overclock unstable.

Also I assume as you are increasing your fsb speed to up your cpu you are remembering to drop down your ram speed? Always best to put the ram on the lowest ratio and underclock it when trying to find you highest cpu overclock. Then you can muck about with your ram settings later.
 
It's pretty certain that you'll need to up the chipset and VTT voltages if trying to run a quad at 450mhz FSB on a P35 board.


By just how much is pretty dependant on your own chip and board. But they'll likely need to go quite high to be stable at 450mhz.
 
Thanks for the advice cob. I've upped the NB & VTT and arn't getting any BSOD. Just one core that seems to fail the 8k tests in P95. I'll up the vcore a bit more and see how I get on. From what I've read the q6600 should be ok for up to 1.55v (under load) Not that i'd go that far but does that sound about right?
 
See, this is where I get confused with the voltages. Theres such a discepancy with idle and load voltages and vdroop on my board that when people say 'I run at 1.55 vcore' I get confuddled... I'd hate to put 1.5v through under load if people mean 1.5v at idle:rolleyes:
 
This is what I need for 3.6 on mine...

Vcore of 1.52 bios (1.42 - 1.43 under load according to uGuru)

VTT at 1.23

MCH at 1.29

the others are default.

Just keep an eye on the PWM temps. I need to run a little 60mm fan above mine to keep them below 70c at load. They can get very toasty, and WILL affect your overclock.
 
No it was under water. Also consider the bolt mod for the mobo. From memory I found it dropped my nb and pwm temps by 10c allowing me to reach 3.8Ghz. Before that it was holding me back despite the watercooling.
 
Ok, so with a bit of trial and error (BSOD's;)) I'm at a point where I run P95 for 30 mins with no further BSOD or cores failing the torture tests. I know its a while away from stable but its definately progress.

I used the following settings

FSB 400*9
vcore 1.5325 Bios - 1.5 idle - 1.44-46 under load
CPU VTT 1.27
MCH 1.44

The reason I stopped P95 is the PWM temps are 50 idle and after 30mins of 8k fft's hit 78. :(

Will these sorts of temps be ok to continue testing with? If they are and it turns out 3.6 is stable then i'd look to do a bolt mod and active cooling for the VRM's for everyday use. Whats the highest I'm safe to test with?
 
I think (at least the old 680i based boards were) these are rated to over 100c, although it's not a great idea to be that hot. You should be fine. As you say, look to do the mod and actively coold them.
 
My settings for 3.6 GHz - my PC is stable through 5 cycles of Intel Burn Test (IBT) which needed a small boost of Vcore from my P95 settings!

Q6600_Settings_3_6.jpg
 
You must log in or register to reply here.
Back
Top Bottom