nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Child Item

Help with SEVERAL failed Overclocks!! Q6600

ibFoxer

Hi Guys,

Hoping i can get a bit of help from the tech geniuses here, having a bit of trouble overclocking my Q6600 (G0). The kit i'm using is:

Q6600 (G0) with Tuniq 120 cooler
Asus P5Q-E Motherboard
Corsair XMS2 PC8500 CAS5 (2 of these kit CM2X2048-8500C5)
Antec NeoPower 550
Coolermaster 430 Elite

On my first attempt i followed the advice here: http://www.overclock.net/intel-motherboards/387626-overclock-q6600-p5q-e-bios-help.html

It went well, everything booted, flew through the OCCT test over 2 hours @ 59degrees. Put it through PRIME95 and it failed out after approx 2 hours.

I then changed the settings as per here: http://forums.hexus.net/hexus-hardware/110267-c2d-overclocking-guide-beginners-p5k-add.html

With a few tweaks i managed to get it to run through OCCT overnight (5hours) @3.17ghz, but again PRIME95 failed out after approx 2 hours.

Didnt boot on first attempt, booted on second when i changed the ram back to 800mhz with the FSB Frequency @ 333mhz, but again it crapped out on PRIME95 after roughly 2 hours.

Then it refused to boot, i cleared the CMOS, tried it again and have not had a stable boot except for when i set it back to stock. I'm 99.9999% everything is ok hardware wise, cooler is seated correctly, RAM is in PWO, etc.

Just wondering if theres anything else i can do? Is there any definitve settings that i can run it at to get it to at least 3ghz? Or is it likely a hardware problem? Have looked around the various forums but it seems to be the same info each time.

Cheers.

Find more posts tagged with

Quick Links

Comments

Solitaire

You sure you're dropping the memory divider to compensate for inadvertent RAM OCing? Spread Spectrum disabled? EIST enabled? (if it is, disable it and try again, but most of the time Core2 works better with it on) All voltages NOT set to Auto?

ibFoxer

Solitaire wrote: »

You sure you're dropping the memory divider to compensate for inadvertent RAM OCing? Spread Spectrum disabled? EIST enabled? (if it is, disable it and try again, but most of the time Core2 works better with it on) All voltages NOT set to Auto?

Not sure about the divider, but All the stuff that should be disabled is, spread spectrum, PCIE spread spectrum, speedstep etc. Had voltages set to following:

CPU- 1.35
RAM-2.10

CPu Ratio Control- 9x
FSB Frequency- 333MHz
PCI-E Freq- 100
DRAM Freq- 800MHz (RAM is rated @ 1066)
DRAM Timing- Manual 5,5,5,15

Perhaps i need to bump the northbridge a bit, maybe the CPU Pull & FSB Termination too. Have noticed that with the FSB set to 333MHz the auto RAM frequency is 800MHz but it shows at 5:6, not 1:1, maybe pull that back to 266 and work up? Tired now, have reset the CMOS so many times iv lost the feeling in my fingers :P

EDIT- In relation to the ALL VOLTAGES, the only 2 were the CPU & RAM, theres no definitive guide for the P5Q-e and it differs slightly from the other P5Q boards ever so slightly. Seems to be a few different settings from the board from the hexus guide.

Solitaire

And Asus uses different names from MSI so I'm not going to be much help either

Raise all the other voltages a little and try again. Also make sure the P5Q-e doesn't have a manual FSB switch/jumper, those can cause instability if they're set to a different setting than the FSB set in the BIOS.

At this point I'm guessing you have a bad mobo or bad CPU

Last resort but have you tried updating the BIOS to the latest version?

ibFoxer

Solitaire wrote: »

And Asus uses different names from MSI so I'm not going to be much help either Raise all the other voltages a little and try again. Also make sure the P5Q-e doesn't have a manual FSB switch/jumper, those can cause instability if they're set to a different setting than the FSB set in the BIOS.

At this point I'm guessing you have a bad mobo or bad CPU Last resort but have you tried updating the BIOS to the latest version?

Have decided to try again using the following conditions:

Set the Multiplier and FSB to default values of 9 and 266.

RAM Timings @ 5,5,5,15, 2.1v. I also set my RAM Frequency as DDR2-800.

Core CPU Voltage to 1.35v. (i notice in core tem it is 1.3250)

NB voltage to 1.16v.

SB voltages on AUTO.

Left all the other configurations on AUTO.

I disabled all the settings in Advanced mode except Execute Disable Bit

Running PRIME95 now for an hour, if all goes well i will go back in and change the following:

FSB- from 266 to 333
Core Cpu Voltage- from 1.35 to 1.325

This should give me 3.0GHz, will then run PRIME95 for an hour or so and see how it goes.

EDIT- No joy, failed to boot the first time, changed the VID back to 1.35, booted into windows but BSOD'd immediately. Faulty hardware mayhaps? Or wrong figures?

Solitaire

AFAIK that NB is too low... might be okay for P45, but too low for P35 for sure. Pump the NB and SB up to 1.2-1.3V and set the FSB Terminator and VTT to ~1.3V. IIRC its the VTT that should never ever be set above the CPU voltage to avoid excessive MELTING :eek::D:P I have no clue what SB I/O Voltage maps across to on Asus boards, IIRC its usually set very high with 65nm CPUs...

PogMoThoin

Here's my 24/7 overclock with a very similar mobo (specs in sig), its rock solid stable and will do this overclock with lots less voltage (it'll do 3.8Ghz underclocked actually

), I'm overvolting the cpu as my temps are very good (water).

Don't leave ram timings or other settings on auto, set them manually. 1.16v for northbridge seems very low, I've mine at 1.58v to push my oc through, but my northbridge is on water. I can get to 4Ghz, but any higher than 3.8Ghz and I've to remove 2 sticks (4Gb) of ram.

Hopefully you'll see something you missed.

ibFoxer

Solitaire wrote: »

AFAIK that NB is too low... might be okay for P45, but too low for P35 for sure. Pump the NB and SB up to 1.2-1.3V and set the FSB Terminator and VTT to ~1.3V. IIRC its the VTT that should never ever be set above the CPU voltage to avoid excessive MELTING :eek::D:P I have no clue what SB I/O Voltage maps across to on Asus boards, IIRC its usually set very high with 65nm CPUs...

AFAIK it's P45, but will bump the NB & SB to 1.2v and see what happens. You sure about the VID @ 1.3v though? Should it not be the same as what coretemp reads?

As an aside, i wonder should i try the other Q6600 i have (same stepping), and reseat the cooler, IF i get no hop?

ibFoxer

PogMoThoin wrote: »

Here's my 24/7 overclock with a very similar mobo (specs in sig), its rock solid stable and will do this overclock with lots less voltage (it'll do 3.8Ghz underclocked actually ), I'm overvolting the cpu as my temps are very good (water).

Don't leave ram timings or other settings on auto, set them manually. 1.16v for northbridge seems very low, I've mine at 1.58v to push my oc through, but my northbridge is on water. I can get to 4Ghz, but any higher than 3.8Ghz and I've to remove 2 sticks (4Gb) of ram.

Hopefully you'll see something you missed.

Cheers for that

Should be a big help. I'm curious though, you didn't specify a multi, i see it's sets auto to 8.5, do you get 1:1 timings with that? Also, where did you pick up the figures for CPU PLL etc??

EDIT- Noticed a few differences straight away, in pic 2 you have both DRAM options disabled, Also in pic 3 you have load line calibration ENABLED-that goes against every other guide?? Also, you have CPU Margin Enhancment set to optimized, again, that ontravenes every other quide?? Is there that many differences between the 2 chips or are you working off a stable template??

PogMoThoin

Checking my notes from when I overclocked this last year, load line calibration should have been disabled and I've disabled it now.

I don't follow anyone elses' overclock, I push and find my own values. This has been stable for months.

ibFoxer

PogMoThoin wrote: »

Checking my notes from when I overclocked this last year, load line calibration should have been disabled and I've disabled it now.

I don't follow anyone elses' overclock, I push and find my own values. This has been stable for months.

Right, i guess i will try some higher values. If it fails then i can assume it's a hardware issue

ibFoxer

Im going to put this down to hardware, maybe i will try reseating the CPU, or try a different one. For now looks like im stuck with stock

BigEejit

I had my q6600 running on an Abit board @ 3.4 on air, it ran hottish (55 - 58 under load) but was stable. I had my MCH set to 1.35volts.

Also I had memory issues, after a lot of research it turns out that using 4 sticks of ram required that I drop my memory speed from DDR2-800 to DDR2-667 ... its Corsair ram and this is what was recommended on their forum!!. Dont ask me where on the forum, it was a good while ago.

(MCH = Memory Controller Hub = Northbridge)

BigEejit

http://forum.corsair.com/v3/showthread.php?t=77762
see second post:

You are advised to drop the bandwidth to 667Mhz and raise the NB Voltage +.2v. Also, you are advised to set the timings to the highest latencies of your modules which is 4-4-4-12 2T.

I know in this case it is just some guy on the forum, but I also read that advice from a corsair representative.

Solitaire

Ah feck, good point! If you have four modules instead of two P35/45 no likey

Try adding another 0.2V to the northbridge and whatever Asus calls the FSB memory lines, probably either FSB terminator or NB GTL... you may also have to loosen memory timings.

ibFoxer

BigEejit wrote: »

I had my q6600 running on an Abit board @ 3.4 on air, it ran hottish (55 - 58 under load) but was stable. I had my MCH set to 1.35volts.

Also I had memory issues, after a lot of research it turns out that using 4 sticks of ram required that I drop my memory speed from DDR2-800 to DDR2-667 ... its Corsair ram and this is what was recommended on their forum!!. Dont ask me where on the forum, it was a good while ago.

(MCH = Memory Controller Hub = Northbridge)

Believe it or not that did occur to me but i didn't do it, will give it a whirl tomorrow night alright and let yez know how i get on, thanks very big for the info though

ibFoxer

Solitaire wrote: »

Ah feck, good point! If you have four modules instead of two P35/45 no likey Try adding another 0.2V to the northbridge and whatever Asus calls the FSB memory lines, probably either FSB terminator or NB GTL... you may also have to loosen memory timings.

Any rough idea on the timings Sol??

I must post screens of the actual teak menu, might help putting the differwent names to whatever asus calls them

RobMozza

In my opinion non stable overclock's are usually down to insufficient voltage. Whilst booting failure is often down to the memory running too quickly.

It can be very hard to find the right balance between voltage and operating temperature, and of course this will vary slightly between seasons.

Also as your finding out the very best overclocks are normally with two sticks of low capacity memory. Again it's another balancing act between having a useful overclock 24/7 and an extreme overclock for the odd occasion and using the same hardware components for both!

Hope you get this sorted, if it were me I slacken the RAM timings some more 5-5-5-15-2T and see what happens. Another thing from the screen shots, if at all possible try not to use "Auto" settings. I always found that actually specifying the correct value where know was always better than letting it decide for itself.

ibFoxer

RobMozza wrote: »

In my opinion non stable overclock's are usually down to insufficient voltage. Whilst booting failure is often down to the memory running too quickly.

It can be very hard to find the right balance between voltage and operating temperature, and of course this will vary slightly between seasons.

Also as your finding out the very best overclocks are normally with two sticks of low capacity memory. Again it's another balancing act between having a useful overclock 24/7 and an extreme overclock for the odd occasion and using the same hardware components for both!

Hope you get this sorted, if it were me I slacken the RAM timings some more 5-5-5-15-2T and see what happens. Another thing from the screen shots, if at all possible try not to use "Auto" settings. I always found that actually specifying the correct value where know was always better than letting it decide for itself.

Thanks RobMozza- the sickening part is i had it more or less stable @ 3GHz, it crapped becaouse of the RAM voltage i'm sure, but i havent been able to recreate that. As far as the timings, 5-5-5-15 was the setting, i don't think i specified 2T though, might be worth a shot.

I'm beginning to think it is the voltage because i ran it @ stock on PRIME95 and if it was hardware i'd imagine there would have been issues. So it seems i just have to work at the correct timings,volts, etc. It's just a pain in the a$$ to have to reset the CMOS evry 3rd or 4th time

:rolleyes: