Q6600 Stable 8+ Hours With samllFFT, Fails Soon on Blend

[Tron87]

The title gives the general idea. I have my Q6600 set to 3.0Ghz. Here are some more specs

X9 Multiplier
FSB 333
Vcore 1.168V
(interesting note: Idle Vcore is 1.216V)
Temps never go over 61C.

So I feel like things are solid with the smallFFT tests. I ran over eight hours at 2.8Ghz and the same time at 3.0Ghz with no errors.
The voltage setting was left at Auto, hence the changing I guess.

I dont quite see why I should be failing the blend test.
My Ram is Patriot Viper DDR2 800 2X2Gig. The same kit that TH was able to take to DDR2 1200 at 7,7,7,11 @ 2.2V.
I've got mine set at stock timings and stock voltage (4,4,4,12 and 2.1V).

Should I start running a crap ton of Memtest86 trials to test for bad memory or maybe try adjusting voltages?
What do you guys think is the culprit in this case?

Q6600
GA-P35-DS3L
PV DDR2 800 LLK 2X2gig
650TX PSU
Xigmatek HDT-s1283

 

[Fruity]

Try setting your RAM timings back to 5-5-5-15 and run blend test again. If it still fails, keep your RAM timings as 5-5-5-15 and adjust your FSB VTT and NB Voltage.
At 3.0GHz they might need a *little* extra juice to get stable.

Once everything is running stable try putting your RAM back to 4-4-4-12 and retest.

hth

 

[Fruity]

Try setting your RAM timings back to 5-5-5-15 and run blend test again. If it still fails, keep your RAM timings as 5-5-5-15 and adjust your FSB VTT and NB Voltage.
At 3.0GHz they might need a *little* extra juice to get stable.

Once everything is running stable try putting your RAM back to 4-4-4-12 and retest.

hth

 

[Zorg]

Oh how I want to post, but we lost our moderators.

 

[Tron87]

^What does that mean Zorg?

I did loosen the timing s to 5,5,5,15 last night. Also set NB up one notch.
Also set Vcore to 1.2625V which is .05V over idle draw and .1V over load draw.
Still failed Blend in about five munites.

With my original settings I failed Blend in 2.5 hours, 4 hours and 15 hours across three cores. Last core made it the whole test time, about 16 hours.

I'll set everything back to those settings and fiddle with the FSb and NB volts.
I've also got to post some pics or people wil get bored.

 

[Tron87]

I decided to set everything back to stock.
The Q6600 is back to 2.4Ghz
The Patriot Mem is back to DDR2 800 4,4,4,12. No errors so far. Better not be.

After this test I'll go back to 3.0Ghz. Will I have proven that the mem isn't the problem or am I just wasting time?


Another funny note. My CPU has a 2 degree temp diff from 2.4 to 3.0.
2.4 GHz load runs at 59C. 3.0Ghz load got to 61C for a few seconds but is usually 60C.

Not a bad rise for a 25% OC.

 

[The Taz]

try intelburn test 1.9 i have a q6600@2.4GHZ running at 3.6GHZ.

My setup

CPU----------------Intel core2 quad Q6600 @2.4GHz overclock to 3.6GHz
Motherboard-----P5Q
Ram----------------Corsair XMS2 Dominator 2GB
Case---------------Thermaltake Armor LCS VE2000BWS
PSU-----------------Jeantech Storm 700W Sli/Cross-fire
GPU-----------------Nvidia Geforce 9400 GT


setting

CPU Ratio-------------------9
FSB Frequenc--------------400
DRAM Frequenc------------1066(this is my ram rating)
CPU voltage----------------1.4
DRAM Voltage--------------2.1

Idle temp 33c,30c,30c,30c
temp after runing the above test were 72c,72c,68c,68c


P.S have got it upto 4.5HGz but need more cooling eg nb and another rad than i tr again,going to see what the temp are at 4.0GHZ tomorrow

 

[cjl]

If it fails blend, try upping northbridge voltage and/or memory voltage.

 

[Tron87]

Well I still haven"t gotten it. I did a overvolt blitz across the board
Mem 2.2V
Vcore 1.3V
all others +.1V

Still no stability at 3.0Ghz.
I might have messed it up by loosening the timings to 6,6,6,18 and setting the Dram to 1000.
I'll have to reset that.

Anybody think I should lower the multiplier and go for a 1:1 ratio?
Or should I underclock the ram?

Also, what program should I use to monitore NB temps and such? I dont think HW monitor does it.

This is the first thing that I have ever OC'd but i made a point to choose parts that people have had great success with in the past. I would really appreciate it if someone followed me along and helped out.

Thanks all very much. I'll start a new slew of testing tomorrow and get really specific with the data. I'd like to reach 3.0 or even 3.2 Ghz with stability. And maybe even get my ram to 1066. TH proved it is possible.

 

[eagles453809]

dude what is your fsb set to?

 

[Tron87]

Almost all my attempts are FSB 333. Pretty mellow

 

[Tron87]

So ive made some changes. Mostly these.
Chnaged multiplier to X8 and set Bus Speed to 375.
This gets me 3.0Ghz.
Got the mem running 1:1 for it is underclocked at 750 instead of 800
but it is running spec timings of 4,4,4,12 and voltage 2.1V

Vcore is 1.2625V and all other support volatges are up one notch(My bios doesn't actually display the voltage)

Also, anybody recommend a program for viewing NB temp and such.

Prime95 stable for 3 hours now.