WIP: Karaktu's (and Zap's!) Unofficial Guide to mobile Celeron overclocking success!

[Zap]

Latest word from Karaktu

FYI, the Vintage moterboard is a rev 2.01 and has an AMI BIOS instead of an AWARD that is NOT overclocking friendly. 133FSB is about all you can expect from it.

 

[Karaktu]

Originally posted by: Zap
Latest word from Karaktu

FYI, the Vintage moterboard is a rev 2.01 and has an AMI BIOS instead of an AWARD that is NOT overclocking friendly. 133FSB is about all you can expect from it.

Vintage info added to OP. Forgot to mention in that PM to you Zap that you also have to bend the pin.

 

[ender11122]

can you guys post what your Motherboard temp is in bios? Mine is at 39C. Is this to high and could that be limiting my OC?

 

[Karaktu]

Originally posted by: ender11122
can you guys post what your Motherboard temp is in bios? Mine is at 39C. Is this to high and could that be limiting my OC?

That's a perfectly normal temp. 30-40C idle and 40-50C under load would be completely normal.

 

[DJMiX]

I had a dead Celeron 400 MHZ PPGA 370 ...

That I guted and removed the heat spreader ...

Haven't tried it with a moble celeron yet...
I can post pics if anyone wants..

 

[greenwar]

Got a new board working with mobile celeron 1.8ghz.

Asus P4P800S-X
Intel 848P chipset

Posted without pin mod, voltage defaulted to 1.6ish.

I did the pin mode, now running one of my mobile celery 1.8@2.4ghz at 1.2x volt.

Incidentally, my other mobile celery in the Abit AI7 1.8@3.0Ghz scores 7095 in PCMark 2002. I thought its fairly good score.

Finally, thanks to karaktu for giving the mobile celeries new life.

 

[TStep]

greenwar, was a pin mod needed to get the AI7 to boot?

 

[greenwar]

TStep, Ya, AI7 needed the pin mod. It didnt boot up without the pin mod.

 

[Zap]

I made my own spacer today in under 1 minute. I took a floppy disk with a metal shutter and trimmed the metal a bit. Worked well.

BTW, my mobile Celeron in my main machine locks up when it goes somewhere over 52ºC (that's what it says in BIOS after immediate reboot). Guess my low RPM NMB fan isn't enough... only happens when I play Zero Hour.

 

[Karaktu]

Good idea, Zap. I bet you can make several heat spreaders from one floppy disk.

I have ender11122's board/chip coming to me...we'll see what the problem is.

 

[TStep]

Zap: Did one layer of floppy shutter do the trick? What brand? Also was this enough to take up the full space utilized by the IHS so that say, a stock, flat bottomed, run of the mill, all aluminum, Intel hsf would work? Bascially fail-safe installation with any hsf is what I'm asking.

Kudos for using the obvious!

 

[greenwar]

Do you guys know if the Bus/Core ratio play any role in overclocking success. I was reading that P4 2.26 533Mhz D1 stepping easily goes to 3.4 at 200x4 mhz fsb. Its Bus/Core ratio is 17. Does that mean that it can reach a bus to core ratio of its bigger brother p4 3.06 400mhz which has a bus/core ratio of 24?

As you know fastest northwood p4 is 3.4 at 800mhz which has a bus/core ratio of 17. I was just wondering whether that it something to look at while picking cpus for oc.

Please shed some light on this. I would greatly appreciate your input.

 

[Zap]

Originally posted by: TStep
Zap: Did one layer of floppy shutter do the trick? What brand? Also was this enough to take up the full space utilized by the IHS so that say, a stock, flat bottomed, run of the mill, all aluminum, Intel hsf would work?

Don't know what brand floppy, just one that used a metal shutter without printing on it. It ended up about 1" square and I used it in a Shuttle SS61G mini system. That heatpipe cooler is different from the previous ones in that there is a raised ridge on the bottom around the CPU and the fins at the back are larger than usual, and the edges of the bottom don't quite clear the corners of the P4 retension bracket. The thickness of the shutter was just enough to make it fit, plus make the heatsink clamp down tighter.

Do you guys know if the Bus/Core ratio play any role in overclocking success.

Don't know, but seems "easier" to push CPUs that start out at a low bus speed, probably because when you clock up say a 2.4C chip you will start to push your motherboard and RAM (depending) because you start at the maximum 800MHz FSB while a mobile Celeron 2.4 starts at 400MHz FSB so you will hit the maximum of the CPU before the motherboard/RAM tops out (not counting using RAM dividers).

 

[Karaktu]

Originally posted by: greenwar
Do you guys know if the Bus/Core ratio play any role in overclocking success. I was reading that P4 2.26 533Mhz D1 stepping easily goes to 3.4 at 200x4 mhz fsb. Its Bus/Core ratio is 17. Does that mean that it can reach a bus to core ratio of its bigger brother p4 3.06 400mhz which has a bus/core ratio of 24?

As you know fastest northwood p4 is 3.4 at 800mhz which has a bus/core ratio of 17. I was just wondering whether that it something to look at while picking cpus for oc.

Please shed some light on this. I would greatly appreciate your input.

Ummmm....

All chips of the same core/stepping are going to have the same limitations. Therefore, something with a lower multiplier is going to overclock on a higher FSB, and vice-versa.

The maximum overclock for most C1 stepping mobile Celerons is 3GHz at 1.6v or 3.2GHz at 1.7v (on average). It doesn't matter if you start with a 1.5GHz CPU or a 2.4GHz CPU, your top attainable frequency stays the same.

So, your two 1.8GHz mobile Celerons should do 166/167FSB for 3GHz, and MAYBE you can push them a little more.

 

[greenwar]

Thanks guys for your reply regarding the bus/core speed issue. I read karaktu's posting mentioning that D1 stepping didnt prove to be very successful in overclocking. Is that what everybody else is seeing? I am trying to get a D1 mobile celeron to test it out first hand. If you know anything about this let us know. Thanks in advance.

 

[DJMiX]

I got a moblie 1.5 celeron to be recongize by a ECS L4VXA2 v1.0C HT support
WITHOUT any pin mods..

I can't do other testing since I don't have a P4 heatsink fan, I used a pIII heatsink to see if it would boot that's it.

update on this Board:

Mobile celeron worked ok, but died not sure why, but I had a p4 mobile lying around and it posted without any pin mods:

Some good points about this board:
HAs a jumper on the board to set fsb to 133
vcore can be change
bios fsb chan be changed from 100 - 166 by 1 mhz intervals.
memory voltage can be changed:

currently running at 161 fsb on a p4 mobile

Not bad for a ECS board...

 

[Zap]

Originally posted by: DJMiX
Anyone have a cheap P4 HEatsink they can spare (for the cost of shipping ?)

I have one... that may need some work. You'll need a spacer, plus I think one of the latches is missing. I'll check when I get home after the LAN gaming tonight. Failing that, you can get a decent cooler for $4.99 plus shipping from SVC.com, Arctic Cooling Silencer 4 TC. It doesn't need a spacer.

 

[DJMiX]

Originally posted by: Zap

Originally posted by: DJMiX
Anyone have a cheap P4 HEatsink they can spare (for the cost of shipping ?)

I have one... that may need some work. You'll need a spacer, plus I think one of the latches is missing. I'll check when I get home after the LAN gaming tonight. Failing that, you can get a decent cooler for $4.99 plus shipping from SVC.com, Arctic Cooling Silencer 4 TC. It doesn't need a spacer.

Zap thanks, but I found one off the free stuff forums here...
I appreciate it.

 

[Jon187]

Originally posted by: greenwar
Thanks guys for your reply regarding the bus/core speed issue. I read karaktu's posting mentioning that D1 stepping didnt prove to be very successful in overclocking. Is that what everybody else is seeing? I am trying to get a D1 mobile celeron to test it out first hand. If you know anything about this let us know. Thanks in advance.

I always found the D1 cores were the best OC'ers. I have run several at 3.6-3.7ghz whereas any C1 CPU seemed to die around 3.2ghz.

Jon...

 

[locab]

I got Mobile Celeron 2.0GHz SL6QH and able to POST in my ASUS P4S533-X with no pin mod. CPUID reported it running at 1.63V. First couple of times able to boot into XP and then crashed after a little while. Last few times, PC just shut down as soon as it is powered up. Able to catch CPU temp via BIOS - hitting 80+C.

Can anyone suggest a HSF that does not require shimming ?

Is there a way to bring back down voltage to 1.3 ?

P4S533-X has no CPU voltage adj in BIOS, but I am willing to perform pin mod if require.

Any suggestion is appreciated.

 

[TStep]

Well, Zap found that the shutter from a floppy worked, but I haven't tried yet. I have had success with the stock Intel hsf that has the round copper core that is slightly raised, the cheapa$$ AOC EC-NW6-715CA, and Thermalright SP94. Heatsinks that didn't workout for me were the stock Intel all aluminum flatbottoms and Thermaltakes.

 

[greenwar]

I am using a Thermaltake spark 7+ hsf for my mobile celery. Its awfully loud, but comes with two fan controllers. Keeps my celery pretty cool at full load, around 43 degrees when the fan is kept to 48xx rpm. It works awesome. I did not have to use any additional thing to get it to touch the cpu core. Best part is I picked it up from ebay for $7 bucks used.

Check ebay, you might get lucky like me.

 

[TStep]

Geesh....I have a Spark7 w/ the X fan controller. Maybe it's user error on my part or may the mobo heatsink bracket was slightly taller. I'll have to recheck.

 

[Zap]

The floppy shutter can also be used on HSFs that don't need it to make them press down harder on the CPU.

 

[TStep]

Originally posted by: Zap
The floppy shutter can also be used on HSFs that don't need it to make them press down harder on the CPU.

Sounds like a good excuse to try it w/ the Spark7. I've used aluminum roof flashing in the past, but not since I've found the few hsfs that work straight up. Green: is there alot of pressure when you use the Spark7 or is it barely making contact?