Cel 1.2 in BH6 with PL-iP3/T no go @ 124fsb

[Lyron]

Heya...
Just got my Cel 1.2 going in my Abit BH6 (ss bios)... It runs great @ 1344 (112fsb) but I am unable to get it working @ 124fsb... It will post, but with no video. But the odd thing is the monitor will powerup like normal, but it just wont show anything. I've gone all the way up to 2.05v and still nothing... I've got 512mb in 2 sticks of mwave noname pc133 that worked great @ 103fsb cas2 with my cel 633 @ 980. One stick at a time still does nothing... It also works great @ 83fsb so I'm 99.9% sure it's not a pci/agp clock prob. I was really wanting to get 1488+ out of this thing... _ANYONE_ with _ANY_ ideas PLEASE reply...
:(

Thanks
-Lyron

[William]

Its not voltage, the AGP bus is probably too much for the video card.

[ol' man]

Do you have a multimeter. If not I would suggest you try to find one and check the vtt voltages on that chip. Once you know that the vtt voltages are okay then we will help you get further but until then I am not going to help you raise your vcore any more than what they have it at 1.32v under spec. Not sure why they do this but I doubt your AGP card is the problem. My GF2 will do 105MHz AGP bus no problem and maybe higher. The main problem you may see is your PCI bus go but I am highly sure it is your vcore being at 1.32v. If you can check your vtt voltages with a multimeter and I will tell you which pins to check then I will help you get your vcore higher but until then you may wreck your chip if the vtt and vcore are on the same circut.

You can not adjust the vcore anyway in the bios. You have to do it on the slocket. As I said though I wll not help youuntil you find out forsure if your vtt is in spec. My theory is the reason they run the vcore so low on the slockets is because the vtt and vcore are ont the same circut for some odd reason. If they are both kept at 1.32v then you could keep the chip running albeit undervolted as also the vtt voltage in spec at under 1.36v under load.

[Lyron]

Well, the bios shows a vtt voltage of 1.48...
Also, both the bios and sandra the vcore change as I change it in the softmenu... The PL-iP3/T has no jumpers for voltage anyway.. Right now I'm running the chip @ 1.75vcore @ 1344mhz.

Also, the agp clock @ 83 and 124 are the same, so I don't think thats it...
I have a gf3 ti-200 so it should be able to take a high(er) agp clock...

Ok, so how am I going to change the vtt?
Or is that not what I want to do?
I'm not really clean on this :(

Thanks for your input :)

-Lyron

[ol' man]

Originally posted by Lyron
Well, the bios shows a vtt voltage of 1.48...
Also, both the bios and sandra the vcore change as I change it in the softmenu... The PL-iP3/T has no jumpers for voltage anyway.. Right now I'm running the chip @ 1.75vcore @ 1344mhz.

Also, the agp clock @ 83 and 124 are the same, so I don't think thats it...
I have a gf3 ti-200 so it should be able to take a high(er) agp clock...

Ok, so how am I going to change the vtt?
Or is that not what I want to do?
I'm not really clean on this :(

Thanks for your input :)

-Lyron

Your vcore can only be changed on your slocket by a trick but until you find out with a multimeter I will not tell you how as you may fry the chip. I will ask you again do you have a multimeter so you can check the vtt on the back of the slocket while it is running? If not can you get access to one? The OC'n community needs to know some things about that device and if it is running everything in spec like it is supposed to. Your vcore is not at 1.75v as it may say in the bios in reality it is stuck at 1.32v. If you could hit 1. 75v with that chip you would be at 1600+MHz. I would just leave you bios set at defualt for now. It would suck if the vtt is at 1.5v on the slocket too. This is what we need you to measure for us. Once you do this I will personally show you how to hit high HURTZ with that chip:)

Once again~again do you have access or can you get a multimeter? If you answer yes we will proceed further. If not I can't help you any more.

[Lyron]

Yes, I have one... I'll have to find out a way to get the damn cover off the PL-iP3/T thing...

What I don't get is how I can change the vcore in the bios via softmenu and have sensors lie to me about it :(

lyron@freak:~$ sensors -f
lm79-isa-0290
Adapter: ISA adapter
Algorithm: ISA algorithm
VCore 1: +1.79 V (min = +1.58 V, max = +1.92 V)
VCore 2: +1.48 V (min = +1.58 V, max = +1.92 V) ALARM
+3.3V: +3.28 V (min = +2.97 V, max = +3.63 V)
+5V: +5.07 V (min = +4.50 V, max = +5.48 V)
+12V: +11.78 V (min = +10.79 V, max = +13.11 V)
-12V: -12.00 V (min = -10.78 V, max = -13.18 V)
-5V: -5.06 V (min = -4.50 V, max = -5.48 V)
fan1: 0 RPM (min = 3000 RPM, div = 2) ALARM
fan2: 0 RPM (min = 3000 RPM, div = 2) ALARM
fan3: 4411 RPM (min = 3000 RPM, div = 2)
temp: +84.2


Edit: Wait, I think I get it... the PL-iP3/T is whats sending the 1.45 (Or 1.32, whatever) to the cpu... The 1.75 I'm setting in the bios is not going anywhere... Ok, I'm smart... :x

[funnyperson1]

because the mobo may send out 1.75 volts, but the powerleap scales down the voltage to the chup to 1.32...

[Lyron]

Originally posted by funnyperson1
because the mobo may send out 1.75 volts, but the powerleap scales down the voltage to the chup to 1.32...

Heh, just as I edit my post
hehe
Thanks :D

[funnyperson1]

Originally posted by Lyron


Heh, just as I edit my post
hehe
Thanks :D
lol

[ol' man]

Here is the pin layout. I got it as small as I could. I think you will want to test it by touching the ground to the vss pin which is black and the hot side to the vtt pin in red. I could be wrong though so let me ask a couple people first alright. I know for sure the vtt goes to the red but I am not sure about the ground wire. I know for a fact that the vss pin is the system ground and it makes sence that that is where you would put the wire. Maybe you would know your self? Once we find out what is up with your vtt and vcc voltage we will figure out what is going on here okay. I think you can check your vcc voltage the same wya as your vtt. It is supposed to be at 1.475 but I read a report by a taiwan website that said the adapter is keeping the vcore to 1.32v below spec which would explain why no one can get the adapter to go above around 110 fsb.

Here is the pin layout..

http://hhrol.bobsville.com/pins.jpg

[Lyron]

Sounds good...
Once you are 100% sure about as much as you can, I'll get started... Also, do you have a bigger pic that than? It would be nice to print up a nice big image to work with :)
Thanks
Feel free to icq me @ 2811683

[ol' man]

Originally posted by Lyron
Sounds good...
Once you are 100% sure about as much as you can, I'll get started... Also, do you have a bigger pic that than? It would be nice to print up a nice big image to work with :)
Thanks
Feel free to icq me @ 2811683

You are going to need these anyway to get all your vcores.

ftp://download.intel.com/design/celeron/datashts/29859601.pdf

It is a .pdf file so hopefully you have adobe acrobat reader.


First goto page 60 and you can expand the pic as much as you want to print them.

[doer]

I tried that wiretrick, connecting vid1-> vid2 in my Powerleap adapter socket. Well at least my cpu didn't get fried. But I couldn't get Cel1.2 booting @ 1.6Ghz either. I got GlobalWin WBK-38 cooler. My mobo is Abit VT6X4. I haven't tried other FSBs than 100 and 133. Though I don't have a faintest idea if that voltage was raised or not by that wire. But it didn't boot @ 1.6Ghz.On 1.2Ghz it worked just fine with the wiretrick.

Now I'm running @ 1.2Ghz w/ normal celeron heatsink+fan without the wiretrick. Case temps 23*C and 32*C. I guess those forthcoming 1.0 &1.1 Ghz tualatin Celes would be a better choice... 1.33Ghz and 1.466Ghz oc:ed @133FSB.

But if any1 could confirm does that wiretrick actually raise the voltage on Powerleap adaptor. And what's the default V on the adapter, that would be great..:)

[Lyron]

Well, had some good reading with my coffee this fine morning and it looks like the 1.325 setting has all _5_ voltage pins connected...
Looks like I may have to drill out a pin or 2 :x

I have yet to read the vcc or vtt so I'm still a step or 2 back from doing that... Will post an update when I read the vcc/vtt.

"I could be wrong though so let me ask a couple people first alright." Hear back yet?
Thanks :)

-Lyron

[ol' man]

Originally posted by Lyron
Well, had some good reading with my coffee this fine morning and it looks like the 1.325 setting has all _5_ voltage pins connected...
Looks like I may have to drill out a pin or 2 :x

I have yet to read the vcc or vtt so I'm still a step or 2 back from doing that... Will post an update when I read the vcc/vtt.

"I could be wrong though so let me ask a couple people first alright." Hear back yet?
Thanks :)

-Lyron

Good call man. I am hoping that this is a security measure that they think may thwart OC'n with this. You may be able to remove some pins to get to at least 1.6v which should be enough to hit 1.6MHz. Does your BX board have a 1/4 divder? There is no wire trick after seeing how they have the 1.325v setup that will work with this. You will definatly have to do some hard modding. Ie, pin removal or other. You have to remember that the chip is set to 1.475v and you will have to pull pins accordingly. I would not do anything over 1.65v without decent cooling. I am thinking they are keeping the vtt in check as the way they have the vcore vid pins setup it would be impossible to get higher vcores with a wire trick. Hopefully this is true. Sounds like the only true way to check vcore is with a multimeter with this device.

[ol' man]

You know thinking about it again I would do something where you can maybe adjust the vcores between 1.5v and 1.65v. I know they have some jumpers at radio shack that supposedly could be possibly attached ot this device. One may even be able to do this off of the device too.

[Lyron]

"Does your BX board have a 1/4 divder?"
Nope... I have the Bh6 rev 1.0x... But I have run this board @ 124 before with a very close hardware setup... I'm not too sure about 133fsb, but I would _REALLY_ like to be able to try it. And the voltage looks like whats holding it back (Duh).

I bounced this off my dad (electrical/optical engineer) and he said we may get lucky and find a lead comming off the pin/package on the back and we may be able to cut and add a jumper/switch.

Or,( just sitting here typing and thinking about it) putting some kinda cover over a pin.
Ok, anyway...

Just removing vid3 would give me 1.725.
And that would for _SURE_ give me 1488mhz. And if the pci/agp bus holds up, a VERY good shot @ 133fsb for 1596mhz.
I've got good but not super good cooling on my chip. Globalwin fop32-1 (http://www.globalwinusa.com/products/fop32_1.html)
So, I should (keyword) be able to do 1.725vcc with this... I may start to run a little hot... I've also been looking at getting a new heatsink if it runs too hot...

Wish me luck on reading vcc/vtt...

Edit: Btw, I was able to do 980mhz @ 2.05v with my Celeron2 633 and the heatsink was just warm to the touch... The bh6 has no cpu temp sensor monitor so I can't give ya a true temp... So lets hope I'll be able to run @ 1.6gig/1.725v

-Lyron

[Godfodda]

Originally posted by Lyron

Or,( just sitting here typing and thinking about it) putting some kinda cover over a pin.


Someone suggested in here a while back (outhouse, maybe?) that instead of removing pins, one could cover them with nail polish. I've never had to attempt this, so do so at your own risk. :)

BTW, the suggestion was to use a colored or black polish so it would be easier to see where the goop ended up and could be removed from any other pins that may have been partially coated.

[Lyron]

Originally posted by Godfodda


Someone suggested in here a while back (outhouse, maybe?) that instead of removing pins, one could cover them with nail polish. I've never had to attempt this, so do so at your own risk. :)

BTW, the suggestion was to use a colored or black polish so it would be easier to see where the goop ended up and could be removed from any other pins that may have been partially coated.

Wow, thats a really good idea :)
Thanks :D

Edit: Well, after readin the intel spec paper, it looks like it just can't be connected to nothing... It shows (page 21) 0 being connected to vss (ground), and 1 being "Open to processor". So, would it work if it was conncted to nothing and not vss like intel spec asks for? :(

-Lyron

[ol' man]

Originally posted by Lyron


Wow, thats a really good idea :)
Thanks :D

Edit: Well, after readin the intel spec paper, it looks like it just can't be connected to nothing... It shows (page 21) 0 being connected to vss (ground), and 1 being "Open to processor". So, would it work if it was conncted to nothing and not vss like intel spec asks for? :(

-Lyron

Actually it was I that suggested using black nail polish. Thing is you have to let the polish dry for a awhile and then it works very good.



Well i found out I guess which pins you should test.

These are the ones.



I also assume you are using pins AB36 and AG1 (for vtt) and X4 (for vss).

http://discussions.hardwarecentral.com/Forum2/HTML/012013.html

This is what it seems glussier is saying to test with. Not sure why but find these exact pins I guess and test them. I am thinking that your adapter is okay but I think you should still test it.

As for running 124fsb. That will give you a very nice speed. i am running 125fsb for 1500MHz and am going to give this to my brother for christmas. I have found it to be a tank at this speed. Not bad for a $85 mobo and a $110 chip. Of course the PCI is fine at 31.5MHz.

[Lyron]

Wow, do I have a lot to post about...
First off, it looks like the ip3/t gives the cel 1.2 1.425 volts.
Also, the 5 voltage leads coming off the package are RIGHT on the back... So, my dad and I get an idea... Cut, all 5 vid traces coming off the cpu.
Take a set of 5 dipswitches, one for each vid pin, make off go to a vss pin, and on do nothing (open)... Oh my... It works :D
I've still got the multimeter hooked up to 2 wires going to a vcc and vss pin. In the 5min I've been playing with it I've posted @ 1.65v (meter told me 1.632 or something). I have yet to try clocking any higher, but I plan on doing that in just a sec...

But, before you crazy people try this, let me document this stuff better first and TAKE PICS!

WoooOOOoOOOOO! :D
I'll post an update soon...
Let me get a'clockin :)
*jig*

[ol' man]

Originally posted by Lyron
Wow, do I have a lot to post about...
First off, it looks like the ip3/t gives the cel 1.2 1.425 volts.
Also, the 5 voltage leads coming off the package are RIGHT on the back... So, my dad and I get an idea... Cut, all 5 vid traces coming off the cpu.
Take a set of 5 dipswitches, one for each vid pin, make off go to a vss pin, and on do nothing (open)... Oh my... It works :D
I've still got the multimeter hooked up to 2 wires going to a vcc and vss pin. In the 5min I've been playing with it I've posted @ 1.65v (meter told me 1.632 or something). I have yet to try clocking any higher, but I plan on doing that in just a sec...

But, before you crazy people try this, let me document this stuff better first and TAKE PICS!

WoooOOOoOOOOO! :D
I'll post an update soon...
Let me get a'clockin :)
*jig*

This is what I was suggesting above man. Sweet. But what was the vtt voltage?

Also quit posting how you are hitting 1.65v and start posting how high of hurtz you can get on that BX board:D

Also dude only use like 1.5v to start. You may not need more than this to hit what you desire. If you got equipment that can do 133/3 MHz PCI then you could do the 13 dance but I would be careful and only do it with the HD at first.

So you are saying that the slocket was giving the chip 1.425v instead of the 1.325 like some others have had it give. Even 1.425 is below spec but not as bad as some of the others I have heard. Maybe this is why you could hit 112fsb which usually is opnly 105 to most with that adapter.

[ol' man]

Originally posted by Lyron
"Does your BX board have a 1/4 divder?"
Nope... I have the Bh6 rev 1.0x... But I have run this board @ 124 before with a very close hardware setup... I'm not too sure about 133fsb, but I would _REALLY_ like to be able to try it. And the voltage looks like whats holding it back (Duh).

I bounced this off my dad (electrical/optical engineer) and he said we may get lucky and find a lead comming off the pin/package on the back and we may be able to cut and add a jumper/switch.

Or,( just sitting here typing and thinking about it) putting some kinda cover over a pin.
Ok, anyway...

Just removing vid3 would give me 1.725.
And that would for _SURE_ give me 1488mhz. And if the pci/agp bus holds up, a VERY good shot @ 133fsb for 1596mhz.
I've got good but not super good cooling on my chip. Globalwin fop32-1 (http://www.globalwinusa.com/products/fop32_1.html)
So, I should (keyword) be able to do 1.725vcc with this... I may start to run a little hot... I've also been looking at getting a new heatsink if it runs too hot...

Wish me luck on reading vcc/vtt...

Edit: Btw, I was able to do 980mhz @ 2.05v with my Celeron2 633 and the heatsink was just warm to the touch... The bh6 has no cpu temp sensor monitor so I can't give ya a true temp... So lets hope I'll be able to run @ 1.6gig/1.725v

-Lyron

Whoa dude I just read over this???????

As I said you will not need that much juice(1.725vcc) with this chip most likely. Only try the lower vcores first as I have found and many others that the extra vcore sometimes does not allow higher OC's. I know if I keep my vcore at 1.8v I can not keep it stable at 1.5GHz but if I drop it too 1.7v I can keep it stable to 1.54GHz or better. You take it easy with that Phillipine made chip there bud;) You got a great chip and you wouldn't want to muck it up man:)

[Lyron]

I'm running @ 124fsb for 1488mhz @ 1.65v :)
Been running sandra/3dmark2001/tribes2 for about 2 hours now with zero lockups/crashes/<insert bad thing here>.
After a 45min game of tribes2, I shutdown to feel how hot the heatsink was getting...
Just warm to the touch :)

Will try for 133fsb soon, but I don't want to push my luck too hard, too quick :D

Also, I never did read the vtt... Sorry :/
Oh well?

Anyway...

I have NOT used a voltage lower than 1.65 for 1488mhz... I will do this at some point, but for now, without the heatsink heatting up much, and it LOOKS (keyword) to be stable, I don't feel like touching it :)

[ol' man]

Originally posted by Lyron
I'm running @ 124fsb for 1488mhz @ 1.65v :)
Been running sandra/3dmark2001/tribes2 for about 2 hours now with zero lockups/crashes/<insert bad thing here>.
After a 45min game of tribes2, I shutdown to feel how hot the heatsink was getting...
Just warm to the touch :)

Will try for 133fsb soon, but I don't want to push my luck too hard, too quick :D

Also, I never did read the vtt... Sorry :/
Oh well?

Anyway...

I have NOT used a voltage lower than 1.65 for 1488mhz... I will do this at some point, but for now, without the heatsink heatting up much, and it LOOKS (keyword) to be stable, I don't feel like touching it :)

Ummmm, dude it should do that at default vcore with a phillipine made chip?????? You didn't read the vtt but you upped the juice to 1.65v off the bat???????? You had the multimeter that close to the vtt pins without checkiung it for US. Dude I mean there are alot( STRESSES ALOT) of people that want to know what the story is with that adapter I don't mean to be a sour puss but why in the world did you do that. You probably only need 1.55v at the most to do that MHz. I am still going to tell people to stay clear of that adapter till I know what is up with it. It is probably okay but I am not sure and would really need to know before I ordered it. It is pretty dissapointing that you had it that close to the vtt pins and you won't even tell us. Pretty crappy dude. Sorry. O well?

[Ruiner]

Since he can only use the wire trick with that card, he is probably limited to default and 1.675 (unless he was successful with the dipswitch idea he was discussing).

I'm guessing he's maxed out because of his AGP/PCI speeds, but I am curious about the Vtt.

He may have cooling issues as well.

[Lyron]

Originally posted by Ruiner
Since he can only use the wire trick with that card, he is probably limited to default and 1.675 (unless he was successful with the dipswitch idea he was discussing).

I'm guessing he's maxed out because of his AGP/PCI speeds, but I am curious about the Vtt.

He may have cooling issues as well.

What is this, "unless he was successful with the dipswitch idea he was discussing". It DOES work :) I pick a voltage out of page 21 of the Celeron 1.2 datasheet, and set the dipswitches... I'm getting a nice digital cam tonight, so I'll take lots of pics and document this much better...

On the Intel Cel 1.2 datasheet, it has pin combo's up to 1.825v, I have yet to try it, but 1.75v worked (meter showed 1.72v).

Well, right now, I don't feel like taking out the board and putting an extra wire on a vtt pin... Sorry? :D

Ok, I'm off to try 133fsb...


-Lyron

[Ruiner]

Sorry, I wasn't sure if you had finished the dipswitch setup ( I thought you were using wire wrap only).
I'm looking forward to the pictures.
What temps are you getting at load?

I'm in a quandary here. The Powerleap costs about 60 bux or so after you account for the celly.
A brand new ECS Sis635 mobo (with both SDRAM and DDR support, and reportedly faster than the 815 even with SDRAM) is about $50. http://www.hardforum.com/showthread.php?threadid=281018&perpage=5&pagenumber=2

Neither offer Vcore adjustments, so it's wire trick or dipswitch/hack. The ECS has only 66/100/133 FSB's and the PL has whatever your mobo has. (so it's 1600 or bust on the ECS).

One factor to swing it to the PL....I wouldn't have to reinstall Windoze and all my apps. A question regarding that: can you reinstall an OS over itself after a mobo replacement or does it have to be clean? I have heard of windoze recognizing everything automatically, but that sounds dicey.

[Lyron]

Well, I was able to post @ 133fsb/1.75v, but Win2k would bsod on boot...
I'm not to sure If I want to try 1.8+v. I think I want to get a better feel for how hot the heatsink is getting first... I'm going to go down to radio shack and get a temp monitor thing...
Once I can watch to see how hot it's getting, I'll try higher voltages... Until then :D
1488mhz is still nice on my 3 1/2 year (?) old mobo :D

Btw, I'm running @ 1.6v now and it LOOKS to be stable (1hour of tribes2)... I'm sure 1.65v was stable, I ran prime95 overnight...

Also, load temps. The bh6 lacks the ability to monitor the on die temp sensor, so thats why I'm getting a little lcd indoor/outdoor thermometer (Like this http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F007%5F013%5F003%5F000&product%5Fid=63%2D1009) today...
I do have a hole in the heatsink for a sensor but I'm not sure if the sensor will fit... I may have to sand it down or something... Anyone have any idea whats inside a little sensor like that?

Thanks :D

[Ruiner]

yeah, there's a tiny thermistor inside that plastic 'bullet'. You can buy a bare thermistor for 3 bucks at Radshack and connect it to the wire. They are tiny...like 2-3mm.

[Lyron]

Originally posted by ol' man


Ummmm, dude it should do that at default vcore with a phillipine made chip?????? You didn't read the vtt but you upped the juice to 1.65v off the bat???????? You had the multimeter that close to the vtt pins without checkiung it for US. Dude I mean there are alot( STRESSES ALOT) of people that want to know what the story is with that adapter I don't mean to be a sour puss but why in the world did you do that. You probably only need 1.55v at the most to do that MHz. I am still going to tell people to stay clear of that adapter till I know what is up with it. It is probably okay but I am not sure and would really need to know before I ordered it. It is pretty dissapointing that you had it that close to the vtt pins and you won't even tell us. Pr
etty crappy dude. Sorry. O well?


Well, hrm, I don't like to be rude, but, did you spend the $170 for it? I know other people would like to know this info, but each time I solder an extra wire on that thing I'm taking a risk... And yes, I know you did help me out a lot, and I will be writing up a document with all the steps on how to do this. Also, I didnt know the vtt was this important for overclocking... I had it in my head it was all about the vcore... Next time I pull it out (for taking pics) I will take the time to do it... Anyway...

When I get a new cpu, I like to start out at what I would think would be a stable speed/voltage and work my way DOWN on the voltage side. I dislike having to reset my cmos a lot. I'm happy with it running a little hot at first until I find the lowest stable voltage...
1.65v is _NOT_ REALLY high... Hell, it's still in intel's spec. 1.75v is the "Absolute Maximum Rating" (Page 26) for the vcore. If I was NOT able to post @ 1.425v @ 1488, _I_ would think that it would need at LEAST 1.5v to get it booting an os, and at least 1.55v to get it 100% stable...
But, I will find out as I work my way down.

Sorry if I ****ed anyone off...

(Edit: I take it p.i.s.s is a bad word? Very sad...)

-Lyron

[ol' man]

Originally posted by Lyron
Well, I was able to post @ 133fsb/1.75v, but Win2k would bsod on boot...
I'm not to sure If I want to try 1.8+v. I think I want to get a better feel for how hot the heatsink is getting first... I'm going to go down to radio shack and get a temp monitor thing...
Once I can watch to see how hot it's getting, I'll try higher voltages... Until then :D
1488mhz is still nice on my 3 1/2 year (?) old mobo :D

Btw, I'm running @ 1.6v now and it LOOKS to be stable (1hour of tribes2)... I'm sure 1.65v was stable, I ran prime95 overnight...

Also, load temps. The bh6 lacks the ability to monitor the on die temp sensor, so thats why I'm getting a little lcd indoor/outdoor thermometer (Like this http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&category%5Fname=CTLG%5F007%5F013%5F003%5F000&product%5Fid=63%2D1009) today...
I do have a hole in the heatsink for a sensor but I'm not sure if the sensor will fit... I may have to sand it down or something... Anyone have any idea whats inside a little sensor like that?

Thanks :D

Well no doubt dude I was trying to tell you that your vcore was too high and you wouldn't need all that juice to get it too go but you wouldn't listen. Well now after running at those high vcores you may have wrecked your chip a little bit. Many with that chip have said that the higher vcores actually caused instability with the chip. That vcore was around 1.625v for the phillipine made chip.. Well I am done here. Nothing left for me in this thread. I try helping someone and watch them possibly wreck a sweet chip that I showed them what not to do in the first place. You went and did it all anyway. See ya.

[DeepScience]

Originally posted by Lyron

1488mhz is still nice on my 3 1/2 year (?) old mobo :D

Btw, I'm running @ 1.6v now and it LOOKS to be stable (1hour of tribes2)... I'm sure 1.65v was stable, I ran prime95 overnight...

Thanks :D

You da man. I have a question though. How on earth do you get Tribes 2 to run stable for a whole hour??

I'll be watching out for your article.

[Lyron]

Originally posted by ol' man


Well no doubt dude I was trying to tell you that your vcore was too high and you wouldn't need all that juice to get it too go but you wouldn't listen. Well now after running at those high vcores you may have wrecked your chip a little bit. Many with that chip have said that the higher vcores actually caused instability with the chip. That vcore was around 1.625v for the phillipine made chip.. Well I am done here. Nothing left for me in this thread. I try helping someone and watch them possibly wreck a sweet chip that I showed them what not to do in the first place. You went and did it all anyway. See ya.


Ok dude, I don't know what you have up your ass, but calm down, dude... "higher vcores actually caused instability" Yes, heat does that...
I'm not looking for someone to hold my hand as I play with my new chip. I've been overclocking for some time now... Started out with a Cyrix 150+ (120mhz) and clocked it to 133mhz... I'm just a newbie to HARDWARE mods... I can do what I want with my new chip... Don't be a jerk just cuz I don't do things JUST as you would... I don't get why you would say things like that. You think my vcore is too high? Thats fine...
Well, thanks for your help.
But I don't see why I should take the extra time/risk to read the vtt for you (And all the "others" I hear so little about) if your going to be a jackass about it and the way I run my chip...
Thanks again...

-Lyron

[Lyron]

Originally posted by DeepScience


You da man. I have a question though. How on earth do you get Tribes 2 to run stable for a whole hour??

I'll be watching out for your article.

I'm not sure if this is a joke or not...
Tribes2 does segfault from time to time... But never more than once in 2 _DAYS_...
Are you sure your setup is stable?

[ol' man]

Originally posted by Lyron



Ok dude, I don't know what you have up your ass, but calm down, dude... "higher vcores actually caused instability" Yes, heat does that...
I'm not looking for someone to hold my hand as I play with my new chip. I've been overclocking for some time now... Started out with a Cyrix 150+ (120mhz) and clocked it to 133mhz... I'm just a newbie to HARDWARE mods... I can do what I want with my new chip... Don't be a jerk just cuz I don't do things JUST as you would... I don't get why you would say things like that. You think my vcore is too high? Thats fine...
Well, thanks for your help.
But I don't see why I should take the extra time/risk to read the vtt for you (And all the "others" I hear so little about) if your going to be a jackass about it and the way I run my chip...
Thanks again...

-Lyron

Yeah and I wasted a good 30min finding out the info you wanted. You help me I help you? You don't help me, ahhhhhhhhhh find another chump.

[William]

Be nice guys.....

[Yodums]

Listen to William heh..

Lyron I believe ol'man was juss trying to help bro, were all in here to give each other tips and suggestion on how to go all out to the max as everyone would.

[Lyron]

Originally posted by tsunami
Listen to William heh..

Lyron I believe ol'man was juss trying to help bro, were all in here to give each other tips and suggestion on how to go all out to the max as everyone would.

Well, the thing is...
I was going to take the time/risk and read the vtt...
"Next time I pull it out (for taking pics) I will take the time to do it... " and the next thing I know I was flamed... I'm not going to help someone if they are going to be a jackass to me... :(
Ok, I'm done talking about this...

[funnyperson1]

you may want to take the vtt anywayz, if they are running it on the same channel as the VCORE as some people including olman suggested you could fry yuor chip....

[ol' man]

Originally posted by funnyperson1
you may want to take the vtt anywayz, if they are running it on the same channel as the VCORE as some people including olman suggested you could fry yuor chip....

That is the whole point summed up in a very short sentence by funnyperson. I was trying to help but in the end you would have helped a very large amount people gain trust in that device. At the moment I have NONE.

Anyway I did not know you had to solder anything on it. I figured the pins stuck out far enough like they do on my ST6 that you could have just carefully clipped your leads to them and in a few short seconds known what the vtt was at.

[Lyron]

Originally posted by ol' man


That is the whole point summed up in a very short sentence by funnyperson. I was trying to help but in the end you would have helped a very large amount people gain trust in that device. At the moment I have NONE.

Anyway I did not know you had to solder anything on it. I figured the pins stuck out far enough like they do on my ST6 that you could have just carefully clipped your leads to them and in a few short seconds known what the vtt was at.

Oh no way would I be able to hold the meter probes in the case on the right pins for more than 1/1000 of a sec, then slip and make a connection from the vcc to some other pin and fry the chip... As you'll see in the pics (Cam just got here), My dad and I solderd on 2 wires to a vcc and a vss... That made it VERY easy, once you solder on a wire to the pin...
Anyone have any idea if what the bios is being told about the vtt could be true? From the first time I booted up, the vtt reported in the bios has been 1.45v. It would save a lot of time to be able to know if this is a true reading or not...
And no, as I change the vcore in the BIOS (that does nothing) it does NOT change with it...

Sorry about getting rude with you ol' man :/
It's just that, I take this (what I would call) big risk with my NEW cpu you just got and have it work. Then have someone ask you to do more to your cpu without much if any gain for you. :/ I don't have $170 to go out and buy this combo again...

One last note, the Intel spec for max voltage going to vtt is the same as vcc...
:D

Cam just got here... Time to take pics :)

[DaveB]

Don't take it personal Lyron. The ol' man was harrassing me over at HWC to get the same info for him. He acted real annoyed when I suggested he go and buy one if he's so interested. I wasn't about to mess with it since I was gonna either sell it or RMA it. He was saying over there he got someone to do his testing for him at some other forum. I guess he meant you! :D

[ol' man]

Originally posted by Lyron


Oh no way would I be able to hold the meter probes in the case on the right pins for more than 1/1000 of a sec, then slip and make a connection from the vcc to some other pin and fry the chip... As you'll see in the pics (Cam just got here), My dad and I solderd on 2 wires to a vcc and a vss... That made it VERY easy, once you solder on a wire to the pin...
Anyone have any idea if what the bios is being told about the vtt could be true? From the first time I booted up, the vtt reported in the bios has been 1.45v. It would save a lot of time to be able to know if this is a true reading or not...
And no, as I change the vcore in the BIOS (that does nothing) it does NOT change with it...

Sorry about getting rude with you ol' man :/
It's just that, I take this (what I would call) big risk with my NEW cpu you just got and have it work. Then have someone ask you to do more to your cpu without much if any gain for you. :/ I don't have $170 to go out and buy this combo again...

One last note, the Intel spec for max voltage going to vtt is the same as vcc...
:D

Cam just got here... Time to take pics :)

Forget it dude. Sorry to waste your "time." I always test my stuff out of the case before I put it into the case. I guess you don't. You don't need your computer in a case for it to work. As for helping you wreck your chip?????? I think you are doing a decent job of that your self without our help if indeed the vtt is out of spec. I read where they say that the vtt can be upto 1.75v but then in another part it said it should not be higher than +/-9% of 1.25v which equates to 1.36v for the vtt.

Here is the exact pic.

[ol' man]

Originally posted by DaveB
Don't take it personal Lyron. The ol' man was harrassing me over at HWC to get the same info for him. He acted real annoyed when I suggested he go and buy one if he's so interested. I wasn't about to mess with it since I was gonna either sell it or RMA it. He was saying over there he got someone to do his testing for him at some other forum. I guess he meant you! :D

Yeah don't take it personal:rolleyes:

[Lyron]

Originally posted by DaveB
Don't take it personal Lyron. The ol' man was harrassing me over at HWC to get the same info for him. He acted real annoyed when I suggested he go and buy one if he's so interested. I wasn't about to mess with it since I was gonna either sell it or RMA it. He was saying over there he got someone to do his testing for him at some other forum. I guess he meant you! :D

I feel, so... Loved...

Oh wait no... :D

Ok, got some nice pics.
http://www.speakeasy.org/~norton/all1.jpg
http://www.speakeasy.org/~norton/cutleads.jpg
http://www.speakeasy.org/~norton/pins3.jpg
http://www.speakeasy.org/~norton/vttreadout.jpg
http://www.speakeasy.org/~norton/voltagereg.jpg
http://www.speakeasy.org/~norton/dipswitches.jpg
http://www.speakeasy.org/~norton/pins2.jpg
http://www.speakeasy.org/~norton/pins1.jpg
http://www.speakeasy.org/~norton/installed.jpg

No document to go with it yet...
The vtt readout is just for you ol man :P
MuHaHaHAHaHAhhhaHAhAhAhah!

[Pinky]

This whole thread makes my stomache sick, not because of any one person's comments but for the tone that some have chosen.

Some of us like to get things running and stop. I have an awesome system with existing and/or free components. There's an end to my madness (usually the bottom of my wallet). Tweaking stops there for me, if I need more juice I'll resell what I got and buy an upgrade. I don't mod for others, I mod for myself. I overclock my machine because it's fun and challenging and cool to brag to my friends ;) (not to mention the nuances I've learned in the past 8 months with PC hardware I'd never bother to learn otherwise). I like to do this. I don't like getting hot and heavy over some lousy voltage settings for a half-a$$ed adapter that's probably frying the caps as we speak on this system.

I think it's mostly been said, but as an active member here I would be doing myself an injustice not to say anything.

[Pinky]

Originally posted by Lyron


I feel, so... Loved...

Oh wait no... :D

Ok, got some nice pics.


I need to say I like the mod.

[ol' man]

Pinky's and my old celery are exceptions to the reasoning that anything over a certain amount will cause instability. I used to run my old celery at 1.86v and it would stay stable. I guess pinky can run his at 1.8v and it stays stable. I can't run my new one over 1.725v without instability for some weird reason. It seems those phillipine chips are even worse.

Well any way the vtt you show me in the bios don't mean didily:D

I guess all I can do is laugh this off and I guess without knowing for sure if everything is okay we will know in a few months anyway:D I would have liked to have known sooner as I have a BE6-II with the 1/4 divider and all would be great with this on that you know as many other BE6-II owners. Still though $170 is a bit much. Well if it starts smokin as people say it is hard put the smoke back in. Plus a hard lesson learned of which I have had many.

All that aside it was a very nice mod although it looks as if there is a great deal of congestion around the fan. Hope everything goes alright, that is all I really wanted for the situation in the first place plus you would help out the large OC'r/PC community. What you did is great but it is known that that could be done what isn't known is if it is alright. I will probably chuckle my self to sleep over this one you are as crazy as me. Jump in head first without lookin':D

[Lyron]

Well'sir...
I've got 3 things holding me back from shutting down right now, pulling it out, and soldering up a wire to a vtt pin...

1. If the vtt is being changed to the same voltage as the vcc, @ 1.60v I'm still under the Intel spec (Absolute Max). Ie, it wont fry now, it will take a good 2-3 weeks or more before it does something...

2. The powerleap site says... "* Meets Voltage Regulator Module (VRM) specifications 8.5, 8.4, 8.3, 8.2, and 8.1." And, (please correct me if I'm wrong) They would be _WAY OUT_ of spec if the vtt had the same voltage as the vcc or changing vcc changes vtt in some way... And yes, I did read over the vrm 8.5 datasheet (not all of it, but a good deal :P).

And 3. I plan on doing this next weekend if I don't find some info to let me know for sure...


If you have some info about why powerleap may have gone out of spec for the ip3/t, let me know...

I will be going over the vrm 8.5 spec tonight (again) to see if I can find more info about all of this...

[ol' man]

Originally posted by Lyron
Well'sir...
I've got 3 things holding me back from shutting down right now, pulling it out, and soldering up a wire to a vtt pin...

1. If the vtt is being changed to the same voltage as the vcc, @ 1.60v I'm still under the Intel spec (Absolute Max). Ie, it wont fry now, it will take a good 2-3 weeks or more before it does something...

2. The powerleap site says... "* Meets Voltage Regulator Module (VRM) specifications 8.5, 8.4, 8.3, 8.2, and 8.1." And, (please correct me if I'm wrong) They would be _WAY OUT_ of spec if the vtt had the same voltage as the vcc or changing vcc changes vtt in some way... And yes, I did read over the vrm 8.5 datasheet (not all of it, but a good deal :P).

And 3. I plan on doing this next weekend if I don't find some info to let me know for sure...


If you have some info about why powerleap may have gone out of spec for the ip3/t, let me know...

I will be going over the vrm 8.5 spec tonight (again) to see if I can find more info about all of this...

Dude that is alot of work, I didn't think you would have to solder a wire to it to figure it out. Hey I understand if you don't want to do it. It would just be a good precautionary measure that is all. It would be nice to see the BX live on through one more round of licks and I would really like to see the mem scores at 133fsb on the BX.

As for a reason why. Simply the fact that they undervolt the chip on the slocket is my reasoning why they may have had to do a quick hack or simply did not want to do more work on it. I know there is a [H] thread about this adapter where a guy did the exact same thing as you. He said the vcore was at 1.32v on hids adapter. That alone would explain why everyone I know with this had only got the chip to 105fsb on the average. I at first thought there was something else majorly wrong with the adapter but then in the end I was starting to think maybe the vcore was low and low and behold I found out later it was.

[DeepScience]

Originally posted by Lyron


I'm not sure if this is a joke or not...
Tribes2 does segfault from time to time... But never more than once in 2 _DAYS_...
Are you sure your setup is stable?

heh, yeah mostly, well, no actually. hmmm if it's running sort of smooth now maybe I'll try reloading it again.

And i was trying to lighten things up a bit - sheesh, sometimes a little help isn't needed if you end up receiving an attitude along with it from the vtt guy. ah well.

Can u email me your article when you're done? Sounds great, good on ya

[ol' man]

Originally posted by DeepScience


heh, yeah mostly, well, no actually. hmmm if it's running sort of smooth now maybe I'll try reloading it again.

And i was trying to lighten things up a bit - sheesh, sometimes a little help isn't needed if you end up receiving an attitude along with it from the vtt guy. ah well.

Can u email me your article when you're done? Sounds great, good on ya

Spose I am kinda like grandpa in a ferrari. I tell the kiddies, do as I say not as I do:D

[zPacKRat]

I actually was able to use nail polish on my p3 700/cusl2-c. but i had to booger the hole where the painted pin went. It's a close fit. So that would be my first shot in getting higher vcore voltage.
that way ifin it don't work, you can undo it!


Originally posted by Lyron
"Does your BX board have a 1/4 divder?"
Nope... I have the Bh6 rev 1.0x... But I have run this board @ 124 before with a very close hardware setup... I'm not too sure about 133fsb, but I would _REALLY_ like to be able to try it. And the voltage looks like whats holding it back (Duh).

I bounced this off my dad (electrical/optical engineer) and he said we may get lucky and find a lead comming off the pin/package on the back and we may be able to cut and add a jumper/switch.

Or,( just sitting here typing and thinking about it) putting some kinda cover over a pin.
Ok, anyway...

Just removing vid3 would give me 1.725.
And that would for _SURE_ give me 1488mhz. And if the pci/agp bus holds up, a VERY good shot @ 133fsb for 1596mhz.
I've got good but not super good cooling on my chip. Globalwin fop32-1 (http://www.globalwinusa.com/products/fop32_1.html)
So, I should (keyword) be able to do 1.725vcc with this... I may start to run a little hot... I've also been looking at getting a new heatsink if it runs too hot...

Wish me luck on reading vcc/vtt...

Edit: Btw, I was able to do 980mhz @ 2.05v with my Celeron2 633 and the heatsink was just warm to the touch... The bh6 has no cpu temp sensor monitor so I can't give ya a true temp... So lets hope I'll be able to run @ 1.6gig/1.725v

-Lyron

[Ruiner]

We're still waiting for some benchies on your nice OC...:o

[Lyron]

Originally posted by Ruiner
We're still waiting for some benchies on your nice OC...:o

Ok then...
Here are some random "benchies" that I have been writing down in a little text file...
All 3dmark2001 runs with Gf3 ti-200 @ 220/487 (23.11 drivers under win2k)

Celeron2 633 @ 980 cas2-2-2
Sandra 2001te:
Cpu: Alu 2630, fpu 1316
Mem: Int 331. Float, 375
3DMark2001 (Build 200): 5061

Celeron2 1.2 @ 1.2 cas3
3Dmark2001: 5893
cas2-2-2 Sandra:
Mem, Alu 316, fpu 338
103fsb cas2-2-2 sandra:
Mem, Alu 341, fpu 360

@ 1.344 cas3 Sandra
Cpu, alu 3785, fpu 1807
Mem, alu 281, fpu 309

@ 1.488 cas3 Sandra
Cpu, alu 4189, fpu, 2004
Mem, alu 313, fpu 346
3Dmark2001: 6554
2-3-3 mem alu 328 fpu 356
2-3-2 mem alu 367 fpu 398
3-3-2 mem alu 344 fpu 379
2-2-3 mem alu 351 fpu 381
2-3-3 (with Precharge control enabled) mem alu 338 fpu 376
2-3-3 (Precharge on) 3dmark2001: 6596

Only 2-3-3 (Precharge on or off) was stable...
But thats what I get for getting cheap pc133...
2-2-2 would bsod before win2k got done booting...

Also, when I'm talking about cas settings, I'm talking about:
a. SDRAM Cas Laency time
b. SDRAM Ras Recharge Time
c. SDRAM Ras-to-Cas Delay
So 2-3-3 would be Cas Latency time = 2, Ras recharge time = 3, and ras to cas = 3
Not sure if thats the way everyone does it, but whatever...
I'm not sure if I want to spend cash on getting some pc133 cas2... It would help in some games like q3, but I don't play much q3... But at 70bucks for a 512mb stick, I may just do that... Heh, one last thing, would a 512mb stick work in a bh6 rev 1.0x?

[ol' man]

Originally posted by Lyron


Ok then...
Here are some random "benchies" that I have been writing down in a little text file...
All 3dmark2001 runs with Gf3 ti-200 @ 220/487 (23.11 drivers under win2k)

Celeron2 633 @ 980 cas2-2-2
Sandra 2001te:
Cpu: Alu 2630, fpu 1316
Mem: Int 331. Float, 375
3DMark2001 (Build 200): 5061

Celeron2 1.2 @ 1.2 cas3
3Dmark2001: 5893
cas2-2-2 Sandra:
Mem, Alu 316, fpu 338
103fsb cas2-2-2 sandra:
Mem, Alu 341, fpu 360

@ 1.344 cas3 Sandra
Cpu, alu 3785, fpu 1807
Mem, alu 281, fpu 309

@ 1.488 cas3 Sandra
Cpu, alu 4189, fpu, 2004
Mem, alu 313, fpu 346
3Dmark2001: 6554
2-3-3 mem alu 328 fpu 356
2-3-2 mem alu 367 fpu 398
3-3-2 mem alu 344 fpu 379
2-2-3 mem alu 351 fpu 381
2-3-3 (with Precharge control enabled) mem alu 338 fpu 376
2-3-3 (Precharge on) 3dmark2001: 6596

Only 2-3-3 (Precharge on or off) was stable...
But thats what I get for getting cheap pc133...
2-2-2 would bsod before win2k got done booting...

Also, when I'm talking about cas settings, I'm talking about:
a. SDRAM Cas Laency time
b. SDRAM Ras Recharge Time
c. SDRAM Ras-to-Cas Delay
So 2-3-3 would be Cas Latency time = 2, Ras recharge time = 3, and ras to cas = 3
Not sure if thats the way everyone does it, but whatever...
I'm not sure if I want to spend cash on getting some pc133 cas2... It would help in some games like q3, but I don't play much q3... But at 70bucks for a 512mb stick, I may just do that... Heh, one last thing, would a 512mb stick work in a bh6 rev 1.0x?

Yeah it should. I have heard of many getting their BX boards to take 1 Gb of memory.

You know at 1488 your 3dmark2000 scores are very high man. Much higher than what they have on the first page regarding the celery vs. XP crap. I am going to make a post about this. Feel free to cut in. Do you have a older gf2 to do some 3dmark testing with?

[Lyron]

Originally posted by ol' man


Yeah it should. I have heard of many getting their BX boards to take 1 Gb of memory.

You know at 1488 your 3dmark2000 scores are very high man. Much higher than what they have on the first page regarding the celery vs. XP crap. I am going to make a post about this. Feel free to cut in. Do you have a older gf2 to do some 3dmark testing with?

Btw, thats 3dmark2001 :)
And yes, I do have a gf2 mx in one other comp... If you REALLY REALLY (Really) want me to pull it out, I could, but I'm lazy... So you'll really have to bug me about it...

Oh and btw what is this "celery vs. xp crap" page you speak of? Is it this http://www.overclockers.com/articles494/ ?

[ol' man]

Originally posted by Lyron


Btw, thats 3dmark2001 :)
And yes, I do have a gf2 mx in one other comp... If you REALLY REALLY (Really) want me to pull it out, I could, but I'm lazy... So you'll really have to bug me about it...

Oh and btw what is this "celery vs. xp crap" page you speak of? Is it this http://www.overclockers.com/articles494/ ?

Yep!

I also wanted to ask if you were using 2X agp speed?

And if so how much of a affect does this have on the performence of 3d games over 4x speed?

[Lyron]

Originally posted by ol' man


Yep!

I also wanted to ask if you were using 2X agp speed?

And if so how much of a affect does this have on the performence of 3d games over 4x speed?

BX boards only have 1x and 2x. So I have no idea. I'm sure it would help 3dmark2001 a little. 100-200 points or so... But with games that don't push much over the agp bus, I would think it may only be a 1-2fps at best...

Oh, one last thing, I may start a new post for this... My old Celeron2 633, I would like to get it running in an old ex chipset board that has a Celeron300a @ 333 right now...
But when it posts with the 633, it reports a cpu speed of 93mhz and hard locks... ANY ideas? And yes, I have the lastest bios update...

[Ruiner]

3dmark2k1 scores that high are not surprising. That benchmark is almost all vid card, and really favors the DX8 features of the GeForce3.

Q3, wolfenstein and dronez (at low resolutions) love mem bandwidth and scale directly with it.
(like here, AthXP Alimagik vs. Via266a: http://www.overclockers.com/articles497/ )
Those are the same programs that the P4 loves when given RDRAM or DDR.

[D2DM]

512MB dimms on BX..

I think that 's not possible.
256MB dimms = maximum.

1GB on BX as ol'man says can be done yes. with 4x256MB.. Like the abit BX6.
For 256 MB I recommend (16x8)x16 architecture based dimms.

---------------

AGPx4 is not much better than agpx2.. no big impact.

Maybe because at 133fsb the Ram systems pumps data at 1066MB per second.

And agpx2 on a BX communicates with that RAM at (133x0.66)x2=178mhz = ~713MBps.

If a bx would support AGPx4 the max for agp would be higher but anything above the max for ram would not be helpfull..

---------------

and'em ehh BH6 rox :)