Hello! i would like to OC my Q6600. I have 0 exp, can you help

[jasjeet]

also i ran memtest on boot and had 0 errors after an hour, this was at 1.86v ram but blend still failed at the same time and core

 

[jasjeet]

seems like is was over volting the ram back to 1.8v seems to be working fine for 10 minutes blend test, temps are lower on this test at 49 49 46 45

 

[jasjeet]

confused! it failed blend at 14 minutes through with ram at 1.8v so i pu it up to 1.82v, that failed in 3 minutes. NB is set at 1.34v and vcore at 1.41v.

i dont know what to do now???

 

[jasjeet]

well i think my ram didnt like the slightly higher clocks, im back at 3.2ghz prime stable at blend and small fft with ram at 801mhz.

 

[V3NOM]

loosen the timings a bit and put the voltages as what they say on the website or box.

 

[jasjeet]

i tried 1.8v with 6 6 6 18 timings but it still didnt like it, small fft pass for over an hour blend fails pretty quickly. i ran memtest for an hour or so and it had 0 errors.

i got a stable 3.25ghz with ram 1:1 and all voltages on auto apart from vcore which is passing blend.

by keeping a 3.25ghz clock how am i effecting life expectancy?

 

[iluvgillgill]

i think you will need more voltage on the NB then just 1.3V. and what Vcore you have now.

and what V3nom have said on a quad core Core 2 the heatpipe should be on top of the two cores which meaning the fan on the cooler should be blowing towards the exhaust fan BUT the Antec 300/800/1200 will have it at the upward position due to the much bigger top fan which the case has a much bigger updraft then backdraft.

 

[V3NOM]

^but that means the heatpipes won't be touching as much surface area of the cores! less heat transfer= higher temps haven't you read lupi's xigmatek guide? 😛

 

[iluvgillgill]

yeah i have.lupi told me in MSN. but the thing is the rpm on the stock xiggy is not exactly great considering this. if he placed it on the up position there will be more air flowing through the fins.

and there is the heatspreader anyway so if on the Core area where there are less heatpipes being "touched" the heat will eventually spread out which then will get transfer to the other heatpipes which didnt get heated up before.

but i think jasjeet could find that out for us? do it will all the fans max out.

 

[jasjeet]

you could be right, maybe thats why the core 3 and 4 run 5 degrees cooler than core 1 and 2 at all times!

but yeah on the antec 300, which i have, the 140mm fan up is the larger exhaust. also at *idle* my cpu fan is running at 800rpm 120mm (Vendetta 2), with 30 29 25 26 right now, (been online for 6hrs).

 

[jasjeet]

i totally didnt think about the core position on the cpu and heat sink, all i thought was that heat rises so may as well face it to blow air up wards lool

 

[jasjeet]

rpm ranges from 800 - 1450rpm on Vendetta 2 fan as well.

umm i just re read your post, what do you want me to do? LOL

oo just found quick edit, i was trying edit but it never worked hehe

ok so i have vcore at 1.40V 406X8= 3.25Ghz. Ram 825Mhz 1:1 4-4-4-12. i put everything else on auto and it stabilised prime95 blend.

whenever i use my own NB, SB and VDimm i get unstable because im guessing. i had NB 1.34v, SB 1.20v, Vdimm 1.9V and i couldnt pass blend!?!?

edit: looking at http://benchmarkreviews.com/index.php?option=com_content&task=view&id=156&Itemid=1&limit=1&limitstart=6

it seems that at 20-21 degrees ambient temps went to 34 degrees 100% load on a Core 2 Duo 2.66GHz (Overclocked to 3.6 GHz)

maybe it is better to have it the other way around?!

 

[iluvgillgill]

have you change the heatsink to face and front and blow out the back?

 

[jasjeet]

nah i cant because i used the TIM that came with the heatsink so i have none left atm. also its my first build so i dont really want to temper with it really.

 

[V3NOM]

this is a question thats been bugging me for a while... can you apply the TIM, mount the HSF onto the CPU, then take it off again to observe the spread of TIM, then mount it again without re-applying the TIM?

 

[iluvgillgill]

well i think you can!people will say the microscopic dust and particle will get trap on the TIM. but have people really test it out if it will actualy affect the temp up to a point where the monitoring software report 1-2C rise?i dont think so.

 

[V3NOM]

hmmm good point wow lol this is weird typing in the box in Chrome! it like turns brown! :O

 

[iluvgillgill]

i still prefer Firefox to be honest even though Chrome is more animated(the tab open and closing bit)

 

[jasjeet]

ive been trying chrome too it seems good apart from no adblockplus kinda thing for it. fast nevertheless.

 

[iluvgillgill]

somehow chrome load webpage specially tom's forum slower then firefox.

any OC going on jasjeet?

 

[jasjeet]

lol thanks for the interest, i just tried 3.3ghz again at 412X8 but blend fails me again. i upped NB to 1.40V this time, maybe the chips a dud?! i get an error, resulting sum was -1.XXXXXXXXXXXXX expected:8.XXXXXXXXXX. the number always changes.

not sure what to do?
i had 3.25ghzz stable, only manually entered Vcore, the rest on auto.

now i heard the q6600 like high NB so i put it up but still nothing.

 

[iluvgillgill]

above i meant specially slow SLOW page loading.

you need to up FSB termination voltage and MCH voltage if its available in your voltage option.

 

[jasjeet]

i upped fsb termination voltage to 1.28V, i cant see MCH anywhere. its getting better. it usuall errors at 1minute or 3 minutes. ive got it to 3 minutes so far and the error number is closer to the expected number so i think im getting closer.

also what is GTL refernce voltage, ive set both to 0.670, ive read its optimal for my mobo.

i have ram 1:1 still 5-5-5-15 at 825mhz

thanks

 

[iluvgillgill]

you actually might need more then 1.28V on FSB term at 3.25Ghz.

just a reference i done Q6600@425X7@1.38V vcore + 1.55V FSB term.

 

[jasjeet]

really ok ill up it to 1.40v and see what happens. it isnt dangerous is it?