E6400 Overclocking Problem, Please Help

[wrathloki]

Ok, so I'm following graysky's guide and I've run into an error problem during the process. In my system I'm running:

-2.13ghz e6400
-Gigabyte 965p-ds3 mobo
-2x1gb corsair cm2x1024-6400
-antec truepower 2 550w psu
-seagate 500gig 32meg cache hdd
-nvidia 7600gt vid card

Where I ran into the problem is the step where I start raising the bus speed of the cpu, (I'm using a multiplier of 8 and the default is 266) I raised it to 276 at 1.250 vcore and ran Prime95 for an hour with no errors. I raised it to 278 and started running into all kinds of errors in windows, so then I ran memtest and got about 52 errors just 44% in. I reset the bus speed to 266, vcore to 1.325, and the DRAM multiplier back to 3 from 2 (synchronous) and everything's working fine with no errors in memtest. Now I'm wondering where this problem manifested, is it just because my vcore was too low or was it because when I lowered the DRAM frequency multiplier to 2 my ram timings dropped from 5-5-5-18 to 4-4-4-12 (which is below spec on my ram)? If that's the case do I just leave the multiplier at 3 and let the DRAM frequency climb or will that cause problems by the DRAM frequency going over 800? I can't imagine the bus speed is too high since the average people seem to overclock this chip to is 3.2ghz, so I doubt I'm getting errors because of that.

Another question I have now is for some reason now when I start my computer, before post and before anything really turns on there's a loud ~3 second beep, I can't figure out why. Any ideas on why this beeping would be happening? I'm not going over any temperature thresholds or anything and all the fans are working.

 

[tricky trees]

Dunno about the beeping but as far as your overclocking goes first off what are you using a stock fan or something else? I would recommend going for a bigger jump first off. Try going for a 350 fsb at a 1:1 ratio (hich will be 2.0 in your bios list) and raise your vcore to 1.325 or even 1.350 and if it boots check your temps, run a fractal test and if all is ok try bringing down your vcore. If it boots but fails the test try adding some mem juice. No point doing little hops when you first start, do them later when you start to hit high temps and FSB walls.

 

[Evilonigiri]

You're suppose to leave the timings on 5-5-5-15. 4-4-4-12 is faster, more tighter, so it requires more voltage to run.

Run Prime95 blend test for 8 hours. If it fails, tweak the memory settings.

 

[cowspoo]

I've got the same setup basically as you and i'm at 3.0 msg me on aim Cowspoo19 i'll help you out.
does the computer boot after the three beeps? i'd think not, 3 beeps for me is a ram error, my friend has an amd and 3 beeps on his asus crossfire led poster says DET DRAM. i bet you have a bad chip. careful about frying it. if it's not posting at all try with one ram chip, then the other i BET one is fried

 

[wrathloki]

I've got the same setup basically as you and i'm at 3.0 msg me on aim Cowspoo19 i'll help you out.
does the computer boot after the three beeps? i'd think not, 3 beeps for me is a ram error, my friend has an amd and 3 beeps on his asus crossfire led poster says DET DRAM. i bet you have a bad chip. careful about frying it. if it's not posting at all try with one ram chip, then the other i BET one is fried

It wasn't 3 beeps, It was one ~3 second beep that went off before the normal single post beep (which I still get). I managed to overclock to 2.8ghz stable with my stock cooler and AS5 right now, I'm going to go farther as soon as I get a better cooler for it. I think what the problem was is that I was trying to overclock it too low for my DRAM multiplier and I was getting too low of timings on the ram.

Another question since no one's answered my other thread. When figuring out the Tcase temperature do I just take the core temp from the core temp program and minus it by whatever my processor allows. And on an E6400 with stepping B2 is it 15 degrees minus the core temp or is it 10? I couldn't seem to figure this stuff out from that temperature guide.

 

[Evilonigiri]

Another question since no one's answered my other thread. When figuring out the Tcase temperature do I just take the core temp from the core temp program and minus it by whatever my processor allows. And on an E6400 with stepping B2 is it 15 degrees minus the core temp or is it 10? I couldn't seem to figure this stuff out from that temperature guide.

Just use Coretemp and be done with it.

 

[wrathloki]

Just use Coretemp and be done with it.

I need to know what temp limit my cores can handle and I can't do that without figuring out the temperature differential. Core temp says TJunction Max (which is core temp correct?) is 85 and the thermal limit of my cpu is 61.?, but then that would mean I have a 24 degree differential between cpu case temp and cpu core temp which is wrong... So I need to know what it is so I can just add that to the 61.? and know the limits of my cores.

 

[Evilonigiri]

I need to know what temp limit my cores can handle and I can't do that without figuring out the temperature differential. Core temp says TJunction Max (which is core temp correct?) is 85 and the thermal limit of my cpu is 61.?, but then that would mean I have a 24 degree differential between cpu case temp and cpu core temp which is wrong... So I need to know what it is so I can just add that to the 61.? and know the limits of my cores.

65C is fine for that cpu, but if you're worried, make that 60C.

EDIT:Yes Tjunction is the coretemp, and 85 is the DTS value. This does not mean your CPU can handle 85, it's just a scale. Just don't let your core temp go past 65C.

 

[wrathloki]

So I've managed to get my cpu up to 3.2ghz stable (at least a couple hours in) but if I start to go higher I need to increase my vcore a lot. I tried 3.4ghz at 1.4825 and got no errors in about 10 mins, I could still get one because I got one right away at 1.475, but I had to stop for now with the stock cooler because of heat problems. Is there a chance I could still get 3.4 stable under 1.5vcore or have I most likely hit the limit of this chip? Or could it be because my DRAM is up to 850mhz from it's stock 800 and the memory isn't very good at overclocking?

 

[Evilonigiri]

So I've managed to get my cpu up to 3.2ghz stable (at least a couple hours in) but if I start to go higher I need to increase my vcore a lot. I tried 3.4ghz at 1.4825 and got no errors in about 10 mins, I could still get one because I got one right away at 1.475, but I had to stop for now with the stock cooler because of heat problems. Is there a chance I could still get 3.4 stable under 1.5vcore or have I most likely hit the limit of this chip? Or could it be because my DRAM is up to 850mhz from it's stock 800 and the memory isn't very good at overclocking?

Your chip seems normal. You need to raise the voltage exponentially as you go up in frequency.

 

[wrathloki]

Your chip seems normal. You need to raise the voltage exponentially as you go up in frequency.

Exponentially in what kind of ratio? I was at 1.375 at 3.2ghz so that would be what for 3.4?

 

[Evilonigiri]

Exponentially in what kind of ratio? I was at 1.375 at 3.2ghz so that would be what for 3.4?

Every chip is different but for 3.4GHz of that cpu, I would say somewhere near 1.45v

 

[wrathloki]

Every chip is different but for 3.4GHz of that cpu, I would say somewhere near 1.45v

Except I got errors right when I started the test at 1.475 even...

 

[Evilonigiri]

Except I got errors right when I started the test at 1.475 even...

That's why I said every chip is different. Perhaps yours isn't a good OCer?

 

[wrathloki]

That's why I said every chip is different. Perhaps yours isn't a good OCer?

That's what I'm thinking, I just want to see what the experts think before I go buy an aftermarket cooler.