Overclocking my q9450

[linford585]

I currently have my Intel q9450 overclocked to 3.2ghz, and stable after 24 hours of prime95.
Before I go on to ask for help, Ill post what my computer is:

Intel q9450 with arctic cooling freezer 7 pro and arctic silver 5 thermal compound
Biostar TP45HP Motherboard
4gigs (4x1gb) 1066mhz DDR2 Dual Channel G.Skill ram
Asus 4870
BFG 650watt power supply

My vcore is set to 1.25v, ram is set to 2v, and everything else is on auto.
My multiplier is on 8x, and my bus speed is 400mhz.
Running prime95 with large FFT's, my cpu temp never goes over 51 degrees, and for the most part stays in the upper 40's.

I wanted to overclock further to 3.4ghz, and tried putting my FSB at 425mhz, and tried many different voltages ranging from 1.25v to 1.32v, and also tried auto, which set it at 1.29v. My computer would POST fine every time, but shortly after loading programs that start after bootup, my computer would either automatically restart, just lock up, or BSOD and start a memory dump, which was very quick and would then restart. I'm fairly new to overclocking, and just starting trying/researching in late july. (However I overclocked to 3.2ghz originally back then, and havnt overclocked anything at all since then.)

Any help is appreciated!

Oh, before I forget, even though auto vcore was lower than the highest voltage I tried, it seems to take longer to crash while on auto (seems ok while on stable, I can actually open prime95... but my computer crashed about 3 seconds into running prime95).

 

[Guest]

It actually kind of sounds like a psu issue...

I mean I have an 880w psu... with lke 90 amps on 4 rails... but I think my psu is not all that great for my build as my build sometimes just randomly reboots... as in i think one of the rails is failing and not providing enough power for my setup...

so list the specific specifications for PSU

also its possible you just can't OC further... I personally would think your not getting stable power at that but i can't do anything to test mine as i don't have a second power supply... and don't want to spend upwards of 150 just to have a guess at what is possibly causing my issues


Also if you have any GTL reference function in bios try upping the voltages on those...

those specify the specific amount of voltage increase per core on quad cores... on auto it seems to set them higher than if you set them though... that could explain why it lasts longer on auto

next try turning the multiplier to 7x and have the FSB at 425... if you can't achieve stability there then you know its the FSB of your mobo that is stopping you... but if you can achieve stability at that multiplier, your chip is probably the issue

I still would think its a PSU issue but i have no way of knowing

 

[50bmg]

I would not set auto for anything. Set everything manually; CPUv, RAMv, PCI to 100mhz, etc, etc. The 45nm chips can go a little higher than 1.32v. If the temps stay low, i would try 1.33v -1.34v. My e7200 runs at 1.33v and has been stable for 6 months and never reaches 57c. You quad would get a bit warmer, but you have a much better HS&F than i have.

 

[linford585]

PSU:
> ATX 12V 2.0/EPS12V
> Dual 12V Rails
> SATA Connectors
> PCI Express Ready
> BFG Silent Control Technology
> Protection Circuitry
> MTBF: 140,000 Hours
> Safety Approval: UL, CE, FCC CLASS B
> Dimensions: 5.9" W x 3.4" H x 6.6" D

 

[cory1234]

Don't go by the specs the PSU manufacturer gives you. The actual maximum power that it can run 24/7 is more than likely less than what is written on the box.

What are you complete specs? GC?

 

[linford585]

At stock, 333mhz x8 - 2.66ghz, my mobo does the following:
CPU: 1.200v
NB/SB: 1.088v
FSB: 1.096v

If I put everything on auto, except my RAM, it does the following:

CPU: 1.296v
NB/SB: 1.176v
FSB: 1.184v


I tried turning my mutiplier down to 7x, and keeping my FSB at 425mhz. It crashed again, but im hoping it was just a coincidence, because this time when it crashed, it went to a BSOD, but it never rebooted itself, and it had a different error code than usual (d1 was the error code)... Also, when I rebooted, windows told me it had an error report ready, and asked me if I wanted to check for a solution. If however, the crash was related to me overclocking, could it not be turning my FSB high enough? I just checked the voltage while at a stable 3ghz (400mhz), and its still 1.184v. I have never adjusted the chipset or FSB voltages manually before, so I dont want to ruin anything unless I'm told more about them (such as maxes, or previous used values with my motherboard and/or cpu, etc). Also, I know every chip is not the same, but Iv heard of people going to 3.6ghz, and even heard one person say they went to 4ghz on a q9450. If mine cant even go to 3.4... Well... That would just suck lol.


(Sorry didnt completely read through the PSU specs before I pasted them, looking back Im noticing there isnt much there that helps)

PSU:
http://www.newegg.com/Product/Product.aspx?Item=N82E16817702003

 

[bpogdowz]

It's yo RAM dude. Stick to the ratios 1:2, 3:5, 5:4, 3:2

 

[linford585]

Every ratio I choose, even at 400mhz, and my computer wont even POST... So I just left the ratio on auto, its the only way my computer actually boots up no matter what Im overclocking to.

 

[linford585]

So I think my power supply went bad. Im getting 5.6v out of my 5v, and 11v out of my 12v. (Says HW Monitor). I just put everything back to stock, and now my 12v is at 10v.

Edit: My mobo, if on auto, puts my RAM on 1.8v... My RAM is rated for between 2 and 2.1v. I turned my RAM voltaged up to 2.096v (The closest I can get to 2.1, since I cant type in a number in this program) and now my 12v is putting out between 11.5 and 11.6v. However my 5v is still putting out 5.6v.

I also noticed that if I put my multiplier on 7.5x or 7x, cpuz shows my multiplier as 8x, and both cpuz and windows still identify my clock speed as being as if my multiplier is 8x.