Vid - PIN wire wrapping for Tualatin Celeron

[Pinky]

Much of the information below is a repeat from an overclocking article (http://www.overclockers.com/tips630/index.asp) that was written prior to the release of the Tualatin core. Besides the spec sheet/voltages available, little else has changed (even appears that the PINs are the same ;) ).

On page 61 of the Intel Specs Sheet you will find the diagram of the pins of a Tualatin Celeron. For a close up view, visit here (http://www.overclockers.com/tips630/P3pinout.gif). Please note VID1 and VID2 towards the upper right of the diagram. These are the PINs we will be wrapping.

Vid Pinning your cpu changes the default voltage for your cpu to a higher value, in this case from 1.475 to 1.675 With this additional voltage you will be able to run your processor at higher clock speeds. Beware that increased voltages will generate more heat and require better than retail heatsink cooling. Perform voltage modifications at your own risk!!

There's some additional information/discussion here (http://www.asusboards.com/forums/showthread.php?threadid=7009) and here (http://forums.overclockers.ws/vb/showthread.php?s=&threadid=44329).

This is relatively simple, but careful attention needs to be paid to only connecting the proper pins together and ensuring no stray wire touches the surrounding pins.

Personally, I used a single strand of copper speaker wire cut to about 3 inch length. Created a tiny hook with needle nosed pliers on one end, and patiently wrapped the pins, pulling very slightly to keep the tension. Once wrapped 4 to 5 times, I pulled with a little force. The first few times pulling the wire caused the wire to unwrap and I had to start over, but it's necessary in order to keep them tight once you flip the processor and place it in the socket.

I cut the remaining wire away and ensured the left over wire would not become loose. In this (http://forums.overclockers.ws/vb/attachment.php?s=&postid=451467) picture it appears that the wire does not run a risk of touching the other pins (NOTE: this was not a picture of my own wrapping ;) ).

This takes a little common sense, a slightly steady hand, and much patience. Total time: 15 minutes. Results (http://www.aoaforums.com/forum/showthread.php?s=&threadid=1017): 1530mhz (127mhz bus speed) at 1.82V on an Abit VH6T motherboard.

Good luck!

[ol' man]

Thanks Pinky! We have needed this for awhile:)

[Pinky]

Originally posted by ol' man
Thanks Pinky! We have needed this for awhile:)

Actually, it wasn't my idea, but I was more than happy to help :)

[zoopa_man]

Hey bud. Thanks....If you want give me a call. I'll take you out for a few beers :) You diserve a few, nice work :beer:

[Yodums]

Nicely written Pinky.

Thanks for taking my idea into consideration :D

Yodums

[Pinky]

Originally posted by zoopa_man
Hey bud. Thanks....If you want give me a call. I'll take you out for a few beers :) You diserve a few, nice work :beer:

I drink coffee, if you're buying then I'm drinking :highlycaffeinated:

[oldfart]

Its great we have a sticky on this. IMHO, a much better method is the pin hole method. Bend a tiny strand of hair fine wire into a U. Drop it in the VID1 and VID2 holes in the socket. The CPU will still drop in as it normally does. This way you don't run the risk of the wire touching the other pins, it also allows the CPU to sit all the way down in the socket instead of being raised up slightly in one corner. It is also a heck of a lot easier than trying to wrap wire around pins.

[Yodums]

Originally posted by oldfart
Its great we have a sticky on this. IMHO, a much better method is the pin hole method. Bend a tiny strand of hair fine wire into a U. Drop it in the VID1 and VID2 holes in the socket. The CPU will still drop in as it normally does. This way you don't run the risk of the wire touching the other pins, it also allows the CPU to sit all the way down in the socket instead of being raised up slightly in one corner. It is also a heck of a lot easier than trying to wrap wire around pins.

Sounds really easy and not risky where the wire is like really small and doesn't create any bumps when everything is hitting the flat surface. Where the twisting and turning the wire gets me worried since I'd be afraid of doing it wrong or bending a pin or something.

Yodums

[Pinky]

Originally posted by oldfart
Its great we have a sticky on this. IMHO, a much better method is the pin hole method.

I just don't trust laying something in there... I need something reinforced in some way... it seems that it would be easier for a piece of wire to "slip" that was placed in a socket rather than wrapped tight around the PINs...

The first article I link to covers the pin hole method.

[Sempei]

Originally posted by Pinky


I just don't trust laying something in there... I need something reinforced in some way... it seems that it would be easier for a piece of wire to "slip" that was placed in a socket rather than wrapped tight around the PINs...

The first article I link to covers the pin hole method.
I took ol' mans advice and used the pin hole method. Once the CPU is locked down there is no place for the wire to go. I've even pulled and reinstalled my chip 3 times and the wire stayed in place each time.

[oldfart]

Agree with the above. I also used the method recommended by ol' man. It worked perfectly. Once the CPU is in, there is no place for the wire to go.

[takiwa]

I've tried the pin hole method, and wrapping...and I have to agree with Pinky on this one...wrapping those pins just give me a sense of security, even if it is a false sense. Both methods work, so it falls back on the old adage "to each his own"...whatever works, right?

[gg]

I tried laying the wire method with too thick wire and broke a pin off A variation on this theme would be to take apart the 370 socket and place the wire under the top of the plastic 370 socket.
Haardcore users could purhaps solider the pins on the reverse side of the board.

[Pinky]

Originally posted by gg
I tried laying the wire method with too thick wire and broke a pin off A variation on this theme would be to take apart the 370 socket and place the wire under the top of the plastic 370 socket.
Haardcore users could purhaps solider the pins on the reverse side of the board.
-----------------------------------------------------------------
XXXXXXXXXXX.XXX

That's the best method IMHO. But, alas, I don't solder (yes, some of us don't :p ) so it would not be an option for me.

The concern I expressed regarding placing wire in the socket is not what happens once the CPU is inserted, it's while the CPU is inserted, and that it could push, pull, tug, etc the wire loose if not done properly.

There's risk involved with both methods, I find less risk with wrapping since I can clearly see and have more control over the movement of the wire...

BTW, gg, that link you have/had in your post is probably not allowed by forum rules. You may want to edit your post to remove it ;)

[Blimpyboy]

Has anyone got the wire trick working with a powerleap pl-ip3/t adapter?

I have the adapter + celeron 1.2Ghz, and have done the wire trick by wrapping a thin piece of wire around VID2 and VID1 but get no post at 112Mhz FSB. I've tried the 'make a u-shape and put it in socket holes' method as well, and it didnt work either.

has anyone got it working *with* the powerleap adapter?

[Yodums]

I remember someone here I forgot who modded their adapter to be able to do some voltage adjustments.

Yodums

[Blimpyboy]

Originally posted by Yodums
I remember someone here I forgot who modded their adapter to be able to do some voltage adjustments.

Yodums

Yeh I read the thread about modding the adapter. I want to know if it works without modding the adapter and just doing the wire trick.

[Pinky]

Originally posted by Blimpyboy


Yeh I read the thread about modding the adapter. I want to know if it works without modding the adapter and just doing the wire trick.

Sure it would, the mod to the adapter is only so you don't have to continuously remove the CPU and rewire to change voltages... they both do the same thing

[Blimpyboy]

Originally posted by Pinky


Sure it would, the mod to the adapter is only so you don't have to continuously remove the CPU and rewire to change voltages... they both do the same thing

Sure this is true in theory. But I'd like to hear from someone that its actually worked for them (using the adapater). I have the feeling that the powerleap adapter is setting the voltage based entirely on the VID signals like it should.

[Pinky]

Originally posted by Blimpyboy


Sure this is true in theory. But I'd like to hear from someone that its actually worked for them (using the adapater). I have the feeling that the powerleap adapter is setting the voltage based entirely on the VID signals like it should.

All I can say is scan recent posts on tualatin celerons, somewhere in there a member (gets into a long discussion with Ol'Man) about doing this mod, even took photos of the results... I was pleased to see his results... and that's the only place I base my information from. I can't say I'd blame you for being skeptical, it took me 3 different threads (here and elsewhere) to finally try the pin mod at all :p

[zoopa_man]

Ok, I finaly got the....ohh what's the word.....Insanity.......Courage....Guts......To do this mod.


I have one question....How do you know if your sucesfull ?

I have an abit ST-6 and don't see any more voltage options in my bios so I assume you have to add .2 v to what you see in bios or how does this work or did I not get it right ?

I used a small piece of wire and droped it into the vid1 vid2 holes in my socket..maybe their not touching ??? or did I get it right...I just need to add .2v to what I see in bios ?

[takiwa]

if I'm not mistaken, you should flash youtr BIOS after this mod...your BIOS will pick up your vid-pinned vcore as defualt, and give you new voltages to choose from in the menu...if I'm wrong, someone straighten me out on this one...

[Sempei]

Originally posted by takiwa
if I'm not mistaken, you should flash youtr BIOS after this mod...your BIOS will pick up your vid-pinned vcore as defualt, and give you new voltages to choose from in the menu...if I'm wrong, someone straighten me out on this one...
No bios flash is needed. After I vid pinned my ST6R, the default voltage in the bios showed up as 1.675 with options up to 1.8v. If you're not seeing this, something is wrong.

[zoopa_man]

I got it. :D Check out my other thread to see how i'm doing. Titled Tualatin 1.2@1.44 Stuck

The first time I had it...just didn't realize it until I went into bios and set the voltage option to default and saw the 1.65 instead of the 1.45 :D

Let me just say this too......What a rush...Goofing around with $200 worth of bran spanking new stuff is insane. LOL I see why this is so addicting.


Thank You for all your help.

(Maybe one of the mods can kill this so our sticky doesn’t get so cluttered. I would be helpful to show how you know you did it right though....that was the only thing I noticed was missing. Other then that a stupendous job Pinky. Thank You)

[Pinky]

At least this one won't go to the lab, eh Zoopa ;) ??

We should leave successes here, who cares how cluttered it gets after the first post :p

[zoopa_man]

Some more celibration clutter then. hehehe

I made my mark 133 fsb !! Temps are way down right now...I turned off the heater in my room and let it get cold. 19C room temp. Inside ambent case temp is 22C and CPU temp idle is 32C full load 35C Looks like i'll either have to wear a coat in my room or work on my water cooling setup. So far at this speed I'm ok as long as my temps dont go too high. I usualy crash / reboot at 41-43C range.

another note. I had to go to the full 1.75v to get 133 to post....maybe after a burn in I can lower my voltage. that will help with the heat issue some. I'm also going to try some higher CMF fans on my radiator to get more air moving through it.

Othen then some tweaking to do I'd call this a sucesfull overclock. Now I have to feed my addction with case fan mods and playing with a neon light and some dye lite. (Future projcts :D )

Thanks for your help.....looks like this one will live :cool:

[Pinky]

Burn in earned me 2 more mhz from the bus (24+ mhz processing power) which allowed my components to run closer to specs (127 bus versus 125).

You may see some improvement from a burn in (including lower voltages at current-highest clock speed)... just remember the keys to burning in... 1- Slowest clock speed (that means underclock it to 66mhz bus), 2- Highest voltage possible, and 3- Coolest temperatures... I usually do it for 24+ hours, but you will reach a point it won't improve any more, I did.

[ErazorAMG]

Does this vid trick also work with the p3-s tualatin? Have the same vid 1 and 2 to be connected? And are the pins at the same position like the celeron tualatin?

[John1975]

I did the vid pin thing on my 1.2 celly and it brought my voltage up to 1.7v. I can run for able 5 mins when cold at 1.6, but then it crashes. I am still stuck at 1.5 :(

[Pinky]

Yes, the P3 version is an identical chip with more cache (512) and lower cache latency, but same core and PINs.

This only increases the default voltage, but then you can up it .125v at .025v increments... if you can't run 1.6ghz at 1.7v you'll have to raise the voltage some more.

[Mit Map]

Hi, I used the pin hole method for my Celeron 1A. It was so easy. It took my about 2 mins to get 12 Mhz more on FSB.

1A@1.47G 1.85 Vcore
Volcano 7 w/ AS II
Asus 7100Magic
SB Live
ANTEC SX830 CASE (I love my case so much!!!).

[groo4ever]

I would like to thank Pinky. Thanks to him I'am able to run my Tualeron 1.1A at 1.47GHz.

Muito Obrigado (That's "thank you very much" in Portuguese)

PS: Used the "pin hole method"

[Rainman]

.

[Pinky]

Originally posted by groo4ever
I would like to thank Pinky. Thanks to him I'am able to run my Tualeron 1.1A at 1.47GHz.

Muito Obrigado (That's "thank you very much" in Portuguese)

PS: Used the "pin hole method"

I am only the messenger :)

[CANiSLAYu]

Does anyone have pictures of how they did the pin hole way?

[Blimpyboy]

Originally posted by CANiSLAYu
Does anyone have pictures of how they did the pin hole way?

No picture but all you have to do is bend a small piece of wire into a U-shape, and drop the legs of the U-shaped wire into the two holes. Makes sure the wire is thin enough. 0.2mm is good.

Thats all there is to it.

[CANiSLAYu]

Where do you get really thin wire?

[Pinky]

Originally posted by CANiSLAYu
Where do you get really thin wire?

A single strand of speaker wire works well.

[CANiSLAYu]

So I just drop it in there and that's it? Do I make it so it goes down all the way into the bottom of the socket? I did the wrapping way, but it sticks up a bit when I put it in the socket. When I tried the pin hole method it didn't stick up as much. I just want to be extra sure.

[CANiSLAYu]

How does this look?

http://www.bol.ucla.edu/~imaddela/pin.jpg

Left side to show how long I made it, right side how looks when put in.

[bldegle2]

duh, so easy, single strand out of very small stranded wire, stuck in zif, reinsert. Even with the MC370 in da way, five minutes total

Tually 1.13 running as high as 185FSB so far, 1573MHZ. Actually much faster @180FSB.

Thank you, no electrons were used in this upgrade.

baldy:cool:

[56k]

i`ve successfuly done that pin wrapping thing to my cel.1200 and now i can easily select the max 1.825 voltage in bios but i still can`t get 133fsb (@1600) to boot up (emergency beeps during post) Maybe it`s because i`m using the standard cpu cooler and the cpu temperature is higher than 52C? the highest fsb speed that seems to be running stable on my comp is 125 (@1.575V and the temperature is usually about 45-46C)

[JBatOC]

Originally posted by 56k
i`ve successfuly done that pin wrapping thing to my cel.1200 and now i can easily select the max 1.825 voltage in bios but i still can`t get 133fsb (@1600) to boot up (emergency beeps during post) Maybe it`s because i`m using the standard cpu cooler and the cpu temperature is higher than 52C? the highest fsb speed that seems to be running stable on my comp is 125 (@1.575V and the temperature is usually about 45-46C)
If 52C is what you getting out on POST, I'd say it's way too hot....even if you could boot in, it won't stay too long in Windows. Yeah, definitely try a better HSF.....I've been told that Millennium Glaciator 2 is very effective at low cost.

[bldegle2]

you need better cooling. temps way too high. For comparison, I am running my Tually 1.13 @1.825, 180FSB (1.530 gig), temps are about 36*C with a MC370 from Swiftech.

Swiftech, Alpha, both cool to the max.

Also, you may have a chip that won't do 1600, you got to think about that.

baldy:cool:

[surfinguru]

canislayu, excellent pics! as they say, a picture is worth a 1,000 words.......

[LesterP]

surfinguru - who's riding that wave?

[surfinguru]

Dude! That's me!!:D

Seriously though, the guy is Shane Dorian. The picture was from "The Eddie" big wave contest held last year at Waimea Bay.

Here's the whole picture....

[LesterP]

excellent surfer - awful actor.

[surfinguru]

In Gods Hands. Bah! What a gay movie.....Hollywood trying to cash in on the big wave tow in phenomenon......

Keep an eye for a new movie comming out called Liquid Assets. It's by Bruce Brown's son Dana Brown. (Bruce did the both Endless Summer Movies.) Should be much more respectable than that debacle called IGH's

[Peter007]

Originally commented by Pinky on why "Wire-Wrap is prefer over Pin-Hole"
____________________________________________
That's the best method IMHO. But, alas, I don't solder (yes, some of us don't ) so it would not be an option for me.

The concern I expressed regarding placing wire in the socket is not what happens once the CPU is inserted, it's while the CPU is inserted, and that it could push, pull, tug, etc the wire loose if not done properly.

There's risk involved with both methods, I find less risk with wrapping since I can clearly see and have more control over the movement of the wire...

BTW, gg, that link you have/had in your post is probably not allowed by forum rules. You may want to edit your post to remove it _____________________________________

I FRIED 2 Celerons doing the Pin-Hole Method!

Apparently, A strand of the thinwire GOT PULLED under during the CPU insertion and touch all kind of VID1+2+3 etc....
and create a Dangerous Voltage to Fried my CPU

[bldegle2]

It couldn't be easier the hole way.

Get very thin copper wire, cut, bend, insert, then add CPU.

With the proper gauge wire the CPU reinserts with little effort.

Helps if you have a heatsink that clips on, I am using an MC370 by Swiftech.

Total time, about two minutes to install, daze thinking about it.

baldy:cool:

[ShiFtY2001]

The pinhole method works well, only took a few minutes, practised first with an old celeron on a slocket.

One thing i found useful was to put a mark with a pen near the relevant pins on the bottom of the cpu, much easier to visualise where the right holes are on the socket and avoid that crispy frying CPU sound.

also check your voltage in the hardware monitor in the bios, dont want to over volt that cpu too much without good cooling. I am running 1.75V while burning in with prime95 torture test, hope to drop it later.

~ShiFtY out~:burn:

[G Hunter]

Hi Guys its also good to hook them up on the back of the board.
I did this with one of my boards next I am going to add a dip switch or jumper so I can adjust it with out removing cpu and or
mother board

[Prodius613]

OK NOW

EVERYTHING I READ HERE IS FOR A P3 WHAT ABOUT A CELERON IS IT THE SAME?

[@crilicM@n]

Hmmmm, first good new idea i´ve readed in the latest minutes of (me too...) my tualerons upgrade guides reading :-) Even with a not so thin wrap, it can be soldered by any man with at least one good eye :-) in the back of any slotcket... Hmmm that´s the way I´m gonna do (just don´t know when :rolleyes: )
About insulating the other 3 pins, hmmmm? Silicon works?

Thanks

Originally posted by G Hunter
Hi Guys its also good to hook them up on the back of the board.
I did this with one of my boards next I am going to add a dip switch or jumper so I can adjust it with out removing cpu and or
mother board

[Rio71]

Here the Results of the vidpinning are cel-t 1300 on TUSL2-C.
Unfortunately that is not the desired effect:

default (w/o vidpinning):
VID25mV=0
VID3=1
VID2=0
VID1=1
VID0=1
V-Core Range=1.500V-1.800V

vidpinning VID2 to VID1:
VID25mV=0
VID3=1
VID2=0
VID1=0
VID0=1
V-Core Range=1.600V-1.800V

vidpinning VID2 to VID1 to VID0:
VID25mV=0
VID3=1
VID2=0
VID1=0
VID0=0
V-Core Range=1.650V-1.800V

was this wrong :confused:
As i can increase the V-Core OVER 1.800V
has someone an idea :(

EDIT: TUSL2-C Rev. 1.04 ; Made in Taiwan
Bios 1010 ; 1011 ; 1011@Mad --> all the same

[RIPDOTCOM]

I need the Vid pinout fro 1.7v, 1.75, 1.775 and 1.8v

I have a Tualatin Compatible mainboard however I have no voltage settings in the BIOS. I have vid-1 and vid-2 connected for 1.675v but I need more juice so help a brother out please.

[sour]

ftp://download.intel.com/design/celeron/datashts/29859602.pdf

see page 21 for the table.

[Pinky]

Originally posted by Rio71
was this wrong :confused:
As i can increase the V-Core OVER 1.800V
has someone an idea :(

No, the motherboards followed the specs handed by Intel and the max they permitted motherboards to run was 1.825v

There are probably power supply level mods that could be tried, but I never bothered and upgraded to AMD and blew the celeron's performance away :p

[Chris_AD]

The PIN Hole method seems good, but it's brand new, so if someone could get a picture of it... Plus this might sound stupid, but i'm new to all this, Where's VID1 and VID2 ?

Also, In my BIOS I can see the voltages, etc, but they are greyed out, so I can't change them ? Or it might just be me, Any idea's on this please ?

Thanks,
Chris !

[o770]

good to see no-one here burnt the chips.
i doubt if removing the mod will still leave the voltages available in the bios. i have a TUSL2-C and heard that the wires could be removed after the voltages were released. is anyone sure if that will work?
thanks.

[Pinky]

Originally posted by oTTO
good to see no-one here burnt the chips.
i doubt if removing the mod will still leave the voltages available in the bios. i have a TUSL2-C and heard that the wires could be removed after the voltages were released. is anyone sure if that will work?
thanks.

Otto's correct, if you remove the wires it returns the processor to default voltage. The wires are changing the default value, but this is auto-detected during each bios startup, not one time after the initial cmos setup.

[ataxy]

want to know what pin is what check this out
http://forum.oc-forums.com/vb/showthread.php?threadid=89864

[ataxy]

vcore pin setup

[GeForceG]

Is this vcore mod safe to do??

[Edward2]

Yes, but you need to keep an eye on your CPU temps because they will increase as you increase the Vcore voltage. The general concensus seems to be that you can go to around 1.8 Vc using air cooling, as long as your CPU temps are around 38-43C (loaded).

See my sig for my setup.

[Bonka]

The voltage mod is very safe to do, barring that you read the instructions carefully. The extra 0.25 increment is crucial.

Personally, to eliminate the chance that cooling is the problem for your Tualatin to not to be stable, I recommend keeping your load temps at 40c or below. This should be with a 1.8/1.825v setting in the bios.

[Yoda715]

Hi guys, got a quick question, Im currently running a Celeron 633 o/c to 894 on a Asus Cusl2-c Black Pearl Edition. I know that it does not officially support Tuatulin core's, but will my board accept it with or without doing the voltage mod? I want to get a Celly 1.2, and stick it in this board, and it seems to be that they will not burn up at 1.65v, which is the lowest my board goes. I think that it should be fine, but I just wanted to see if anyone knew what I should do? :) Thanks

Yoda

[lclark2074]

i want to try this a abit bx2 can eney one tell me if this will work igot a slockit II this will be my last upgrade fore it

[Aihyah]

not officially supported = u gotta do the pin trick.

[Homer]

i was reading the article about the pin mod and it says to lower the voltage as low it can. can i just leave it at default then do it?

[Pinky]

Originally posted by Homer
i was reading the article about the pin mod and it says to lower the voltage as low it can. can i just leave it at default then do it?

I cleared the cmos before doing it.

[False Christian]

Any way to vid mod the 1.5 volt Tualatin Celerons? In my case it's a 1.4GHz one with a defaukt voltage of 1.5 not 1.475. I've tried wrapping vid1 to vid2 but, although that worked on my 1.2GHz Celeron, it doesn't work on my 1.4GHz chip.:eh?:

[Pinky]

Look on intel's site for the spec sheet on the 1.4ghz chip, be sure the vid pins are still in the same spots/assignments...

[puppet]

I have stumbled on another approach to the actual pin-wrapping.
It is particularly useful for "long distance" connections.
I did ths yesterday and it works pretty well ... here goes.

Using the same gauge wire Pinky describes , you wrap it once around a needles' shaft. The needle diameter should match processor pin diameters. A touch of solder on the loop holds the ring shape and won't stick to the needle. Do the same to the opposite end of the wire.

Between the loops lay out a piece of tape (Scotch, masking,etc.) and fold it over the wire (in half lengthwise). Trim the tape back to the wire leaving enough for insulation. Slip the loops over the pins to be connected and lay the wire down between the pins without fear of shorting at all.

If the needle diameter is just right ... the loops go down with enough friction to stay put.

[JCLW]

*** DO THIS AT YOUR OWN RISK - IT'LL VOID YOUR WARRANTY ***

This will work on any Tualatin core (celeron, PIII, PIII-S) with a core voltage of 1.45, 1.475 or 1.500.

So you want to raise your default (cold boot) CPU voltage?

What we're going to do here is connect VSS to VID0, and VID1 to VID2. This will make them all "0" in the following chart, with VID3 remaining "1". Note that with a 1.45 or 1.475 chip all you have to do is connect VID1 to VID2 as VID0 is already "0".
http://www.cansailexp.com/jtemp/wt_vid_chart.gif

#1: SET YOUR PROCESSOR TO ITS DEFAULT SPEED AND VOLTAGE If you have the voltage set to +0.30v in BIOS you'll get +0.30v (ie: 1.95v) after the mod when you first boot up which could damage your processor.
http://www.cansailexp.com/jtemp/wt_bios_start.jpg

#2: Shut down your PC and pop the hood. Ground yourself on the case. Remove the heatsink. Remove the memory if it's close to the socket.
http://www.cansailexp.com/jtemp/wt_socket_start.jpg

#3: With a marker identify VSS, VID0, VID1, and VID2. Draw a line from VSS to VID0, and VID1 to VID2. I won't be offended if you look at another guide to double check the position of the holes to be marked.
http://www.cansailexp.com/jtemp/wt_socket_marked.jpg

#4: Use a utility knife to make a notch/trench along the lines.
http://www.cansailexp.com/jtemp/wt_socket_notched.jpg

(Continued below)

- JW

[JCLW]

(Continued from above)

#5: Find some small stranded wire. Tinned copper wire is better then copper because it is more corrosion resistant. I used the speaker wire that came with my Creative Labs Soundworks speakers.
http://www.cansailexp.com/jtemp/wt_wire.jpg

#6: Cut off 8~9mm of wire and use a thumbtack (or anything else that's small and handy) to bend the wire into a "U". Repeat.
http://www.cansailexp.com/jtemp/wt_wire_jumpers.jpg

#7: Drop the two jumpers into the two notches/trenches you cut in step #4.
http://www.cansailexp.com/jtemp/wt_socket_wired.jpg

#8: Put your processor back in place. You may have to push it a little to get it in the socket but don't push too hard! If it won't go try shifting the jumpers around in the holes. If it still won't go you may have to use smaller wire to make your jumpers. Don't forget your heatsink and memory!
http://www.cansailexp.com/jtemp/wt_bios_after.jpg

No processors were harmed in the making of this article.

BTW: Here's which pins are which when orientated as above.
http://www.cansailexp.com/jtemp/wt_vid_ident.jpg

- JW

[f00t]

Are the voltage changes the same for all tualatin chips (celeron and p3). I ead that to achieve a default voltage of 1.65v on a 1ghz celeron A chip, I had to conect the VSS and VID25MV pins, which I did and it fried my psu. What pins do I conect on a 1ghz celeron A to get a default voltage of 1.65v.

[JCLW]

They are the same for all Tualatin chips (celeron, PIII, and PIII-S). Connecting VSS to VID0, and VID1 to VID2 is a universal mod that will work on any Tualatin core (as of time of posting).

If you know you have a 1.475 default Vcore chip you can make it easier and just connect VSS to VID0.

There isn't much point in playing with the VID25mv pin. 0.025V won't make a huge difference in the grand scheme of things. However, connecting VSS to VID25mv should not wreck your PSU. Are you sure you connected the correct pins? In fact on a 1.50 Vcore chip VSS is already connected to VID25mv.

- JW

[f00t]

I am very sure that I connected the right pins. I even checked about 5 times on the cpu and on the mobo.

[False Christian]

Thanks, JCLW. I really appreciate the time and effort you put in to help me.

Again, thank you very much for the help.:beer:

[macklin01]

Nice guide! :D

As a small point, the PIII-S has a default voltage of 1.45V. -- Paul

[JCLW]

Originally posted by macklin01
As a small point, the PIII-S has a default voltage of 1.45V.

Thanks - fixed now :)

- JW

[macklin01]

So, this method would get a tualtin working on an 810 board, for example? Too bad you can't control the FSB speeds on an HP board .... :( -- Paul

[JCLW]

Originally posted by macklin01
So, this method would get a tualtin working on an 810 board, for example?

It's part of it, but not all of it. Flip through this thread: http://forum.oc-forums.com/vb/showthread.php?s=&threadid=67417

- JW

[Pinky]

Thanks for keeping this updated JCLW, I traded my tualeron a long time ago.

[BadCop]

what vid pins do i have to use for lowering the vcore ?
i underclock a cel 1200 to 66fsb (800mhz) and want to get the lowest possible vcore . plz help me connecting the correct vid pins

[JCLW]

What vcore do you want?
Which vcore do you have to start with?
Are you working with a Tualatin board (VRM 8.5) or Coppermine board (VRM 8.4)?

Between 1.125 and 1.275 is easy.
Between 1.300 and 1.450 can also be done, but will be a little more work because some of the pins will need to be insulated.

- JW

[BadCop]

how about 1.05v / 1.1 v core? isolation of pins is ok, if i have to do it. dont know what vrm i have, board freetech p6f135 mini-itx (official tualatin support).

[JCLW]

FYI: The LV PIII-800-S runs at 1.15v and makes ~11W of heat if that helps at all.

If your board is Tualatin capable out of the box it's VRM 8.5

Option #1: With a Celeron 1.500v tB1 1200 jumper VID3 t0 VID25mv to get 1.100v.

Option #2: With a Celeron 1.475 tA1 1200 make a long jumper between VID0 and VID3 - it will cross over the VID25mv hole. You'll get 1.050v if the VID25mv pin touches the wire when you put the processor in, 1.075v if it doesn't. 0.025v shouldn't make much difference.

I'd go with option #1, myself.

- JW

[f00t]

I finally got my celeron 1A working with my vl6 mobo. I isolated the three pins and did the vid-bridge with the four othe pins. The only way it boots is if I set the "cpu hardwire ioq" to 1 instead of 4. with the 1 setting enabed, it only runs at 75%-80% of its full potential (3Dmark 2001 score=about 5500 with 8500le/ 1ghz@1.4ghz+256mb pc133). At 1.4ghz, cpu runs at 40C after 4hours of mohaa and 1.65 v-core.

My question is: Would the bridge between the ak4 and ak26pins do the trick and should I connect the ak4 to ak26 or the ak4 to an11.

I know that this proc has great potential and I would like to bring that potential out with full force.Thanks in advance for any help.

-f00t-

[JCLW]

Sorry, I didn't realize you had a non-Tualatin board and that's why you were playing with the VID25mv pin. I understand what you were doing now.

As for AK4, there has been a lot of discussion on it and you'll probably find your answer in one of these two threads:
http://forum.oc-forums.com/vb/showthread.php?s=&threadid=96915
http://forum.oc-forums.com/vb/showthread.php?s=&threadid=67417

- JW

[f00t]

Thanks for the reply. I wil do it as soon as I get some conductive paint (very rare in my part of Canada) over Christmas. Then it should perform at its full potential.. :)

[JuMpY]

Just wanted to thank everyone for their help. Got my celeron 1.1A to 1540 (140fsb) @ 1.675V using the pin hole method, because I haven't got voltage options on my board :cry:
Previously was only stable at 124fsb with default voltage. Pin hole method was easy, just used some copper speaker wire dropped into VID1 and VID2. Thanks guys.

BTW - my first post :)

[JCLW]

Originally posted by JuMpY
Just wanted to thank everyone for their help. Got my celeron 1.1A to 1540 (140fsb) @ 1.675V using the pin hole method, because I haven't got voltage options on my board :cry:
Previously was only stable at 124fsb with default voltage. Pin hole method was easy, just used some copper speaker wire dropped into VID1 and VID2. Thanks guys.

BTW - my first post :)
I'm glad it worked well for you. What is the sSpec (http://support.intel.com/support/processors/sspec/icp.htm) of that 1.1A?

Welcome to the forums :)

- JW

[JCLW]

I've had a few questions along the lines of:

"I have a 1.500v processor and some people say just connect VID1 to VID2 and others say connect both VID1 to VID2 and Vss to VID0 like in your picture post. Which way is the right way?"

Depending on the voltage you start with, different pins will do different things. My post was a universal mod that would make any Tualatin processor request 1.650~1.675v.

On the VID pin chart in that post, the 1.500v has zeros in the VID25mv and the VID2 columns. Anything connected to Vss (the processor ground plane) is "0".

If we connect VID2 to VID1 it will make VID1 a "0" as well. Scanning down the chart, we see that with VID25mv, VID2, and VID1 all "0" we get 1.625v.

If we connect both VID2 to VID1 and Vss to VID25mv it will mean everything except VID3 is a "0". Looking at the chart we see that this will give us 1.650v.

So both methods are correct, they just give you different results. Starting with a 1.500v processor, only connecting VID1 to VID2 will give you 1.625v, and connecting VID1 to VID2 as well as Vss to VID25mv will give you 1.650v.

(A 1.450 or 1.475 processor already has VID0 as "0", so connecting it to Vss won't do anything.)

I can tell the next question is going to be: "So which should I do?"

In my opinion, you might as well make the jump right up to 1.650v because it will help you get a little more speed out of your processor.

Then again, if you have a 1.0A, 1.1A, or 1.2 and "only" :) want to run at 133FSB and use silent cooling you might want to try 1.625v first.

- JW

[JuMpY]

JCLW - Only a SL5ZE. Haven't seen any tB1's out here yet (australia) but I saw a SL68P 1.2 the other day, a newer tA1. Now if I just had some money.....
::gets angry at all my relatives for needing expensive xmas presents :mad: ::

[Spajky2]

Hello to all you guys! I am running more than 8 months Tuallie on modded Slotket & my 1.0A CPU can go TL stable @1,6GHz w 1,8Vcore in another setup (friends one), but there disipates more than 50W of heat (double than original!) & there needs serious cooling! Intel does not recommend higher Vcore than long term still safe 1,75V & myself i do NOT reccommend more than 1,8Vcore at any case & doing Burn-In slowly with time upping frequency & Vcore with better cooler than stock one /takes weeks!/! WHY?
Because you do NOT want to damage your CPU like last developed case with P4 CPUs known as "Sudden Northwood Death Syndrome" which can develop for Tuallies also if hurrying & esagerating too much. Both CPUs have the same manufacturing process /its fragile/ & same thermal protection, so the "normal" heat is not such a problem, voltages are & upping Vcore more than max reccommended is long term contraproductive!
And remember: the REAL inside core temps are much higher full load than revealed even from probe mounted under cooler tight to IHS; the difference is sometimes even more than 10°C! With in-Socket probe on MoBo could be even 15° higher or more-depends on how the software is calibrated!
Running Tuallies at high real temps is contraproductive /think on summertime & decent airflow/ because even if nominally is set to certain frequency it starts throthling down & slowing performance, so it is better some MHz & Vcore less & getting better stability & overall speed even at lower nominal speed at extreme using condition & so you can dump all those PC noise down a bit too by slowing fans...so providing your CPU survival for longer time; it is really a wonderful CPU (especially if mounted on BX MoBo..)
Thermal Throthling kicks in @75°C /thermal hystheresis 75-80°C/, after 80°C starts slowing down tremendously; at 85°C is sloooowwww like an invalid turtle, 486 cpu is faster than that!!! All these temps I measured days ago from CPUs inner thermal diode (& Win98seLITE did not crash!). Think about it! :-)

[JuMpY]

Spajky, tuallies don't have thermal throttling built in like the p4's.

quote: "All these temps I measured days ago from CPUs inner thermal diode (& Win98seLITE did not crash!)"

are you saying that you got your tuallie up to these temps, are you f#%kin crazy?

[Spajky2]

Sorry guys, I have to correct myself, I was wrong about thermal throthling; there is only thermal protection built in: today I discovered that this is not a fact of my Tuallie, but a Feature of my MoBo...there is nothing written anywhere in the owners manual or elsewhere mentioned that I would know, that additional auxiliary thermal probe input on my MoBo has such effect! Even if having this mobo for more than 3 years, I still discover time to time something new about it. Well this can be for me a nice & usefull feature! :-) ..

@ JuMpY: >are you saying that you got your tuallie up to these temps, are you f#%kin crazy?<

:-))))))))))) YYYYEEEEEEEEEEEEEEESSSSSSSSSSSSSSSSSSSSSSSS !!!!!!! :-))))))))))))

with OC-ed Tuallie 1.0A@135fsb/defaultVcore & well done AirFlow thru case & very silent machine!

The max working was 88°C if I do not count 97°C for few minutes while checking calibration of
my measuring Gadget with Cpu on the desk & praying god, that later will still work in my setup!:
http://users.volja.net/jerman55/AdaptPreProtTyp.jpg with HairDryer like this:
http://users.volja.net/jerman55/HeatingFront.jpg Gadget schematics:
http://users.volja.net/jerman55/CpuDtempAdapt.jpg

All additional 3 fans in my closed desktop setup are normally running using 7V fans setup like
this-rpm sensing enabled! & the case:
http://users.volja.net/jerman55/rpmSensFan7v.jpg
http://users.volja.net/jerman55/CaseInside.jpg

During tests, fans were controlled by SpeedFan, except PSU thermalControled 6-9V one, which was running all the time. Temps are shown like this: MoBo=MoBo; Motherboard=hotest air pocket inside; Cpu=MoBo-sensor=Socket-sensor 2-3°C cooler than sensor mounted under HSF closest to IHS; (Amb+)AUX= REAL inside core temperature taken from CPUs On-Die diode!

The system was running idle normaly with additional fans @12V, than @7V, than @5V & than WITHOUT them-
this last FULL LOAD!!! :-) While using fans, the full load Cpu temps were just a bit /2-3°C/ higher than
on idle! Here are snapshots:

http://users.volja.net/jerman55/12vFanTemps.jpg
http://users.volja.net/jerman55/7vFanTemps.jpg
http://users.volja.net/jerman55/5vFanTemps.jpg
http://users.volja.net/jerman55/NoFanTemps.jpg

With this last the temperature of a heatsink passed 60°C, vhich diminushes using fans, but it is still about 10° less than real core temps with normal cooling, so everybody should add this value to their readings to get approx. actual core temps! I really wonder, how high are the AMDs core temps... Ouch! :-) if Tuallies can be so high! Any additional comments? :-)))))

BTW, do you now that non-OC-ed Celly400 with only stock heatsink NO fan still normally works @80°C.. :-) ..

[macklin01]

Spajky2, those were some interesting schematics on getting the fan RPM sense to work on the 7V mod! I thought about that problem for quite some time awhile back when designing my fanbus. Ended up just using the method of varying the + voltage with a pot/NPN transistor setup.

BTW, I would appreciate your opinion on whether my description here (http://forum.oc-forums.com/vb/showthread.php?s=&threadid=141022&perpage=30&pagenumber=4) on
how the RPM monitoring works is accurate or not, as I'm always trying to get a better understanding of that.

Thanks! -- Paul

[Spajky2]

@ macklin01: not to be too much OT here, lets move to the link you gave...to the last page

[macklin01]

sounds good, and thanks. ;) -- Paul

[muddocktor]

I'd like to thank JCLW for the nice pics and explanations he posted in here on the socket trick. I pulled my P3 1.26S system down today to install the components in a better case and decided to do the wire in the socket vid mod on my GA-6OXET board, which has the cold boot problem when overclocking too high for the default vcore of my P3S proc. After doing the socket mod and getting the rig back together, I was able to go into bios and set my fsb to 166 and enable the pll divisors to keep my PCI bus speeds down and successfully run. Previously, the highest speed I could boot up air cooled was around 158-159 fsb and not use the pll divisors at all. I'm going to leave it at 166 fsb and 1.75 v vcore for a few days folding and check overall stability but I think this proc can still go higher air cooled. About a year ago, I had this proc up to a 180 fsb using water cooling and a 120 watt pelt but I never could get near this speed before using air cooling.

[macklin01]

An amazing aircooled overclock. way to go! -- Paul

[muddocktor]

Yeah, I know that I got a pretty good chip with this proc. Another thing that is helping me out is this mobo I'm using, which has the pll circuits that give me 1/5 and 1/6 divisors. That way, my PCI bus doesn't go screwy on me with these high fsb speeds. I'm cooling the proc with an original Kanie Hedgehog, which was one of the first all copper hsf's on the market. It's working well; my cpu temp is only 34 C while folding right now.

[Techno Pride]

how do I make a "0" into a "1"?

insulate that pin? I need more than 1.65v for my 1.5v chip

[JCLW]

Yes, to turn a "0" into a "1" you must insulate it.

Some methods people use include:
- Nail polish on the pin(s)
- Remove the contacts inside the socket (but then they're gone for good)
- Wrap pin with tape
- Find some very small wire (ie: 80 pin IDE ribbon cable) and strip a bit of of the sheath off and stick it on the pin
- Or you could just remove the pin, but I'd only do that as a last resort

- JW

[BadCop]

at last i got my mini itx board and managed to underclock my 1.2 tul tb1 to 12x71mhz with 1.1vcore :) , lowered the cpu fan voltage to 5v . cpu temp never goes over 38.

[macklin01]

Hi, and welcome to the forums! Sounds like a nice, extremely quiet setup! How's the performance hit? -- Paul

[mike111]

How do i get 1.725v and 1.75v on a tualatin compatible mobo(ga60xt-a)?
I already did vid1 to vid2 connection which gave me 1,65v,and can only go upto 131fsb.
I have a 1.1a sl5ze and can go upto 140 on a bx board using 1.75v

[asw7576]

My three SL5ZE Celly 1.0A runs 1450 at 1.475V, and all of them actually could run 1500 but I lost the onboard soundcard due to 150FSB. These chips run on generic mobo ECS P6-IPAT. Stabililty is 100% stable with Mushkin High Performance Rev.3 SDRAM PC133 @ CL2-2-2-5/7. The temp. is 33C - 39C.

I have two SL68P Celly 1.2A runs 1680 at 1.600V with Gigabyte GA-6OXTA. My OC is currently limited to 140FSB due to generic SDRAM or incompatibility problems with Hynix Modules. As soon as I get SDRAM replacement ( no more Hynix!!! Micron 7E Only) I will test again to hit 1740 MHz. It boots at 1740 but I see so many dll. warnings.

[mike111]

you are right.My problem is the generic sdram.I replace the 1 of the 2 256/133 sticks i have and i can go upto 139 fsb but not very stable .i get errors all the time and this have to do with my crappy ram.
it seems that gigabyte mobos dont like the generic rams.

[macklin01]

It's probably more a case that the generic (PC100?) RAM doesn't like the overclock. -- Paul

[@crilicM@n]

Originally posted by Peter007
Originally commented by Pinky on why "Wire-Wrap is prefer over Pin-Hole"
____________________________________________
That's the best method IMHO. But, alas, I don't solder (yes, some of us don't ) so it would not be an option for me.

The concern I expressed regarding placing wire in the socket is not what happens once the CPU is inserted, it's while the CPU is inserted, and that it could push, pull, tug, etc the wire loose if not done properly.

There's risk involved with both methods, I find less risk with wrapping since I can clearly see and have more control over the movement of the wire...

BTW, gg, that link you have/had in your post is probably not allowed by forum rules. You may want to edit your post to remove it _____________________________________

I FRIED 2 Celerons doing the Pin-Hole Method!

Apparently, A strand of the thinwire GOT PULLED under during the CPU insertion and touch all kind of VID1+2+3 etc....
and create a Dangerous Voltage to Fried my CPU


:rolleyes: Wow.... sorry man... odds of courage...
To avoid this problem I usually put a very very small amount of simple glue in the minuscle space between the holes, where the "u" stays... I can shake and turn the adapterup/down... and it sits there...

[gamefoo21]

k i am a retarded goof that hasn't gotten enuff sleep this weekend and stayed up way too late, but to get my cele-t 1.4 to a 1.8 vcore??? which pins?? i am way too confused...

[gamefoo21]

ok ok i learned, sumthing discard that last post, but why is the wiretrick acting wierd on my mobo it has official tua support. i mean i can hit a 2.075 vcore. mind you my slot is a universal slot(blue) the mobo is only supposed to support vcore of 1.8.

[Ponte]

Is it posible to connect some of the other Vid pins to get higher voltage? Like connect 1 and 3 instead of 1 and 2?

[Ponte]

Or mabye isolate some of the pins?