Overclocking limit? CPU or Mobo?

[OperativeSix]

Well, I managed to get my little P4 2.4c to 3.4ghz on stock voltage, and it's sitting pretty right now @ 283 FSB // 2-3-3-5 memory timings (3:2). If you're asking "well, is it stable?", yes, it is. It's been up for 5 weeks straight like this without a single error/crash. P95 DOES error out, but will not when I slacken my memory timings. Aside from that though, it's peachy.

However, there ARE issues with higher overclocks.

Going to a 288 FSB with slackened timings is OK, and I can run just fine there.

However, 289 is where my motherboard shoots itself. Every time I try to post at 289 FSB or higher, the mobo decides to corrupt the BIOS, and i'm forced to go back to my old 1006 betas. I've tried upping my AGP voltage and CPU voltage, neither of which allow for any high of an overclock than I'm already achieving.

So whats holding me back? My CPU, or the mothboard. Or is it something else entirely I havn't thought of?

CPU: P4 2.4c w/ Zalman 7000 Cu (1.55 vcore)
Memory: PC4200 GeIL Ultra Platinum (2x 256mb. 2.85 vdimm. 3:2 FSB:memory. 5:4 results in no POST, even at 3-4-4-8 :eek: )
Mobo: Asus P4P800 deluxe w/ Tiger1 chipset cooling.
PSU: 380w Antec

Mobo:
PCI/AGP is locked @ 66/33
Onboard sound is DISABLED
RAID chip is DISABLED
Spread spectrum is DISABLED

[shiyan]

not sure what's holding you back, but that is a great overclock already.

but you say that increasing the CPU voltage does not help, so it is probably not the CPU holding you back. Although I might be wrong on this point, as 3.4Ghz is a very good overclock.

it is stable for what you use it for, but it is not 100% stable, seeing how it fails P95.

this means you are near the limit of one of the components.

how long does it last in P95 before an error comes up?

is there any possibility for increasing the chipset voltage?

[OperativeSix]

P95 lasts for a few hours before erroring out. This is just due to the memory timings (2-3-3-5) on the GeIL stuff. It's meant to run @ 266 3-4-4-8, so 2-3-3-5 is stressing the hell out've it. If I loosen the timings to roughly 2.5-4-4-6, P95 sits perfectly happy.

AFAIK, it's impossible to change the chipset voltage. I know of a few boards where vAGP changes the voltage running to it, but i'm fairly certain this isn't one.

[stan03]

is it possible that your psu is too weak? i don't really know about psus, but that would be my only guess since everything appears to be ok.

[OperativeSix]

350w Antec

System devices:

2x HD
DVD-ROM
CD-RW
sound card
TV tuner
Wireless card
4x 80mm fans
GF4 4200


I have heard of P4P800's not being top overclockers sometimes...could this be a case of it?

[Filter]

i think you just hit the limit of the cpu. the o/c is amazing in itself.

be happy with it.

[OperativeSix]

I know its possible, but I just can't see how its possible.

The system is rock-solid up to 289mhz at stock voltages. The moment I take it to 290, the BIOS on the P4P800 corrupts itself, and requires a re-flash. No amount of added voltage can remedy this.

[stan03]

hmmm yea it doesn't sound like you hit the limit... how do your voltages look? and you did get the most recent bios right?

[PCGUY112887]

"Well, I managed to get my little P4 2.4c to 3.4ghz on stock voltage, and it's sitting pretty right now @ 283 FSB"

Sorry, i'm new to this, but it started with an 800FBS, how/why did it drop down to 283? Sorry again, still very new to this.

[OperativeSix]

800fsb = 200mhz x4
283mhz x4 = 1132mhz :cool:

P4's use quad pumped busses to essentially run at 4x the "real" FSB speed.

Processor is at 283.4x12 = 3400.6mhz (give or take a few. stats are according to CPU-Z)

For those wondering my voltages:

vCore: 1.568 (P4P's overvolt , and I bumped it up to 1.55 since it isn't adding any noticable amount of heat to my temperatures)

12v: 11.916v
5v: 5.026v
3.3v: 3.28v

Idle temperature is ~29 degrees (I keep the windows open :p )

BIOS version is 1010.02. I've tried other BIOS versions (1008, 1009, etc.). All crap out at 290FSB. Some crap out earlier! (1006 beta that it re-flashes to upon corruption)

[PCGUY112887]

OIC. Thanks for the clearup. You guys here are nice to noobs, I am a moderator on the "Spudtech" forums, they people are mean to noobs :-(

[shiyan]

well welcome to OC forums PCGUY112887!

I feel that it is simply wrong to be mean to noobs, as even the most experienced overclockers started out as noobs.

People who are mean to noobs should think about what it was like to be a noob.

[mirko_3]

vAGP increases chipset voltage...

[Zuzzz]

I think he is at the top of his mother board. Seems it would at least post. The fact the BIOS corrupts tends to make me lean in that direction as corruption of the BIOS shouldnt happen even with no or a bad CPU. What is the 865/875 chipset rated at? I have heard others talk of a so called 290 mHz FSB plateau on some mother boards. You might want to look into this. Seriously? SWEET overclock. I dont think I wouldn't mess with it (I cant get my 2.6 anywhere near 285mHzFSB. Wish I could get 270 stable.). Get it stable at 285 and be happy as that is just SWEET.

Good Luck

[OperativeSix]

Well, I guess its time for me to watch newegg refurbs for a IC7, Max3, or P4C board :P.

I'm planning to do the vdimm mod to my board within a week, so with 3.15v, i'm hoping this geil PC4200 stuff can do 300 1:1 like others have gotten.