TIPS TO SUCCESSFULY RUN A TUALATIN CPU IN A BX/SOLT1 MOBO NOT TUALATIN READY.

For the do it job, please read the James Anderson article, it has a very good "how to" in practice... I collected down here some of the most interesting posts from our mega giga thread about running a Tualatin in a BX mobo and from some threads outside the forum. (or, slot1 genericly speaking...) It contains many users solutions for many different troubles that came up when trying to run a Tualatin processor in a non Tualatin ready motherboard. It contains info about mobos, slockets that work, mod´s, advices, solution to problems and alternative mods to the tradicional ak4-ak36 bridge... Excuse-me for typing errors and if it´s not exactly "beaultifull design" archive, I did not pay attention to it, just to the content... The images are linked to their respective links... And some links will soon be outdated, but the words are there and the oc.com forum "protagonists" have their names in italic before each comment... I bolded some text, that´s about what I would like to say about that thread and added some comments... I expect this saves you from a 3 day of reading our thread as I made and go directly to the "Cream of the crop". No need to say who do you have to thanks about this info... The "protagonists". Sorry for my english.

@crilicM@n

Go here for the How to with photos from: James Anderson

 

Other Considerations

VOLTAGE:

                Zbone
"All you have to do to get a default voltage of 1.675 is wrap VID1 and VID2. If you need the Intel spec sheet on the desktop Tualatins let me know...I can either give you a link or e-mail it to you. I had to go up to about 1.65V to get stability at 167 FSB(1.5 GHZ) on my ST6."

jmsandrsn I'll just go ahead and give you the link since I don't have your e-mail address: http://developer.intel.com/design/p...ts/24976501.pdf
"The voltage chart is on page 19. If you look at the 1.475V setting you will notice that only pins VID0 and VID1 are connected to the VSS pin(it's done internally and there's no way to change it). If you then take a look at 1.675V you will notice that the only additional pin that is connected to VSS is VID2(all the other pins are the same). You can indirectly connect VID2 to VSS by actually connecting VID1 to VID2.

To see where the pins are located, you can get a diagram on page 59. Notice that VID1 and VID2 are both pretty close to the edge and easy to get to. There's probably no way to wire VSS to VID2 directly since they're not next to each other. Once again, connecting VID1 to VID2 does the same thing."

soil "use a dot of non-conductive expoxy to seal the AK4 connection"
'insulation lacquer (if you don't want to cut it)¡C'

 

G Hunter "isolated an3 ak4 aj3 with small peice of black tape and held my breath when I put the power to it. "

Michal[mt] "The matter is: you have to insulate 3 pins(VIDS) and they are:
AN3, AK4 and AJ3 but at the same time you have to make a direct link between the AK4 vid and the AK26 one. How is it possible? Simple: making a mistake...I mean you can t connect the AK4 and AK26 on the MB or slotket background cause you disconnect the AK4 first so the only possible method is to directly apply a tiny insulated electrical wire on the cpu background. This means that the cpu must be seated a pair of mm higher on the socket in manner to don t damage the connection between the two vids.
Personally I DO NOT suggest to do this mod...
Un salutone "

Ruiner "EUREKA!
I am writing this on a Celeron 1.0a on an abit slotket III, on a Via Apollo Pro133a slot 1 mobo...non b-step.
The bios is mistaken on post....6x166=1003, but it reads Vcore as 1.47.
I slipped the insulator from a thin phone wire over the 3 pins, connected the other two pins (AK4-AK26) with coated 26G wire (see previous posts) and did the pin hole trick for Vcore (vss to Vid4).
The strange thing, I can't get into regular windows. Just before I hit the desktop in win2k, a quick box (dos maybe?) flashes by and it restarts. I can get into safe mode. I disabled all startup items, but still only safe mode. The chinese site did mention having to reinstall windows...maybe this is why.
I'll fool with it a bit at stock speed and overclock later.
Core temp reads 18C in bios. So does chipset temp. Both are below ambient, so I don't believe them."
"You'll have to do a pin-wrap or pinhole trick to set the default voltage on the cpu. It won't boot properly at 2.0.
Jumping Vss to vid4 will give a default of 1.47...perfect. I used a short 26G wire, stripped the insulation off the ends, bent it in a 'u' and put it in the holes.
It gave me the right voltage, but I've had other problems, as above.

Ruiner has attached this image:"

Ruiner "Good news guys!!
I fiddled with the bios, and got BOTH chips to work. I disabled the bios virus check and it booted! Does this mean I have a virus?"

Yodums about pin hole method: "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."

Sempei "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."

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."

belinassu after a noon successful boot prob..."It works!!! It works !!!
i'm writing in regular windows desktop with a TUALATIN on
a BX133!!!
i modified the value in bios called "In order queue dept"
from 8 to 1 and it works!!!
and listen ,the only trick i'm using now is pulling off the three
red pins no othe pin insulation or connection (neither the vid trick)
the deafault voltage is 1,45v anyway on bx133!!!!
Only need to insulate/pull off three pins on BX133!!!!!
i also skinned the cpu , the cover (it is in copper , not alluminium)
is good but they put inside a awful pad between core and cover
so i prefer pull off that also..
now yhe cpu is cooler (27-30°C instead of 31-34°C)
great!!!!"

belinassu about in order queue depth x stability after reconected ak4-26 "ok it is fine now... after connecting ak4 to ak26 i colud get in win98
also making "8" in "In-Order Queue Dept"
I also connect VID4 to Vss using that silver conductive paint
and the system seem better (same voltages shown but
less temperature)"

ebc
quote:

Originally posted by belinassu
Hi ebc , i found your suggestions very useful in other forums !!
but now i can ensure i'm running at 1,45v default on BX133
without doing the wire trick on Vid4 and Vss!!
I only insulate AN3, AK4 and AJ3, that's all..
according to intel's datasheet Vid4 is already connected to Vss
inside the cpu , infact it boot at 1,45v naturally
That's my personal experience, i don't know if it is the same
thing on other motherboards...
my problem is that the only way to make stable and workable the system
is putting "1" instead of "8" in "in-order queue depth"...
and this give poorer performance....
the max overclock i may achieve is 1333 (133x10) at defult
voltage (1,5v seen as 1,48v)...
no higher voltage could help... it is runnung at 33-34°C
with Intel's coppermine boxed fan/heatsink...
and i lift away the cpu cover (now it is similar to a coppermine cpu)
any suggestions?
-------------------------------------------------------------

"1.About voltage,this is you case,I guess your mobo is old enough,the Vid4 is connect to Vss in you mobo,not the cpu inside,i found C 1GA five voltage pin not even one connect to Vss inside cpu,you can measure it yourself,in "new" old mobo,the Vid4 do not connect to Vss in mobo,it will cause voltage go to 2.0v and even higher, when you plug the cpu in and boot the pc.so we recommend connect vid4 to Vss,aslo connect vid2 to Vss to get 1.5v voltage,the method test in my cpu and mobo,my mobo is ms-6309 v1.0(694x),for the via chipset problem,it can boot but can't enter windows.
2.I am not a oc expert,I can't help you,you do better than me."

ebc "It's strange, another man from taiwan oc forum also measure the C 1GA CPU five voltage pin,the result is same to me,at first,I aslo regard the Vid0 and Vid1 had been connect to Vss in cpu internal, I suggest only connect Vid25mv(vid4) to Vss so can get 1.45v voltage,but in fact,some people only get 1.3v voltage,and the measurement result explain the reason. my cpu model is SL5VP(OEM), made in malay.there,several man don't connect Vid25mv(Vid4) to Vss and get 2.0v voltage and even higher,because quite a few mobo not connect Vid25mv(vid4) to Vss.for my poor english, i can't explain more detailed."

 

ol' man "Here is my idea of how to disable a pin in a socket after you pop the cap of the ZIF off. I actually suceeded yesterday with this method to increase my vcore to 1.825v on my NEW ST6. PLEASE YOU PEOPLE DO NOT RIP OUT PINS ON YOUR CPU'S FOR OBVIOUS REASONS ABOVE WITH BELISAU(sp.)

ol' man has attached this image:"

As has been mentioned you may experience better stability if you connect ak4 to ak26

 

chosenfool "interesting things to note:

• the 1.2GHz@1.54GHz chip im using is stable at 127MHz FSB, but not as stable either lower or higher (unless i go to default 1.2GHz).
  
• the tualatins have a default voltage of 1.475V. this chip im using is not stable past 1.95V - craps out at 2.0V.
  
• Temp rises 8C more with a switch from 1.75V to 1.8V.
  
• Its stable at 1.75V (current voltage setting). I may try for lower voltages later.
  
• only the 3 pins were needed to be disabled to make this work on a slocket with voltage selector. The same method doesnt work on a generic FCPGA adaptor, though. That needs further modding with the connection of AK4 and AK26, which i may try later."
  

Comment: Some generic slotckets work, some not... The newest ones are more proned to do....

Xaero_toast "I have an abit slotket III, an Iwill Slotket II, and an Asus s370...
they all have voltage, busspeed, and cpu type (ppga/fcpga) jumpers.
The Iwill is the most stable, with the Asus as a near second YMMV"

 

chosenfool "thanks for the tip, ol'man. im just about to do that ak4 to ak26 connection...
actually i have two projects. seeing as the abit slotketIII has been reported to be unstable past 127MHz FSB, ill switch to my Iwill Slocket II (yep, i have both! )
Ill be modding the Iwill with the 3 gored slocket holes for now. I wont be connecting the two aforementioned pins yet, , to see if this stability issue is truly isolated only to the abit slocket, as Xaero_toast and Ruiner said.

here is another way of making the 3 holes bigger so the 3 sleeved pins will fit: USE AN X-ACTO KNIFE. After taking off the slocket top, simply insert the tip of the small exacto blade in the hole and start twisting the knife. This causes the hole to get carved out, and eventually gets big enough for the thick pins to enter. ill post a pic of it as soon as i snap a shot.

Also another way to make the ak4 to ak26 connection is to solder the pins ON THE BACK. Its already been mentioned here, but ill take it a bit further, taking again from ol'man's tip. I used the same IDE wire, cut it to about 2 inches in length, and instead of removing the thin wires from it, i just stripped a small portion on both ends exposing the wires. I then tinned both ends, so whenn i solder it on the back of the slocket it will easily stick. Also this prevents overheating, as it may damage the plastic on the slocket by melting it if youre not careful. (heh, thats funny - worrying about damaging the slocket, when the whole procedure CAN damage the chip and mobo for using an out of spec chip! heh heh! so ironic... )
ill post pics of these as soon as i get back."

 

ebc (About the above message from chosenfool) "your method to connect ak4 to ak26 is wrong,it will cause pc can't boot,the right one is the ak26(pwrgood signal) must be send to the pin of ak4 not ak4's solder and the same time the ak4 must be insulation,do you understand my meaning,sorry for my poor english."

 

Romeo "Hi friends..
I've founf a new way to insulate those 3 pins of the tualatin..
I've used a vinilic glue, the same used for paper and wood..
There are different vinilic glue, some is more fluid, some is less fluid..
I have used one not too fluid, normally used for wood..
The glue cover the pin, and creates a plastic skin over it..
When the glue is dry, the pin is exactly like the other, because the glue become fully transparent..
The dimension of the pin is like the original, so no mod on socket are needed.. 

To place the glue on the pins I've used a small plastic pipe, the same of the BIC pens, but whithout the ink inside..
I've filled it whith glue, and cleaned the outside of the pipe..
.. then a sexual intercourse between pins and pipe..
.. and the games are done :-)"

wirehead "Don't give up on the BH6 just yet. I finally got my C1.2@1.49 as I type this reply. Similarly like you, I tried almost possible voltage, bios settings without any luck until I did this...

With your current CPU installed.
Set the Softmenu 600(100) --This is from my original PIII-600e
Default Voltage to 1.65.
Save and Exit bios.
Hold the power button for 4 seconds.
Remove the current CPU.
VID your Cel-t to 1.50V. --Don't forget to ground the VID25 cause the Bh6 won't do it for you.
Install and Boot...
Crank up the MB's voltage and start OCing...

Thanks to all !!! I was this >< close to toss these C1.2 in the garbage last night."

Xaero_toast "quote:

Oops...I missed that one. You are right. Anybody doing this on a slotket should be sure not to set voltage jumpers to "auto", in case their slotket uses AK36 for vid4.

I wonder, if there is an old PPGA spec that may have used pin AK36 as vid4. It could be the case that mobo manufacturers improperly used AK36 as vid4 just as a carryover from an old design.

What are the pins that set the VTT voltage (supposed to be 1.25 for tualatins)? I have a brand new 1.0A, but I have not yet attempted to mod it because I have an ST6. What I'm wondering is, does running it at 1.5V increase or decrease stability. If it helps stability, BXers out there might actually be better off than a board that actually supports the tually."

ol' man "Yeah sure, you didn;t have to mod your mobo in anyway. I have easily popped the cap off of a sock 370 ZIF and simply covered the "pin catch" with masking tape. I think I even gave a example of it in this thread. Way easier and in the end nothing gets ruined. To each his own I guess."

Irondog "Finally it works! Managed to run a 1.0A Tualatin on Asus P3V4X with stability!
I used an Asus Slotket Rev 1.02, removed the three pins and wired AK4-26, BUT this works ONLY if you have the ICS clock generator... Now it's stable at 1,44Ghz with 1,65v! thanks to all!"

jmsandrsn "Asus CUSL2: This one seemed a little bit easier. Was able to run 1.2@1.5 with about 1.7V. Had to insulate the 3 pins but did not have to connect AK4 to AK26 (it seems that most people with 815 mobos don't have to do this). I also wrapped VSS to VID0 to VID25 to VID2 and this set the default voltage to 1.55V. I believe you will need to wrap the vid pins or else you will end up above 2V. Some of the BIOS's out there will not allow Vcore changes though. I used 1011 beta 5 from http://www.x86-secret.com/ and this one did allow Vcore changes."

Spajky2 "TO ALL:
Another confirmation: the known Tuallie mod works at my friends AbitBE6-IIver1.0 (bios_bh7.0) with MSI-MS6905masterVer2.0 slotket. The mentioned Slotket has on it a chip (greatest surface mount=smd square /7,5x15mm/ one with a lots of legs-pins signed TVC16222 -the same one seen on the PL-370-T of PowerLeap S370 Tuallie converter!
So I think that the only working Slotkets (brand name or not) with this mod are those who have this chip on them, so this could be the breakThru info for finding Slotkets that are working properly before even trying practicaly with them /when searching for them to buy!/
Can you all guys having a working Tuallie on slotkets check please for this chip on your slotkets to confirm this info? Would be very usefull for us others in the future finding a propper slotket! This chip looks like converts AGTL i/o signals from CPU to Slot1 properly.. PLEASE CHECK & POST HERE name & version of a slotket & if this kind of chip has different markings, please give us this markings too if different! TIA !"

jmsandrsn "Maybe this is why my Iwill Slocket II ver 1.0 doesn't work at all with Tuallies. It has no IC chip at all like that one."

jmsandrsn "I was able to come across an ABIT Slocket III ver 1.1 today and it has the TVC16222A chip on there as well.
I tried this slocket(with Celeron 1.2) on an old(made in 1998) MSI 440ZX mobo but it wouldn't boot. This motherboard will barely support coppermines but I just think it's too old to work with Tualatins.
I'll try this Abit slocket on my P3B-F in the next few days."  I found out that the chip with the TVC16222A markings is called a voltage clamp.
I don't read Chinese but if take a look at this link it would appear that Ver 1.0 of the Abit Slocket III does not have any chip like that at all while Ver 1.1 does have the chip.
http://terasan.okiraku-pc.net/hole/slotket/ "

Ruiner "My abit slotketiii does NOT have that chip, but the mod DOES work on it.
I am fsb limited with it to 129 though."

sammy "It looks like slotkets with the TVC16222 chip probably have the best chance of success for using a Tually, but there seem to be a few exceptions.
My first try with the 3pin mod was with an Abit Slotket!!! version 1.0, because that is what chosenfool was successful with. Mine was stable at 12x136@1.65v, on a BE6-2. The slotket has the traces for the chip, but not the chip.
I was successful using a Soyo Sy-370 that doesn't have the TVC16222, with 3 pin mod and a Tually, however, my results with this
slotket are not consistent. I have one that runs a T1.0A at 10x140@1.7v on a BE6-2, but others that won't boot at all.
I have found other results that are not consistent, that I can't explain.
I have a 3pin modified Asus slotket that runs fine with one Tualatin, and when I simply replace the Tually with a different one, it won't even boot.
I thought I had killed one cpu this way, until I put it in my TUSI-M Tually board and it ran 11x133!
Anyone else seeing these kinds of inconsistent results?"

Ruiner "Re: make it past safe mode
I think the problem may have been a poor 4-26 jump, but I am not sure. I did remove the cpu from the slotket and remounted it.
I did a CMOS clear....that may have helped. A few people have noted successful posts after failures...without changing anything.
I used the earlier version of the abit slotket ///.
I had to do all of the mods to get it to post: block the 3 pins, the 4 to 26 jump, and the vid pin.
Even though the slotket has Vcore on board, it still needed a Vid pin mod on the cpu to post. In my case it was Vid4 to Vss, if I recall.
quote:

---

@crilicM@n wrote on 06-21-2002 08:39 PM:
Hy Ruiner... I´m collecting tons of info about the tualatin on slot 1 mobo combo... And seen that you had problems before with make it past safe mode ... How did you correct that situation?
Thanks:
Humberto Hepp"

ebc "asus cuv4x ver 1.03 can do it, i just try it out two day ago,insulate the three pins and connect ak4 to ak26,a german also do it with a asus cuv4x-e mobo,good luck for you.  sorry£¬i miss the important step,please connect Vid25mv(vid4) to Vss."

enrico24 "hi,
How did you insulate the trhee pins and connect ak4 to ak26 ?
I tried whith a tually 1.0 and cusl2-c without luck...till now.."

nucro "I just did it 4-5 hours ago  Now it's running @1332@1.50V stable  
First I tried with connecting ak4 to ak26 bur it didn't even boot. So I skiped it and now it works without a problem."

Comment: Those newest boads like cusl2 usualy don´t need the ak4 to ak26 connection... (like the i815...)

Celemine1Gig (about successfull generic slotcket) "That's what the package of the card looked like! It comes in a see-through plastic package and the scan below shows the back of the package! See that red "FC-PGA Coppermine" in the upper right corner?

Celemine1Gig has attached this image:
"

Comment: Sometimes,  for some, the bios cmos clearing has helped... For some not. I recomend doing it if you´re having problems with the mod not working just for itself....

nucro "Did you read what I wrote above? The ak4 to ak26 connection didn’t work for me. It may be because of the wire I was using. It was a bit thick so the processor wasn’t sitting flat on the sloket. But You have a similer setup to mine (Abit BE6-2 and Asus sloket) so you should try without doing the ak4 to ak26 connection. It worked for me."

ScorpiX "Don't be afraid...
You were right, the connection was problem, so I change the wire, and now it is ROCK stabile!"

Comment: Becarefull with the size of the wire... I think a ltle hammering can make them a litle bit more flat :-) But at the same time, ensure there is a good contact area between th wire end and the pin.... The wire is shurely the easiest way, but maybe it´s not the "golden" procedure...

ScorpiX "Ooo... man those Asus slotket's are worth it's weight in gold...   
O by the way I just wanted to say, that after 4 days of search, I found ONE Asus S370 DL-133 (rev. 1.02) whit TVC16222A chip for about 8 US$ "

Comment: Lucky 8 dollars!!! ... Teh AsusS370 slotcket seems to be one of the most successfull... Abit slocketIII usualy Ok too MSI 6093/6095 mainly the 2.0>rewiews seems consistently good too... IWill slocketII not always do the job... Generic Ones: Difficult, but the generic SuperSlotket-III seems to work!

jmsandrsn "I tried out an MSI 6905 Master ver 2.0 slocket today and I was albe to get the mod to work with this slocket as well (previously tried it on the MSI ver 2.3 slocket). This was on a P3B-F rev 1.04 board.
I also tried out a Gigabyte GA-6R7+ slocket but could not get the system to post (this slocket does not have any jumpers so I had to do wire trick to set the voltage). It probably won't make a difference but I will take out the AK4 to AK26 bridge to see if that makes a difference.I tried the Gigabyte slocket some more but still no luck. I still may try it some more in a few days since I may not have had all of the pins insulated well enough.
I tried an Asus S370-133 slocket and was not able to get it working either. I tried it with and without the AK4/AK26 connection but no luck. I had originally reported that it did work but I accidentally put the CPU back into the MSI slocket rather than the Asus slcoket."

Spajky2 " Imagine how I found stuff about that damn chip: a month ago I came accross a link on VP6.com to a review of new PowerLeaps fcpga-fcpga2 adapter & I don´t know why I saved the picture of showing it & the chip on it. After trying a bunch of slotkets with my friend he came to me saying that it actually works with his MSImasterV2 slotket & before he left I got a clue & checked the biggest chip on it & the markings seamed familiar to me so I checked the magnified picture in my archive: Bingo! DL-ed datasheet from the Net checked the function & here is the discovery!
Some people succeded with generic slotket IMHO because they had the last revision of Slot1 mobos with similar chip already on the mobo /depends on model,version & manufacturer of a mobo!"

DSTA "Agreed, also the other way around -- I think the Soltek and MSI parts don't use a TVC16222 and are modable.
BTW, I've taken a look at the data sheet and from what I can understand it's called a "Translation Voltage Clamp" (hence TVC16222).
I'm curious and would like to know how it is being used on the slotkets, so If anyone's got a scanner and some webspace, you could do me a big favour by scanning a slotket with that TVC and upload it somewhere. TIA!"

belinassu "have a look on my page using conductive paint and in the link
i gave in it
http://digilander.iol.it/grandecigno/Tua_BX_41.htm "

Comment: Another good site description for the mod with good photos, check it...

jmsandrsn "The Asus slocket I'm using is the S370-133 rather than S370-DL. It does have the TVC16222 chip but I can't get it to work with or without the AK4/AK26 connection."

Dave001 "sammy, you're my hero.
Got myself an Abit & 2 Asus slockets that are the same revisions as yours. I'm off to pick up a couple of Cel T's.
So far the list of working slockets i've found, and verified with people that have them working, are:
ASUS S370-DL Rev. 1.02 (has TVC chip onboard)
ASUS S370-133 Rev. 1.01 (has TVC chip onboard)
ASUS S370-133 Rev. 1.02 (has TVC chip onboard)
ABIT Slocket III ver. 1.00 (Missing chip)
ABIT Slocket III ver. 1.1 (has TVC chip onboard, but some of them seem to suffer stability issues above 133mhz)
MSI MS6905 master Slotket V2.0 (has TVC chip onboard)
MSI 6905 Master Slotket ver V2.3 (unsure)
It's taken me hours of searching to make that list, hope it saves someone else the trouble."

Spajky2 "I did a wire bridge for ak4-ak26 on S370 socket: carefully with non-isolated thin wire plugged into ak26 hole & isolated ak4 hole traced ski slalom like & than covered with transparent sellotape on top side to isolate it to prevent shortcutting pins; but you better use isolated one wire...no need soldering, painting or breaking pins; the CPU stays still intacted & not voided warranty when you pull it out of socket..oh yeah, for those 3 isolated pins I widened holes in the socket with 1mm drill & isolated holes with thin plastic wire cover.. After using a sellotape you have to use a needlle to make holes again for the cpu pins at that area; do NOT try to force CPU in without doing that before, you can brake pins!"

mczak "Yes! it works stable now! Thanks for all your help.
I didn't connect ak4 and ak26 however, after studying the datasheets (especially the timing diagrams for the startup-sequence) I thought that connecting ak4 to aj5 (a vcc pin) might just work the same - and it did. This is far easier to do with conductive silver paint.
If you want to try this yourself, one warning however: Make really sure you have isolated ak4, if not you will short Vss to Vcc!
Now running at 1.5V, 1330Mhz, 512MB ram with 3 modules (one of them a pc100 - had to lower ram timings for this one to get it to work...).
I was really sceptical about this mod, especially since I have a bios which is known to have problems with faster cpus (cpus with a multiplier higher than 10 will not work, one of the reasons I got a 1.0A) and an almost noname slotket.
(btw I think the reason why you need to connect ak4 is because if you don't the cpu will not read / or will read at the improper time (before the signals have stabilized) the configuration of some configuration pins during reset - this isn't mentioned in the datasheet, but that's my theory. It could be wrong of course, just a speculation.)"

Comment: Someone elsewhere said to be carefull about it becouse of one of them is npt 1,8v tolerant... So... becarefull with the volt...

Spajky2 "@ protocold: the working chips found till now are TVC16222 /A/ or IDT 74FST or two LVC07A ; they have to be on Slotket or on MoBo present; if not will not work as found till now; but worth trying! :-) maybe we will discover further chips working either..you could still change mobo; but "do it yourself" way is quite a fun just like OC-ing... :-)"

Comment: The only certain: TRY TRY TRY and be shure that other factors aren´t   disturbing, like a bad link of the wire etc, wrong vid pin voltage, cmos clear needed or just lack of "persinting"

Yeah, i think it works. My friend has the same slotket(it looks the same) and he run tualatin 1.0a on abit be6-2. but modification is a bit different.
wait a little, i'll ask him what did he do"

---

Originally posted by Spajky2
@ Neeki: /you´ve already got my mail for your inquiery/
I wanna ask about your friends SuperSlotket-IIIver1.0 he did additional rewiring: has he done any additional isolating pins or just straight additional rewiring on those additional pins?

---

he isolated 3 pins, connected ak4-an11, g37-g35. Thats all."

sammy "To answer some questions about versions that worked for me.
I took the cover off of the Abit Slotket !!!, that works on my BE6-2, and this is what is on the card: AB-FC370 and there is a sticker on the back that has FC370 V1.0
This card has traces for the large chip, but NO chip!
This was the first one I tried - just did the 3pin mod, and was lucky, it runs 12x136@1.65v
Asus S370-133 v1.01 and it has has the large chip on it.
MSI 6905 Master - has the large chip on it - cannot get to it to look for a revision number."

D2DM Slotket!!! version:
Take of the black (SECC2 type) casing.
It is held in place by 2 small pins in the lower left and right of the casing.
Then you can look at the back of the converter.
To the left at half hight there is the version #.
To the right there is a sticker with "FC370 Version#"
(For example on mine it sez to the left: v0.83
On the sticker: V1.1)

Comment: For a compatibility list consultin and adding your experience I found this here: http://www.leto.tk/ , the advantage of it over the plip3 list is that here is our real world with realy different slockets, cpu´s, mobos etc...

Spajky2 "@ joikd:
Do not worry; that slotket I found by yesterday experiments IMHO the Best !!!! (even if has no damn large chip on board) :-)

Today I pushed my Tuallie on SuperSlotket3 ver.1.0-noname Taiwaneese made with no "big" chip on board-generic one /with a bit different mod-posted here by Sammy/ from 1.0A to 1,55G with PC100 Cas2 64Mb ram with 1,85 Vcore /shows 1,8/ on 235W !!! Enlight PSU. I made some benchmarks on 1,5G (still NO data corruption!): It really kicks ass to P4wiliamettes ! :-) Would kick ass even more to new expecting Williamette core Celerons! :-) It was a really cheap uppgrade to my machine to double performance @1,35GHz where I will keep it with 1,52Vcore not to bother spending twice such money & time additionally just to gain another 10-15% of performance buying a new PSU&ram & other stuff to make machine be LOUDER; now it is quite silent! When you pass 1G & 133fsb IMHO really doesn´t matter much except benchmarks or for special purposes; the real bottleneck is in fact HDisc access time, which slows the machine...
I also tested my setup @1,35 with a demo loop RallyMasters game with open case, but I forgot to put the fan on heatsink!!! After realised that (3 hours later!), the game was still running smoothly... LOL (yeah, the heatsink was hot!) ! If it would be AMD Duron would be dead instead & fire in the house..! It was a REAL Burnin-in.. :-) ! I have a Philipphiness SL5VP, 1st week March OEM retail chip! :-) ...

@ ALL: /Thanks to Sammy for info for that mod on SuperSlotket-III ver.1.0/
With my yesterdays experiments /damaged also plastic Slotket holders a bit!/ I found interesting facts that may stabilise your CPU @ higher frequencies: instead of ak4-ak26 bridge /w or no resistor/ I used instead direct hair thin wire bridge on s370 between isolated AK4 & non isolated AN11 (Vtt) pin !!! Advantages: the wire is shorter & bridge easyer to do & so recieving less RF noise to the CPU thus improving stability also because Vtt pin is RF decoupled to GND by capacitors, while ak26 has much higher impedance & not decoupled! So also VttPwrGd is now RF decoupled too & Cpu more stable !!! Also now is no need for safety serial resistor in bridge against higher voltage because on Vtt is fixed on 1,5V needed for VttPwrGd, whatever the Vcore is !! You should TRY this manipulation leaving the ak26 pin alone!"

Spajky2 "@ D2DM:
>Superslotket3.. is that the same as abit slotket!!! ?<
NO, it is a SMP capable, cheap & easy to find universal generic noname no voltage regulation made in Taiwan!
>What different mod by Sammy? You mean the info from his question below? > Neeki - I don't think I've seen the ak4-an11, and g37-g35 connections before. Is this something new, and where did it come from?
EXACTLY that mod!!! It came from "studying datasheets, observation&experimenting" from Sammy on MadOnion thread about the same topic... :-)
> So ak4-an11 is better. Great But not always necessary (some setups didn't need ak4-ak26)..
I did not try without it on this particular generic slotket; sorry I forgot to check before assembling my machine..., but you can TRY it for me, maybe will work without that!
>How does g37-g35 help?<
It provides Vtt reference voltage for G37 pin for Tuallie, which is for P3 marked as Rsv instead & on some slotkets left unconnected or grounded-if this is grounded on majority generic slotkets, this would be one of the reasons while others do not work with Tuallies!
>(1.55Ghz with pc100 LOL)<
Yes with 3,6 V-IO & cas333-it does 130@cas222, it is generic one no name; some time2time makes errors that high but @150Fsb completely ok. I "found out this its feature" a year ago while changing Ram & decided to keep it for testing purposes better than sell it for a bargain price-it is always usefull to have apart some quality old piece of hardware!..; the bigger newer stick PC133 I bought does only stable max 139fsbCas3-crap one! :-("

Spajky2 "@ Falkentyne:
>the MSI slotket didn't have this problem, but then again, the MSI adapter is a lot simpler than the powerleap one, too!).<
The simplier the working slotket is, less electronics the signal passess, better stability & higher OC! Mine generic SuperSlotket-III is as simpliest as can be & more stable @ higher FSB than MSIver2 I was using before!"

Spajky2 "@ protocold:
isolate an3, aj3, ak4; connect isolated ak4 to non isolated ak26 (or isolated ak4 to non isolated An11);
connect (solder in the back of a slotket) G35-G37; thats all; oh jeah do also the Vid job for the default Vcore you want!
@ Bender:
try also connecting G35 to G37 after checking with multimeter that g37 should be not grounded on slotket /connected to Vss/; if so isolate it on CPU also & than connect it (wire it on CPU) with non isolated g35.
Worth a try..."

Spajky2 "ms-6309 v1.0 use tualatin celeron
first step:
use amibcp.exe change bios file£¬make the hidden item "vga frame buffer uswc" and "pci frame buffer uswc" active,the two items in the "chipset setup" of "configure setup data",then save and flash it to the mobo bios.
second step:
modify the mobo use the method in this thread.
third step:
when the mobo boot,enter cmos and change the two item above to "enable",that's all,i think it can work on other version of ms-6309 mobo,one addition information,i use the tualatin celeron 1ga£¬1.1ga or 1.2g i can't make promise,good luck everybody."

Spajky2 "@ s3ppel:
That about 1,8V on ak26 is true, but it has much higher resistance & so not providing much current; AN11 can provide much higher current /very low impedance/ but at max.1,5V & also decoupled with condensers to Vss=GND & so preventing "antenna" efect of wire bridge & thus reducing EMI interference to ak4 so giving better stability to chip IMHO!"

New Black Dak "Woo-freakin-hoo!!
K,since all our local shops suck ass I couldn't find a conductive pin, and I was very leary about trying to wire AK4 to AK26.
Instead I soldered a wire to the AK4 stub(pain in the ass!!) and wrapped the wire on AN11. Then I wrapped G35 to G37, and like I did on the voltage pins used a dab of fingernail glue to hold it in place. HELL YES. Running at 1600(12x133.33)@1.65v now with the IOQD on. Been running CPUBURN for about 3 hours, and still not a problem. I'll let it run for a couple more hours to be sure. After that, I'll bench it and get some scores for you guys. Then I'm gonna go for broke. Maybe I can hit 150 and 1800, but my PC-150 only did 147 CAS2 before, and if I get that high I'll leave it. I don't think 24-48 MHz is worth the performance hit of CAS3 when you get this high."

Spajky2 "Info to All:
other slotket working with mod (IMHO standard one-just got mail from croatian guy tried it!):
QDI ver.5.1 - /has 2 LCX07 chips on board/"

Spajky2 "@ mczak:
I was meaning this: on CPU & slotket there are running strong high frequencies currents & so making a lot of radio frequency noise which can be captured by that wire bridge, which can act as "antenna" & if strong enaugh can spoil the binary 1 on ak4 pin not having decoupling to the GND (Vss). So this wire should be shortest & thinest possible & decoupled with capacitor for RF to GND (Vtt is!) not to left it "semi floating" in the "air" because of higher inpedance on it thus preventing entering to Cpu core additional disturbances. This would prevent accidentally crashing some programs for example time to time or even reseting the pc or data corruption; especially when trying to OC to the limit, where very small things matter! This was also my peronal experience! My friend having even better exemplar than mine went higher than me on same slotket, but when I did AN11 bridge instead, I can go higher with lower Vcore than him without any data corruption! So this "my theory" should be seriously considered IMHO!
Also the IHS is NOT grounded like integrated "ColdPlate" on Mendocinos was, so also a heatsink is acting like capacitive linked antenna to the core & thus spreading EMI inside the case making problems @ edge OC where everything matters! It is just my believe, but I could be also wrong..."

mczak "@Spajky2:
ok I misunderstood you about the emi. But, in this regard this makes my solution even better - you can't have a shorter connection than between ak4 and aj5 (2mm or so that's about 1/10th of an inch) and vcccore (aj5) is decoupled of course too. Plus, it makes using conductive silver paint instead of a wire (I just don't like the wire method) dead easy."

Spajky2 "@ mczak:
If you do not rise Vcore much above default, it is even better to aj5, thats true shure, You are right! :-)"

Comment: Hmm it´s now looking like an11 and aj5 are better than an26... Why? Why would an26 be just the best option just becouse it was the first one described and than, the most used until now... Hmmmm I´m gonna try those :-) Choices are : ak4~ak26
ak4~an11
ak4~aj5

Ruiner "Even thinner than ata or FDD cable is CAT5 or regular modular phone cable. It worked great for me and is plentiful."

ol' man "No matter if the speed is 1200MHz or 1600MHz my chip will not load windows after 1.75v. ANything lower is stable but after this vcore BSOD city!"

Spajky2 "I´ve got an E-mail from Italian guy: ANOTHER SLOTKET WORKS: Soyo SY-370 with standard mod with bridge! /he made it on Abit VT6X4 board. Interesting thing is beside SuperSlotketIII another slotket which is not compatible with P3 Intel specs; another LOL :-)"

Spajky2 (About the superslotcketIII) "@ protocold:
isolate an3, aj3, ak4; connect isolated ak4 to non isolated ak26 (or isolated ak4 to non isolated An11);
connect (solder in the back of a slotket) G35-G37; thats all; oh jeah do also the Vid job for the default Vcore you want!
@ Bender:
try also connecting G35 to G37 after checking with multimeter that g37 should be not grounded on slotket /connected to Vss/; if so isolate it on CPU also & than connect it (wire it on CPU) with non isolated g35.
Worth a try..."

belinassu "i tryed carefylly the ak4-an11 bridge on my BX133 with 1.0a cely
but i CAN'T BOOT!!!
so i tryed without any bridge (as i did the first time)
to verufy if it was stable without any bridge now...
and i was very disappointed , infact it DIDN'T BOOT AT ALL!!!
i was worryed i waste the cpu but after i did
AK4-AK26 connection it BOOTED AGAIN...
so my CPU/MOBO NEEDS AK4-AK26 connection to work
no any solution could help my......
sometimes we have to be LUCKY with CPU/MOBO......."

smokva "Hi guys!
Another succes story.
Tualeron 1.0A working on ABT BX6r2 with Iwill SlotketII v1.1 converter.
Interesting thing is that I havent found any of the mentioned chips on coverter and it still works. There are some 2 small chips with A266 written on them. On PCB they are marked as TC1+L1. They have only 3 pins...now I don't know do these chips have something with the mod but the fact is that Iwill SlotketII is working.
BTW AK4-AK26 is connected."

BrainDrain "Got the 1.2 running on the generic SuperSlocketIII, with the G35-G37 mod. As for the bridge between AK4-an11, my system is completely unstable on 3d apps without the bridge to AK4. This is on both, the 1.0/Iwill SlocketII and 1.2/Generic slocket mods. The strange thing I noticed is that when bridging AK4-an11, on the generic slocket, it drops my vCore from 1.68 to 1.58. The results are the same with the Iwill Slocket. If you remember, I stated that putting a resistor, 270 ohms, in the bridge between AK4-AK26, on the Iwill slocket, dropped my Vcore from 1.68 to something like 1.48. It must be something different on the Soyo board so I'll just bridge AK4-AK26 without resistance.
I ran this 1.2 on a Powerleap Tualatin adapter for months before modding it. On the Powerleap, I could only run up to 118 FSB. On the generic slocket I can run it up to 126 FSB. 133 FSB is completely unstable for this chip, at any voltage and it's not a heat issue because my Alpha pep66T is keeping it nice and cool. This doesn't really surprise me though because this was probably one of the earliest produced chips. The newer ones will probably OC better.
I'm just happy that these things will run on the generic slockets. Some of the better slockets are becoming hard to find. I'll be doing a few of these mods for friends so I needed a cheap, easy to find slocket.
I'll be doing some stability testing and benching with this generic slocket but I'm sure I'll be switching back to the 1.0 on the Iwill just for the 133+ FSB.
Thanks again to all.

Comment: Actual datasheet pdf archive from intel: http://developer.intel.com/design/celeron/datashts/298596.htm

Sorry my bad english... :-) OC Forum: 760 posts and 56665 views until 22/06/02

Humberto Hepp

@crilicM@n