Q6600 n OCZ vendetta 2

[SuPrEIVIE]

Hello, after i applied the heatsink and connected the fan to cpufan on mobo (P35-DS3L), when comp is turned on the fan doesnt function?? I just replaced the stock heatsink and that worked fine. My temps at idle right now are pretty much 40C across the cores now but with no cpu fan involvement. Those temps are pretty much the same at idle with the stock heatsink. So i am also worried about that because i want to OC to 3.6 ( couldnt even do 2.9 with stock heatsink) With fan on for the ocz v2 will it show a much greater difference than with stock intel heatsink?

thanks in advance!

 

[SuPrEIVIE]

I just tried a 333X9 OC and it failed , i followed instructions in the stickied thread, but i keep failing prime. Also how can you be certain that all pushpins are locked in? I push hard on diagonal pins, but i dont always hear a click, but pushing harder doesnt seem to lower the vendetta 2 more than it is rightnow. Just cant understand why a moderate OC fails considering temps are a bit lower than the stock

 

[kpo6969]

Same mobo and cooler here.
In bios enable smart fan control
three settings:
Auto
Voltage
PWN
I use PWN
Push-pins are a bitch, took me almost an hour to get it right.
Diagonal pins, 2 at a time, you should hear 2 clicks if right (I'm still not 100% sold on mine being right)
I may breakdown and get a back plate mount. Xigmatek works (I think)
E8400 @ 3.6
35c idle
51c load

 

[SuPrEIVIE]

Thanks it came alive after the switch, still cant do 333x9 even at 1.325 volts, is there some specific settings to change on this mobo to make an OC pass prime? Or is my CPU not cable of this? thought this would be easy with this HSF

EDIT: im also using ocz freeze TIM and a well ventilated case with 3 120mm fans

 

[Idontcare]

Originally posted by: SuPrEIVIE
Thanks it came alive after the switch, still cant do 333x9 even at 1.325 volts, is there some specific settings to change on this mobo to make an OC pass prime? Or is my CPU not cable of this? thought this would be easy with this HSF

EDIT: im also using ocz freeze TIM and a well ventilated case with 3 120mm fans

I've got multiple (5 to be exact) systems running Q6600 on DS3L with the Tuniq Tower all at 9x367 on sub 1.35V Vcore's.

Surely your system with the vendetta 2 ought to hit 9x367 as the vendetta is better than a tuniq, isn't it?

 

[SuPrEIVIE]

yeah it should, i dont get it, i just decided to see OCs under 3.0 and the furtherst away from stock speed that ran stable prime for 1 hr was 2.8 anything above that fails prime in a few minutes tried many things,

This cooler gotta be better than this, i must be doing something wrong dont know what here are my specs

P35-DS3L rev2
4GB GSkill PC 6400 ddr2800 5 5 5 15
Q6600 G0, OCZ vendetta2 + ocz freeze tim
Corsair 750 W TX
640AAAKS
9600GT
rocketfish case 3x120mm fan

One note when installing the vendetta 2 i didnot remove mobo from case since it was push pins, i just did it right in the case. I have no idea why this is underperfoming, i have reseated this HS so many times

 

[Idontcare]

Originally posted by: SuPrEIVIE
4GB GSkill PC 6400 ddr2800 5 5 5 15

Have you set your ram timing and settings to 1:1 for these initial overclock efforts?

For 9x333 attempts that would mean your ram ought to be clocking at DDR2-667.

 

[SuPrEIVIE]

yeah i did all 1:1 ratio, im going to try and mount the heatsink differently to see what happens, i currently had it with the fan facing the top of the case, going to switch so the fan faces the rams, the 2.8 setup ran for 4 hrs and stable at 1.200 but at full load near 60 across the cores

 

[kpo6969]

I have mine facing front (ram) and blowing out the back.
Also what's you ram voltage?
Board defaults to 1.8
My G.Skill is 2.0-2.1v and have to bump it up +.03 to run it's rated speed (5-5-5-15 1000)

 

[SuPrEIVIE]

my ram voltage is 1.888 (1.89 i think is the limit for this gskill) according to everest and in the bios it is at normal no + adjustments. I just changed the position so now fans facing ram and rear facing exhaust fan the temps when down about 1-3C. I guess i will give a run at 3GHZ+ again, should i do increases to other voltages? in my attempts i only made adjustments to the cpu voltage. However once i did + one step for MCH FSB and i couldnt even log in to os after doing that.

 

[SuPrEIVIE]

my ram voltage is 1.888 (1.89 i think is the limit for this gskill) according to everest and in the bios it is at normal no + adjustments. I just changed the position so now fans facing ram and rear facing exhaust fan the temps when down about 1-3C. I guess i will give a run at 3GHZ+ again, should i do increases to other voltages? in my attempts i only made adjustments to the cpu voltage. However once i did + one step for MCH FSB and i couldnt even log in to os after doing that.

 

[Idontcare]

Originally posted by: SuPrEIVIE
my ram voltage is 1.888 (1.89 i think is the limit for this gskill) according to everest and in the bios it is at normal no + adjustments. I just changed the position so now fans facing ram and rear facing exhaust fan the temps when down about 1-3C. I guess i will give a run at 3GHZ+ again, should i do increases to other voltages? in my attempts i only made adjustments to the cpu voltage. However once i did + one step for MCH FSB and i couldnt even log in to os after doing that.

At stock can you report your idle and small FFT loaded temps? Having ruled out memory as an obvious OC limiter the next suspect is your HSF and those pushpins.

Also look for how much Vdroop is going on, you could possibly have a bad DS3L...happened to me and I only caught it because Vdroop was >0.2V.

There have been countless threads like this in the CPU forum these past 6 months where after great lengths of debugging the HSF really was the culprit and it comes down to those pushpins. Simply reseating the pushpins does nothing to ensure they've actually latched thru the motherboard PCB.

The only way to *confirm* all four pushpins are latched is to actually engage them with the mobo out of the case so you can visually inspect and see for sure that the plastic collet that rings the bottom of the pushpin is in actually on the proper side of the motherboard. If the collet is on the wrong side (that would be the CPU side) then the nib can't properly lock it.

No one likes to pull their mobo (me included), but if you go the cheapy route of using those pushpins and you find yourself not getting the OC you expected then its time to roll-up them sleaves and put some elbow grease into it.

 

[SuPrEIVIE]

Thanks for the informative reply, stock temps at idle and smaff fft load are 31C to 50C

Vdroop appears to be 0.05000 interestingly, running prime, vcore is less than idle. I managed to do a 3GHZ OC around the 1.3V area in the bios that runs prime successfully for 3 hours so far, at load it is = < 61C. 3.2GHZ however is quite a stretch thought since at 1.4 v in bios it fails prime.

Only reason i cringe at pulling my mobo out because its obviously a pain, and i dont expect to improve the HSFs seating on the mobo than it already is, but the only way to really find out is to do it to confirm like u said, so if i ended up wasting time, i'll just try and suck it up (hopefully not)

 

[toadeater]

What temps are you talking about here? CPU or core? According to which program? Download Real Temp or Everest and get the temps from that.

If you're still unsure of the push-pins, you could buy the Thermalright bolt-through kit. You could also try a higher RPM fan, the stock fan's top speed is 1600RPM, but it might be running slower than that from the BIOS fan control.

 

[Idontcare]

Originally posted by: SuPrEIVIE
Thanks for the informative reply, stock temps at idle and smaff fft load are 31C to 50C

Yeah those stock temps make me entirely suspicious that you got an incompletely latched pushpin somewhere as I hit those temps with my Tuniq if I reduce my overclock to 3.0GHz...no way should you be seeing that kind of temps at a measly 2.4GHz with your cooler.

Those are like stock temps at stock clocks with a stock cooler.

 

[SuPrEIVIE]

The temps im referring to are the four cores, im using the program everest. I just took out the mobo and one pins was not fully locked in underneath, took care of that and put everything back in and right now at 3GHZ its #1 40 #2 40 #3 37 #4 42 at idle, not much of a difference before confirming. While running prime for a bit looks like this #1 60 #2 60 #3 57 #4 59 vcore at 1.232

EDIT: also while priming the fan is at 1350+ RPM other case fans running at top speed always at 1600+RPM

 

[Idontcare]

Wow, I'm quite literally at a loss for words.

If you get those temps with your HSF on that specific CPU (which must truly be one of those 1/1000 bad eggs for overclocking viewpoint) then I can't imagine what you'd get were you to use the stock Intel HSF.

The short answer is no doubt your temps are causing the OC instability, why your temps are so high at such low clocks and voltages is beyond me now that you've eliminated the HSF as a cultprit.

(edit: by "so high" I mean you are seeing temps here at 3GHz that really you shouldn't be seeing for another 10% higher clocks - 3.3GHz territory. Could be you jsut have one of those unlucky CPU's)

Unless the HSF and/or IHS are just literally phenomally badly warped?

Do you know if your HSF typically requires lapping to get the temps down? (like the Ultra's were notorious for requiring)

 

[SuPrEIVIE]

I think the ultras are more for lapping the vendettas with the exposed heatpipes i think you cant lap it because it might spill some liquid since they are already flattened on the cpu contact area,

looking at the vendetta 2's contact area it seems fine just as it should out of the box before i put it to use, my q6600 has some faint burnt spots near its corners that i noticed after removing the intel stock, use 100% acetone followed by 91% alcohol for cleaning stock grease etc. I then applied some freeze using the technique from here (the very last demonstration on that page) When i pulled off the vendetta 2 away from the q6600 to see the bottom i noticed that the faint small burnt markings near the edges of the q6600 got on the vendetta's contact surface i couldnt wipe it off with the chemicals neither.

The IHS doesnt look too imperfect and the grease spread well when compressing the vendetta 2 to it. IIRC pairing vendetta 2 with quad chips should be good straight of the box with no lapping