Let's see break it down briefly:
OC'ing the proc to 3.2ghz is considered a moderate OC'ing for the E7300, that wouldn't do significant damage, the failure to P.O.S.T due to not enough Vcore to the E7300, commonly.. so that rules it out atm...
The Ram, its hard to tell bcoz I believe in this situation, the problem might not be the RAM, but the whole system configuration and the most apparent culprit atm is an un-updated BIOS, because I believe it "thinks" that a 4:5 divider is more stable than a 1:2.

The only way to narrow it down is if you could borrow another ram (800mhz) to see if the prob occurs, IF the 'halt' still occurs, then you at least have an idea where the problem comes from...the BIOS..
After you update the BIOS, there's still one more step, to see if the system stable after updating.
IF the 'halt' still occurs, then its the mobo, thats the only deduction that can be drawn from this troubleshoot...roughly.

Another thing not to be overlooked: The PSU.
Sometimes its just the PSU who's causing the menace, but it always slip people's minds because the PSU always comes across as the 'unusual suspect' from the distraction of toubleshooting the other things.(mobo,bios, ram, etc.)So if you can get your hands on another PSU atm, you should give it a try to narrow down the problem.
For testing, a fresh cap ayam PSU can be used, but just temporarily for testing only.

OK bro, thats most of the troubleshoots you can DIY for now.

I hope you can do most of the suggested troubleshooting, which is enough to figure out the main villain, or villains.
It's worth it to do all things boring and tedious procedures, because it will get you closer to finding out what's up


@farscope,
I'm still thinking what's going on with your situation atm, I wont say anything yet until I'm sure with the comments prior to you post.

---

Added on December 28, 2009, 2:50 amSometimes with a 1066mhz ram, with default settings usually it sets the ram to 800mhz, so need to manually set the ram clockspeed to 1066mhz...

It's just the bios...probably needs flashing...but can't tell yet...

This post has been edited by JoBigShow: Dec 28 2009, 02:50 AM

you mean u boot up it takes 1 second and to shutdown it takes 1-2 seconds..do you mean to the bios or to the Disk with OS fullly loaded?....

When power up, the system turns on (hangs) a second or two, shutdowns briefly for the same amount of time, then P.O.S.T. normally.
I've tried overclocking with two of my mobos, an Asus P5Q3, an Asus P5N-D and I think another Asus mobo, P5QPL-AM, and they all replicate the same process of 'P.O.S.T delay' that I have described, and that's something you should anticipated because how can any PC(not any that I know of) fully boot Windows in 2 seconds?

Ok, I might explained things a bit out of context regarding the 'reverting to default setting when PC is turned off', yes my imagination took the better of me when i try to explain it in analogy.
My, bad, I'm not a geek, I only able to project my interpretations subjectively or figuratively in that post.
Having said, I have spent hours researching on Asus mobo's (or at the least the one's I have overclocked) why it "delays the P.O.S.T." and can't find find any explanation, even ASUS themselves don't explain much of details.
(btw of course the battery saves the bios configuration, unless the mobo clears its RTC RAM by its self, which of course is impossible, unless the cmos battery is plain dead)

In essence, I tired to contribute to fellow forumer farscope what I know so far about this similar occurrence (which is natural process in my rig) if it ever to coincide in terms of my mobo's 'P.O.S.T. delay' behaviour,
in which pro-actively I'm hoping that it rectifies and not succumb to what we are sooo not fond off...hardware failure.

But then it occured that farscope's latest post indicates that my presumption were wrong, therefore narrows away from what I hope to be just a simple overclocking matter, into a situation more complicated and grim, but still trying to maintain that it is not a hardware failure, but of course foreshadowed by most people (who read his post) having the gut feeling it is a hardware failure, be it PSU or the motherboard, or etc. But I won't draw any conclusion yet until all possible troubleshoots has been done.

Anyways,
I hope you can donate an explanation prior to this so called 'post delay' on an OC'd configured mobo of mine? This would and will be very helpful.
Or more importantly figure out the solution of the distress forumer trying to figure out his situation... (farscope)

Why do you want to know my age? It's just an inappropriate question bro, especially with that sarcastic undertone.
Yes, I'm a 3 years old infant who manage to beat on a board with many buttons and think my parent's PC (that big metal box with string attached on the back and lights coming out from it) is a huge pacifier to play with and suck on


Jokes, thnx for highlighting my mistake in that post, will edited it later....

---

Added on December 28, 2009, 12:26 pmmobo you say... it's could be the capacitors, vrm, mosfet, etc.
The capacitors (those one's next to the 775 socket) regulates voltage stability to the proc.
One thing I like to share, I've read that it's highly recommended Enabling LCC for a 45nm proc, which is the E7400.
Probably when you set to default, the AUTO setting for LCC is 'Enabled',dunno if this is related, but could be...

What else do you notice within the span of a day bro?



This post has been edited by JoBigShow: Dec 28 2009, 12:26 PM

Aren't we all learners my friend?
It's good to educate each other for the sake of collective integrity of knowledge behold by the standards LY.Net..

This is sweet news, I recommend you enable the LCC, coz like i said before, its recommended for 45nm procs (like E7400)
Maybe this was the root of the cause of the problem....meaning did you remember that if you have the LCC enable/disable during the course when the Vcore fluctuated?
This is important data bro, hope you remember

I just briefly type 'fluctuating Vcore' on the the good ol' Google and the results are plentiful, vary from the problem causes by an OC rig with bad casing airflow, to PSU ratings, to capacitors of the mobo has worn off resulting instability to the voltage supplied to the proc, etc.
There's more answers out there (for your situation) than in this thread alone bro, so surfing around the web might be more fruitful at time to time....

Mine stable with 3.8ghz @ 1.31ghz
But to be stable for 4ghz(400 x 10), I have to bump up to 1.41v Thats a full 0.1v just to increase a mere 0.2ghz..so no point really.
I haven't given up though by lowering the FSB to 380, and multi to 10.5, but manage to stabilize it to 1.38v only.

IMO, anything above 1.3625v seems unnecessary for 24/7 (on 45nm) and desperate if you want to show off to you peers , but for benching, once I briefly pump the E7400 to 1.45v, thats unrecommended, but like I said, for benching only.
I hit 4.2ghz that time but I never posted it on LYN because I never documented it (screenshots, prime95, etc.) for proof/evidence, and don't want to bench it that far ever again because it WAS a scary experience.. (constantly in my mind thinking if the proc ever gonna die on me... )


Anyways bro, I believe you can get much lower vcore for 3.8ghz, keep us updated with your progress
Since this it like a lottery, I will place a bet on 1.34v

Lost bet
U cheated bro, I thought you gonna go for 400 X 9.5
Anyways 1.328v is a good result
U can go lower i think, but IMO i like to at least have a margin of 1 increment between the lowest stable voltage possible and instability...meaning my true lowest stable is 1.30v but I bump 1 increment to 1.31v.

Man my chip is not that good, its an angel at 3.8ghz, but a true deep~in~the~gut devil at 4ghz.

Yup, I tested 3.8ghz and 4ghz on SuperPie, 3Dmark Vantage and 06', etc. The difference are in the realm of 'micro'..
So 3.8ghz @ 1.31v VS. 4.0ghz @ 1.41v, is obvious who is the real winner. But thats just my E7400, most probably there's couple of E7400s out there can manage to stabilize 4ghz under at least 1.3625v...

It all depends on how much stress the plug's fuse can withstand. Its not gonna blow up, just need to replace to new fuse if it cuts out. 1 piece of fuse is like RM0.70 in Tesco
Make sure you spend well of power extension, especially those 'power surge' ones. When you say 'blow up', it usually cause by natural occurrence like lightning, so there's more reason you should consider what I've said.
My extension cost me RM89.90, but I'm free of any worries spent on that ext.

plush, once you've done with finding harmony in CPU OC'ing, now you still got one more big step to go... the PL as -pWs- mentioned (as in Performance level).
PL's purpose is to tighten up the whole system, meaning to tighten up the interaction between the CPU and the RAM and ICH, etc.

For Asus board, it's called 'Ai Transaction Booster'.

Usually the default PL for a system is '12'...you want to get stable at '8' as a starter.
To get to '7', you need to set the FSB to 5:4, so for your system the for instance to need to:
OC the RAM to 1004mhz, FSB @ 400, FSB Strap @ 266. I considered this one of the ultimate combo for C2D cpu with DDR2 800 ram, and P45 or higher chipset of course. I think P35 is considerable.
You can try other divider like 1:1, 3:4, 5:6, but for your hardware I recommend 5:4
If fail to boot, it's because one of the components is the weak link.
'6' is the closest to the 'holy grail' of PL tweaking, this one oftenly requires good mobo..(e.g. Biostar TPower, Maximus Formula, Rampage, etc.)
'5' is the best PL any system can achieve, but need pretty good hardware to maintain stability.
Below '5' are only threaded by Sifus and master overclockers.

I can go as far as '7', but a bit dodgy, so got it to '8' for everyday use, most probably due to my 'Kingston HyperX DDR2 800 mhz' being the weak link (OC to 1004mhz @ 2.1v)...need to get OCZ or Corsair rams to seal the deal..

Can buy those RAM sink, (range from RM9 - RM25) they do a pretty good job though.
Buy yeah, I didn't realize the guy using the good ol' KVR.
Probably without the heatsink 900mhz is alright, but need to get up to 1000mhz for best PL setting for C2D and Dual core CPU's.
FYI, there's a manufacture,(I forgot its name) produced a perfectly 1000mhz DDR2 ram (yes, not 1066mhz) to cater to the optimum 2 cores proc for PL tweaking.
There's something special about the combo of 400FSB + 5:4 + 1000mhz RAM to achieve low PL value...for 2 core procs especially.

@plush, I re-opened one of my closed thread which might enlighten you with this PL thingy, as well as RAM timing tightening, tRD, etc.

http://forum.lowyat.net/index.php?act=ST&f=5&t=1255768&st=0

This article, if you read it and study it carefully, you be in a totally new level of OC'ing:
http://www.anandtech.com/mb/showdoc.aspx?i=3208&p=1

OC'ing the CPU is just the first step like I mentioned, the final step is the whole system tweaking.

D9
These Epida chips are pretty lousy.
At least Kingston should use D9 for the HyperX series, but not.
Dunno where to find these chips anymore, it became pretty much extinct in the market during course of the year 2007 I think

Bro where did you bought these RAM?!!
I can't find them in KL, or probably I'm not looking hard enough.

I really need this RAM to get a lower PL since my Kingston HyperX 800mhz is "rattled under pressure", can't get my system to stabilize with them.
At least with those G-Skill I can get stable @ PL 7.
Underclocking a 1066mhz RAM doesn't quite do the job for me, i need a perfect 1000mhz RAM...

Can you recommend me where to get these RAMs ?(preferable in downtown KL)
I'm in the search for F2-8000CL5D 2GBPK, (2Gbs in enough for me)

No prob... but now I'm poisoned!
I'll manage to get those RAMs somehow



Anyone else who read this, I would like to propose a trade of my 4 months old HyperX with F2-8000CL5D 2GBPK

Don't worry, actually your 10066mhz is more than happy to be underclocked, it can run tighter timing without issue @ 4-4-4-12 with stability if underclock to 801mhz.

Run Everest Memory Benchmark, which is more faster:
1) 1066mhz @ 4-4-4-12 with default PL
2) 801mhz @ 5-5-5-15 with PL 8~10.
3) 801mhz @ 4-4-4-12 with PL 8~10> ( I doubt you can get it stabilize with this, but if you do, than it's awesome)
4) 1200mhz @ 4-4-4-12 with default PL
5) 1200mhz @5-5-5-15 with PL 9~11
ETC.

This is what I can think right now.....

1:1 is old school approach, really, it was made obligatory during DRR1 era if you want to OC the ram, DDR2 is a bit different nowadays..
5:4 is as great a ratio as 1:1 ( that is for 2 core procs with DDR2), so can be said for for other ratios depending on your hardware.

For everyday use I OC my ram to 1066mhz @ 2.1v with PL10 @ 5-5-5-15.

BTW, I'm pretty much off topic because this has nothing to do with OC'ing procs anymore, but the whole system performance.

I c, I never owned a Gigabybte board.. I only know much of the Asus mobos.
I lik Asus bios settings(specifically 'Ai Tweaker'), it's very nice to OC, especially to do with this PL thingy.

The PL thing you need some time of research hours, dunno what it's called in Giga boards, but for now should un~leash the potential of that HyperX 1066mhz.

Now that you mastered CPU overclocking, you should move to RAM OC'ing.
In RAM OC'ing, either tighten the ram timing or increase the clock speed while tweaking the voltage(vDimm) for stability.
(I dunno if the HyperX can do both tight timing and high clocks at the same time...at least stabilization is concerned.)
Then run prime95 or any proper memory stress test software to test the ram out.
After that, you need a memory benchmark test, e.g. 'Everest memory test & benchmark' to compare which configuration is best.

Sound complicated? Not really, just need time grasp the idea.
Even a guy like me who takes these thing as one of my many hobbies took less than 6 months to learn adequately to OC the PC..
Your HyperX is crying to be OC'ed, even a little is good enough...

Btw, I like that Old Town Hazelnut Coffee

The Vcore, still can increase up to 1.32v.

Anyways, I see the rams is downclocked to 750mhz to synchronize with the FSB of 375. Thsi is to maintain that 1:1 ratio.

Bro, try change the FSB to 400, with ram divider to 333, this will allow the ram to be set at it's default clockspeed of 800mhz.
And change the multiplier from 9.5 to 9, this will OC your proc to 3.6ghz, which is a pretty good OC.

Other settings like vdimm should be default speed, but it's better to set it manually than AUTO, find out what's the ram's default vdimm at kingston's site and set to it respectively. BTW I'm using the same RAM and I set it just a little above default: 1.84v.

Always monitor your temps bro when running a stress test. Make sure it doesn't go above 75c when running test like prime95, etc.


Bro, now the best thing is to always update here in this thread so we guys over here can help you out...

Ok,bro, I took couple of minutes Google'ing around and I come up with couple of rough ideas:
1) Most probably your E7200 needs lots of voltage pumps into the Vcore to OC above 3.6ghz. Some rough values I can give you is 1.34v~1.38v and above..
2) Your E7200 can't hold FSB above 375.(when combined with specific mobo, in your case Gigabyte) But I've check and some guys can't get the E7200 stable on 400X9 but stable on 420X8.5, do give this a try.
3) Your Gigabyte mobo needs to update bios, from what I've read the Gig mobo have quite an issue when paired up with the E7200..

Do read this, it's a review on E7200, on the section of Overclocking:
http://www.legitreviews.com/article/690/12/



IMO, I think you gotta do an "FSB hunt", meaning you have to find another FSB value other than 375 and able to be stable on higher fsb, AND with different multiplier, like 8.5 for instance.
But always keep in mind that the odds are that FSB can be set above 400fsb are possible, for example like 420 X 8.5, etc.

You have to try different combo of fsb X multiplier..

I get the feeling your situation is more trickier than most of E7xxx users here. An E7400 would be much easier to OC...

The load temps are pretty okay for 3.2ghz, well as you know it, all procs are uniquely different to one another, those guys who manage to OC with high clocks and low vcore are very fortunate...

Sweet, actually to get the FSB to 400 is an achievement! Now you can run your HyperX at its intended stock speed of 800mhz. (if you did select the dram freq. to 800mhz ,that is...)
btw, with an FSB of 400 and mch freq latch 400, you can OC your HyperX to 1000mhz , but that's for later...
I OC'ed my HyperX from 800mhz to 961mhz (you need to set the latch to 333 with fsb of 400 to get this...

The differences in Vcore value in Cpuid and bios probably caused by Vdroop. You can improve the Vdroop by: Enable LLC (Load Line Calibration) in your Giga's bios.
Actually LLC is encourage for 45nm chips, from and article I used to read, it's healthier for the chip. (too long to explain, just Google it up bro)
This goes for CPU Spread Spectrum, enable it for insurance, and disable it for CPU overclocking ability.
FYI, I set both LLC and CPU S.S enabled on my mobo, it doesn't interfere much at all with OC'ing...just that I would like to keep my proc healthy.

Anyways, your ultimate goal I think should get your proc to boot up @ 400X9.5, 3.8ghz. (at most below Intel's recommended specs of 1.3625v)
Although I think 400X9, 3.6ghz is good enough for the E7200.

So your next logical step is to get the proc stable at 400X9 (3.6ghz) in order to give some hints if you can achieve 3.8ghz...

I hope the Vcore that needs to stabilize @ 3.6ghz is much lesser than 1.3625v, because that way, it will give you much headroom needed to pump juice to your proc when you decide to go to the next step...oc'ing it to 3.8ghz.


This post has been edited by JoBigShow: Apr 6 2010, 04:12 PM

I think with Linx, the guy's temps might reach 100c with kind of Vcore paired with air cooling...
And if so, it will reach TJ Max and auto-shutdown during Linx test-run rounds

@deep_hunter
no need to try Linx if you are patient with using prime95, Linx makes the long hours of test in prime95 done in half an hour or more, but will rev-up your E7200 temp very high.
Btw, are you sure 1.45v is the lowest you can get stable on E7200?
Try lowering it by increments until you stumble on errors during prime95...I hope you can get lower bro..1.45v is indeed an unhealthy voltage for the 45nm proc...