Overclocking P5Q Deluxe / Q6600

HummuH1 · 12 Sep 2011 at 10:27

Guys, just dropped my Q6600 into a P5Q Deluxe.

Got it up to 3.4 by simply adjusting FSB and upping to 1.35v. Haven't fully tested stability yet but all seems well so far.

Ther Q6600 is a G0 stepping with a low VID, 1.235v IIRC.

I'm happy with the 3.4 but wondering if it's capable of more? It's being cooled by a Tuniq 120 which I salvaged from my old 775 system. Played a few games and max temps were about 47 degrees although I did notice a difference of about 9 degrees across the 4 cores (lowest temp to highest temp whilst under load). I seem to recall this is normal, is the difference OK?

Any tips on pushing the OC a bit further? There's so many more BIOS options on this board compareed to my old Abit I'm a bit lost if truth be told!

Any help appreciated.

Cool Breeze · 12 Sep 2011 at 13:11

With a Vid that low you should hit 3.6 at about 1.4 volts easy, I don't know how good the Tuniq 120 is but if you keep your load temps when stressing below 70 then your golden.

I have my 1.2875 vid Q6600 @ 1.42 and 3.6 Ghz under a megaharlems.

the best combo for 3.6 should be 400x9 on the FSB but it's a trial and error job small increments and then test for stability.

Have fun

HummuH1 · 12 Sep 2011 at 14:05

Cool Breeze said:
With a Vid that low you should hit 3.6 at about 1.4 volts easy, I don't know how good the Tuniq 120 is but if you keep your load temps when stressing below 70 then your golden.

I have my 1.2875 vid Q6600 @ 1.42 and 3.6 Ghz under a megaharlems.

the best combo for 3.6 should be 400x9 on the FSB but it's a trial and error job small increments and then test for stability.

Have fun

Thanks. 400 FSB should hopefully work well with 800 MHz RAM.

Any thought on the temos across the CPU? I'm slightly concerned the cooler might not be seated 100% even. Is there any way on ensuring this with the screw-in pins? I have to judge each of the pins by how much thread is showing so not an exact science! Should there be any sort of "click" when the pins have been screwed in enough?

Cheers.

j.col · 12 Sep 2011 at 14:07

i used to have a tuniq tower 120 on my Q6600, and i had upto 10.C difference between the cores.
its normal,
i used to turn the screw until it got hard, iirc i had no thread showing, have you used the springs on the bolts?

Cool Breeze · 12 Sep 2011 at 14:10

As j.col says difference accross cores is normal mine are about 5oC apart and my cooler is mounted just fine. It's because the Q6600 is basically 2 dual cores slapped on one PCB so they will differ as they are not in the same package if you will.

HummuH1 · 12 Sep 2011 at 15:11

j.col said:
i used to have a tuniq tower 120 on my Q6600, and i had upto 10.C difference between the cores.
its normal,
i used to turn the screw until it got hard, iirc i had no thread showing, have you used the springs on the bolts?

Yep, springs are on. Got them screwed in until no thread showing so hopefully OK

Cool Breeze said:
As j.col says difference accross cores is normal mine are about 5oC apart and my cooler is mounted just fine. It's because the Q6600 is basically 2 dual cores slapped on one PCB so they will differ as they are not in the same package if you will.

Thanks. I had about 7 degrees diff in last PC just didn't know if the slightly larger diff was due to poorly installed cooler or not. Will keep an eye on them when stressing.

Cheers guys, hopefully 3.6 on the way!!

C64 · 12 Sep 2011 at 15:24

1.45 volts might be needed for 3.6 there tends to be a big voltage bump required from 3.5 - 3.6ghz and the tuniq is capable of doing that 24/7 the IBT temps might get a bit high but I wouldn't worry these chips are indestructible as long as gaming load is -60c you are good to go.

I would go straight in at 1.4 volts and test 401x9 if it's IBT 50 passes stable then start lowering the voltage a couple of notches at a time til it fails then bung it back up 2 notches where you know it's stable, LLC enabled and leave everything else on auto northbridge bump only needed if running 4 sticks of ram.

HummuH1 · 12 Sep 2011 at 17:13

C64 said:
1.45 volts might be needed for 3.6 there tends to be a big voltage bump required from 3.5 - 3.6ghz and the tuniq is capable of doing that 24/7 the IBT temps might get a bit high but I wouldn't worry these chips are indestructible as long as gaming load is -60c you are good to go.

I would go straight in at 1.4 volts and test 401x9 if it's IBT 50 passes stable then start lowering the voltage a couple of notches at a time til it fails then bung it back up 2 notches where you know it's stable, LLC enabled and leave everything else on auto northbridge bump only needed if running 4 sticks of ram.

Thanks. Will give it a bash once all service packs and updates are installed.

Was the 401 a typo rather than 400? If not, why add the extra 1Mhz? Just curious.

Cheers.

C64 · 12 Sep 2011 at 18:16

well on mine 400x9 = 3.599 ghz and I wanted 3.6 it's 1 louder

setter · 12 Sep 2011 at 18:27

My old 1.2750 vid q6600 hit 3.6ghz on 1.400v on a p5q deluxe, 3.8 at 1.512v iirc, fsb - nb strap at 333mhz, 1.58 cpu pll, 1.34v on nb and fsb termination, (vtt). The voltage jumps do get very signifigant above 400 fsb, keeping the cpu cool is the main problem, i had to lap both the cpu and the TRUE cooler, and use dual 2000 rpm fans on a controller when priming it.

HummuH1 · 12 Sep 2011 at 19:33

C64 said:
well on mine 400x9 = 3.599 ghz and I wanted 3.6 it's 1 louder

LoL, that makes sense now.

fowler002 · 12 Sep 2011 at 20:46

mine's vid is 1.3125v

currently on 3ghz(1.35v I believe, cpu id is say 1.344v but think it shows lower than it is) and wasn't prime stable above that tho I never really spent time trying to push it.

will be building my i2500k system once I decide on the ram, not sure if I should try pushing the q6600 first before replacing it.

HummuH1 · 13 Sep 2011 at 08:51

OK. Managing to get 3.4 Ghz stable but 3.6 is proving a challenge.

In terms of 3.4, I can get there by using 9 x 377/8 at 1.35v, max temps seem to be about 62 degrees when stressing. Is there any other multi / FSB option that might be better when using 800Mhz RAM?

For 3.6, I tried 9 x 400 at 1.425v and it fails on the 2nd pass of IBT with temps hitting 72 degrees. Did not change any other BIOS settings.

I'm happy at 3.4 but interested to see if I can get it stable at 3.6 at sensible temps, I think that is what my hold me back.

C64 · 13 Sep 2011 at 20:12

That simply means you need more voltage for 3.6 don't be scared people have ran these for years with 1.5v + pumped through them.
Just set the memory divider to match your rams speed as close to 800mhz as it gets will probably see 802 mhz or whatever and try slightly looser timings if you are running very tight timings, I ran my ram @ 5-5-5-18.Every motherboard can be different too try upping the nbridge voltage 2 notches and the PLL up 2 notches from auto too.

I would try 1.45 volts and if that doesn't work try 1.46 this is what mine needed for 3.6ghz.
The below pic the actual vcore set in bios was 1.46 volts.

ZeroUK · 13 Sep 2011 at 20:45

Think yourselves lucky, had to go to 1.5v for 3.4ghz

setter · 13 Sep 2011 at 21:07

Yep, just luck of the draw, ss i managed to find of mine, prime95 as LinX/IBT werent out at the time.

Ram at the time was ocz reaper, 800mhz, 4-4-4-12, slackened the timings to 5-5-5-15, later replaced it with reaper pc8500 1066mhz. Pll, nb and fsb termination needed a bit of tweaking to get it stable, ran the cpu at that speed up until i replaced it with a q9550.

HummuH1 · 13 Sep 2011 at 21:42

C64 said:
That simply means you need more voltage for 3.6 don't be scared people have ran these for years with 1.5v + pumped through them.
Just set the memory divider to match your rams speed as close to 800mhz as it gets will probably see 802 mhz or whatever and try slightly looser timings if you are running very tight timings, I ran my ram @ 5-5-5-18.Every motherboard can be different too try upping the nbridge voltage 2 notches and the PLL up 2 notches from auto too.

I would try 1.45 volts and if that doesn't work try 1.46 this is what mine needed for 3.6ghz.
The below pic the actual vcore set in bios was 1.46 volts.

setter said:
Yep, just luck of the draw, ss i managed to find of mine, prime95 as LinX/IBT werent out at the time.

Ram at the time was ocz reaper, 800mhz, 4-4-4-12, slackened the timings to 5-5-5-15, later replaced it with reaper pc8500 1066mhz. Pll, nb and fsb termination needed a bit of tweaking to get it stable, ran the cpu at that speed up until i replaced it with a q9550.

Thanks guys, thats really helpfull.

I tried it at 3.6, 9 x 400 and at various voltages up to 1.45v. Passed 5 IBT runs but hit 82 degrees. Failed at 10 passes so it definately needs a bit more.

I'm running carp RAM in single channel also but cant justify buying new DDR2 as this is a fix to get me thrpough to the next upgrade. Would that have any impact at all? Its PC6400 800Mhz so running at about rated with a 400 FSB.

Will try and pick up something decent 2nd hand.

C64 · 14 Sep 2011 at 00:02

Just try 1.46 volts btw my q6600 hit 90c under a tuniq tower during IBT tests at 1.46 volts and was fine so 82c isn't as scary as it may appear.

troops · 14 Sep 2011 at 02:26

Had troubles on my P5q-e with 400x9, dunno why but 8x450 was ok with q6600

HummuH1 · 14 Sep 2011 at 07:59

C64 said:
Just try 1.46 volts btw my q6600 hit 90c under a tuniq tower during IBT tests at 1.46 volts and was fine so 82c isn't as scary as it may appear.

Cheers, thats a bit reassuring on the temps. Will try the higher voltages later.