Sign in with
Sign up | Sign in
Your question

*Problem Solved* blah OC on e4300

Tags:
Last response: in CPUs
Share
February 15, 2007 5:37:24 AM

hey y'all,
i just changed my system from athlon x2 to c2d e4300
i purchased bad axe 2 and e4300; its OC ain't good as e6000s.
i saw many people overclocking their e6400 or 6600 with FSB of 1600 or above; my CPU cannot go over 2.8GHz.
No matter what I do, including maximum voltage i can give, it won't go over 2.8 GHz.
Is this the limit of CPU?? or could it be the mobo??
RAM got no prob since it's PC6400 800MHz; 1:1 @ 400FSB barely gives rated speed.
so i believe it's either cpu or mobo...
upto 2.8ghz, it's rock stable 12+ hrs of prime 95.
anyone knows if it's CPU limit?? maybe cheap die??
if not.... anyone got a solution about how to fix this problem??
thanx in advance.

Spec:
Intel Core 2 Duo E4300 @ 2.7GHz
Intel D975XBX2 Bad Axe 2
OCZ XTC Gaming RAM DDR2 PC6400 2 x 1GB
SeaSonic S12 600W
February 15, 2007 5:56:04 AM

You know that that OC is more than double what your X2 did right? I mean a 55% OC is very nice.
Even though you are only running the ram @ rated, try upping the v-dimm 1/10th of a volt. Also, make sure your temps are not a problem. The north bridge, and the mosfets are key places to check.
What kind of hsf are you using?
You also might want to check out the overclocking forum.
February 15, 2007 6:02:28 AM

thanx for reply.
i know 55% oc is much more than my old x2 4200+,
but it seems a bit less than what those oc forums and other OCers are experiencing...
temp doesn't seems like to have any problem.
i use thermaltake bigwater to cool it off.
well... any suggestion for an OC forum which talks about e4300??
e4300 is not widely spread yet... maybe it never will...
Related resources
February 15, 2007 6:07:57 AM

Are you using 7x400 or 9x311 to reach 2.8GHz?
February 15, 2007 6:42:35 AM

Quote:
Are you using 7x400 or 9x311 to reach 2.8GHz?


i'm using 9x311.
ram is set to 533, so ram's currently running @ 415MHz = 830MHz DDR2
Bad Axe 2 doesn't let you change the multiplier.
there is a menu about the multiplier, but it's @ automatic; you can't touch them for some reason...
February 15, 2007 7:11:40 AM

With water, the mosfets don't have the fan air blowing on them. They tend to heat too much. A side panel fan can help.
You might want to raise the latencies on your ram, to see if that gives you more room. Does the bios allow 1:1 on ram?
February 15, 2007 7:14:05 AM

I thought the multipliers in the core 2 duo was locked and only the fsb frequency can be altered?
February 15, 2007 7:17:17 AM

Quote:
I thought the multipliers in the core 2 duo was locked and only the fsb frequency can be altered?

A lot of MBs take advantage of the SpeedStep feature to enable you to use lower multipliers all the way down to 6x. Only the EEs allow you to use higher than stock multipliers though.
February 15, 2007 7:20:16 AM

They're locked only upwards
February 15, 2007 2:56:23 PM

A few things I can think of:

1) Try testing with the 8x multiplier. Many boards exhibit odd FSB behavior - they may be stable from 266-300, unstable from 300-340, stable 350-450, and so forth. You might be near one of those unstable zones, and the link above regarding bootstraps is the technical explanation for such zones.

2) Try incrementally lower Vcore values as well. Heat destabilizes overclocks. At a certain point, additional Vcore begins to lower the maximum overclock you can get. Heat dissipation is proportional to clock frequency as well as to the square of voltage.

3) Accept that e4300 is the lowest marketed bin - it might even be manufactured using a cheaper technique that still produces acceptable yields (we already know it's a cheaper die). You could try returning the chip or exchanging it for another one. We normally assume that since Intel's 65nm process is mature, they get high yields at or near the top bin (and very few lower-bin processors), and so a lot of their bulk mainstream processors must actually come from the same bins used for high-end processors - just not all.
February 15, 2007 4:12:11 PM

Quote:
Try changing the Reference Frequency as seen in the below picture to 266 or 333. This will cause the 975x chipset to use looser timings:

http://www.motherboards.org/imageview.html?i=/images/re...


so don't put the reference frequency of the memory on default?
what does reference frequency do??
February 15, 2007 4:16:06 PM

Thats about all you can expect from the Bad Axe boards without doing a volt mod. You can probably squeeze and additional 50 to 70MHz out of it, but you're reaching the max for 975X. The problem isn't your CPU, its the board.
February 15, 2007 4:16:20 PM

Quote:
With water, the mosfets don't have the fan air blowing on them. They tend to heat too much. A side panel fan can help.
You might want to raise the latencies on your ram, to see if that gives you more room. Does the bios allow 1:1 on ram?


my case has side blowing fan, and yes this bios allow 1:1 on ram.
i'm just using 3:4 because i couldn't go with higher fsb.
i thought it would be better to use 830mhz than 622mhz ram.
February 15, 2007 4:21:06 PM

Quote:
A few things I can think of:

1) Try testing with the 8x multiplier. Many boards exhibit odd FSB behavior - they may be stable from 266-300, unstable from 300-340, stable 350-450, and so forth. You might be near one of those unstable zones, and the link above regarding bootstraps is the technical explanation for such zones.

2) Try incrementally lower Vcore values as well. Heat destabilizes overclocks. At a certain point, additional Vcore begins to lower the maximum overclock you can get. Heat dissipation is proportional to clock frequency as well as to the square of voltage.

3) Accept that e4300 is the lowest marketed bin - it might even be manufactured using a cheaper technique that still produces acceptable yields (we already know it's a cheaper die). You could try returning the chip or exchanging it for another one. We normally assume that since Intel's 65nm process is mature, they get high yields at or near the top bin (and very few lower-bin processors), and so a lot of their bulk mainstream processors must actually come from the same bins used for high-end processors - just not all.



well.. bad axe 2 doesn't support multiplier change.
although cpu's multiplier is not locked downward,
bad axe 2 doesn't let u override the multiplier in bios setting.
it's totally locked in my case.
i guess i'll try with lower voltage also...
and i also agree... it might be cheap e4300...
February 15, 2007 4:25:18 PM

Quote:
Thats about all you can expect from the Bad Axe boards without doing a volt mod. You can probably squeeze and additional 50 to 70MHz out of it, but you're reaching the max for 975X. The problem isn't your CPU, its the board.


but sources which reviewed bad axe 2 indicated it supports 1600+ FSB.
is that w/ voltage modification??
oh... maybe... this cpu hogs load of voltage to run in faster speed...
which requires over 1.6v... damn..
voltage mod = out of warranty??
February 15, 2007 4:28:38 PM

First off, run at atleast over 2.2v on vDIMM. Second, raise your vCore to 1.4. Drop the multi if possible to 8, and set you timings at 5-5-5-15
February 15, 2007 4:33:14 PM

ram is over 2.2, vcore is 1.5.
multi is locked @ 9 cannot drop in this mobo.
timing is 5-5-5-15..
w/ this setting, max i got was 2.8ghz.. oh and vcore was 1.55
February 15, 2007 5:17:54 PM

Quote:
Are you using 7x400 or 9x311 to reach 2.8GHz?


i'm using 9x311.
ram is set to 533, so ram's currently running @ 415MHz = 830MHz DDR2
Bad Axe 2 doesn't let you change the multiplier.
there is a menu about the multiplier, but it's @ automatic; you can't touch them for some reason...I'm not totally clear: Are you running your RAM at the same ratio (4:3) when you try to clock higher? Or, can you still not get higher with the ratio set at 1:1?
February 15, 2007 5:31:56 PM

Try updating the BIOS.
I've even seen where an older BIOS has helped certain folks.

Try to hardset all of your voltages to std and not any higher.
The exception may be RAM and just use what the manufacturer recommends.

At a FSB of 311 and a CPU speed of 2.8, you have not reached any point that more voltage should really be needed.

On my GigaByte 965, the Auto settings cranked up alot of my voltages but it ran just as stable and cool when I forced everything back to Auto.

Also try multiple higher FSB speeds. I have seen where MBs have issues with specific speeds but work at higher/lower settings. So if you hit a wall, try moving beyond it a few steps and see what happens.
February 15, 2007 7:16:59 PM

Quote:

so don't put the reference frequency of the memory on default?
what does reference frequency do??

The 975x has a number of "straps" which it can use to boot up and set timings and latencies for the chipset. Each strap is optimized for stability at certain FSB levels, with the E4300's 800MHz FSB the BadAxe2 is probably selecting a strap that is too aggressive to be overclocked to extremely levels. Going to 1066 (the default setting for the E6000 series) or 1333 would loosen the timings and should increase the maximum FSB you can reach.
February 15, 2007 11:57:47 PM

Overclocking is a gamble not a guarantee. Intel is not obliged to provide anything better than the stated rating. If you happen to get more, consider it icing on the cake. You got more than what you payed for; be content.
February 16, 2007 1:00:19 AM

Quote:
Are you using 7x400 or 9x311 to reach 2.8GHz?


i'm using 9x311.
ram is set to 533, so ram's currently running @ 415MHz = 830MHz DDR2
Bad Axe 2 doesn't let you change the multiplier.
there is a menu about the multiplier, but it's @ automatic; you can't touch them for some reason...I'm not totally clear: Are you running your RAM at the same ratio (4:3) when you try to clock higher? Or, can you still not get higher with the ratio set at 1:1?

4:3 setting is just for 2.8ghz... i was trying 1:1 for higher fsb.
still not getting higher w/ 1:1 ratio...
February 16, 2007 1:07:01 AM

Quote:

so don't put the reference frequency of the memory on default?
what does reference frequency do??

The 975x has a number of "straps" which it can use to boot up and set timings and latencies for the chipset. Each strap is optimized for stability at certain FSB levels, with the E4300's 800MHz FSB the BadAxe2 is probably selecting a strap that is too aggressive to be overclocked to extremely levels. Going to 1066 (the default setting for the E6000 series) or 1333 would loosen the timings and should increase the maximum FSB you can reach.

oh... ic... there was something i didn't know...
well i'll try that now...
i guess 1066... is 266 on reference frequency...
i'll try 266 or 333 now; thanx for the tip
February 16, 2007 4:39:34 PM

thanx a lot everybody!!!
i got it to work now!!
it reached 3.7GHz, but hot as going over 80deg...
so i just lowered the clock a bit... stable as rock.
prime 95 9+ hrs..
thanx a lot people!
kudos to accord99
February 16, 2007 5:02:37 PM

Mabye it could be faulty setup?
February 16, 2007 5:02:57 PM

Mabye it could be faulty setup?
February 16, 2007 9:02:08 PM

Wow, 3.7GHz that's awesome! No problem with the help, good luck with your new system.
February 16, 2007 9:10:40 PM

maybe you should edit the Post Title to say "RESOLVED"

When I read it, I thought that there was a newly discovered problem with the 4300 or that the 4300 OC was all hype.
February 17, 2007 5:30:27 AM

what did u do right now im stuck at 313fsb?? on the 4300. i tired multiple higher settings. PLz post ur settings?
February 17, 2007 6:31:33 AM

Quote:
what did u do right now im stuck at 313fsb?? on the 4300. i tired multiple higher settings. PLz post ur settings?


my current setting:
Vcore: 1.6v
FSB: 411MHZ
Multiplier: Auto (x9), cannot changed in bad axe 2
FSB voltage: 1.3v
MCH voltage: 1.625v
DDR2 voltage: 2.00v
RAM Timing: 5-5-5-14
Reference Frequency: 333mhz
Ram Ratio: 533mhz, 4:3

tip: 4:3 ratio w/ 333 ref freq comes out as 1:1 ratio for some reason.. a bug maybe??
when i apply 1:1 ratio w/ 333 ref freq, it actually comes out as 4:5 ratio...
maybe the bug... maybe now.. i dunno
February 17, 2007 2:53:28 PM

vcore at 1.6?? isent that like .3 over??
February 17, 2007 2:57:07 PM

is that a safe overclcok my normal cpu vcore is 1.325 seems like an awful lot to jump
February 17, 2007 2:59:50 PM

1.6 is definitely high. I thought intel said you shouldn't take Core2 over 1.55V.

My E6400 has a VID of 1.25V, and I get my current OC ( 425 * 8 = 3.4GHz ) at 1.3V. But much beyond this, and I really have to start pouring on the voltage.
February 17, 2007 3:44:12 PM

I wanna do a fsb 400. and get 3.6. how can safley do it
February 17, 2007 4:12:57 PM

it seems like e4300 CAN go high clock, but it seems like it eats more voltage than e6000series.
if u got a mediocre cooling and want not much of voltage, try ~3ghz
for 3.7ghz, u WILL NEED a decent water cooling system.
i am keep testing my system with variable clock speed;
i'm testing it right now @ 3ghz w/ 1.425vcore.
it seems stable at the moment, but need more testing.
besides, super pi result between 3.7ghz and 3ghz gives 3sec difference.
3 sec in superpi... a lot??
!