BSEL Mod on a socket P explained with photos

Not a gamer; but GT 540M or even 555M is not 'devasting' for gaming...

The price sure looks attractive - but I would be seriously looking at this with 8GB RAM and the i7 2670QM (at a minimum) at this point in time.

If this system is dual hard drive bay capable: buy one with the cheapest HDD's you can order it with (or, make sure that the wiring/connectors are available/installed when you do want to upgrade this and put in an SSD+HDD combo).

Is there a reason why their systems are offered so cheap? Usually, you get what you pay for.

 

From what I have heard they're actually just simply amazingly designed laptops and price match according to the current economy. They design their laptops to upgrade with ease. Removing 1 large panel underneath to get to everything including the cpu, memory, and gpu. That alone makes it amazing. They also take a lot of care in making sure everything stays cool. I'm now considering the ASUS g53SW which is $899 and offers the 2630QM and a GTX 460M MXM board card. They say it isn't upgradable. I find that extremely far fetched. MXM cards are expensive. According to Eurocom laptop upgrade for the Racer laptop it's pretty expensive but they sell all types of MXM graphic cards. The GTX460M is pretty damn good. I mean I'm used to my desktop which is an i7 930 @ 4,200Mhz with hyperthreading enabled, 6GB DDR3-1600, and 2 GTX470's at 800Mhz in SLI. My desktop is simply devastating. I mean when I started looking for another laptop to replace my gateway m-6827 core 2 duo t5750 and x3100 gpu. I just wanted something that I could simply play games on unlike this stupid gateway with intel graphics. I'm starting to think I want a laptop that's pretty portable but will kill new games at high settings. The downside to the ASUS G53SW is the fact you have to remove the entire damn motherboard to get to upgrading the laptop which kinda sucks but if you want something similar from Eurocom it's gonna run you atleast $1,200 unlike the currently priced ASUS at $899.99 and a $100 mail in rebate that'll bring it to $799.99 from newegg. I mean at this point the ASUS has mixed reviews but I love ASUS and I have had a lot of luck with them in motherboards. I hate to have to dismantle the entire freakin laptop to get to anything. It's a pain dismantling laptops, right? Anyway I can't decide if I just want something like the FOX 2.0 or the upgradable ASUS with the MXM board. I mean I'm not looking to go all out, i'm just looking for something better than my gateway. Say I buy the ASUS G53SW with it's 2Ghz quad with hyperthreading turbo mode core i7 that'll kick it up to damn near 3Ghz per core under load and that pretty damn good GTX 460M. Then later down the road the 460M is outdated, and according to Eurocom's site it's gonna run me between $320-$1,000 to upgrade the gpu mxm board to the 470M, 560M, 570M, (insane) 585M @ $790, Quadro 5010M @ $2,495(otherwise known as the price of a used civic), or even the Radeon 6970M @ but i want Optimus. So what do you guy's think? Sorry for moving this thread off subject but in a way it's kind of on subject (lol not really, you tell me).

My choices now are between the FOX 2.0 Core i7 or if i choose core i5 to save $44, the GT 540M or 555M which would damn well be plenty @ $858 with the core i7. Otherwise I could get the ASUS G53SW @ $899 with a $100 rebate. Or I could just say I don't really care and I want something that'll play all games at low setting and a core i3 or i5 at under $550. So I just don't know, later.

 

Probably off topic.

But a $320 - $1,000 'upgrade' is a poor way to view notebooks.

At that time, simply buy another (modern/current platform) with almost the same money (at least at the higher end of your upgrade estimate) for much more performance and a warranty too.

Don't think that all MXM cards are upgradeable: they're not.

Also, there is no such thing as an 'upgradeable' notebook. If there was, I would still be running my 20 yrs old ThinkPAD instead of the mushy things they call a 'keyboard' now on mobile systems.

With that kind of money to spend, think of spending it continuously (on new systems) rather than spending it all at once on a so called 'upgradeable' notebook or part(s).

The only thing worth upgrading a notebook on is the O/S, RAM and the HDD/SSD (if each is applicable and/or do-able).

Nobody price matches according to the current economy. They price their products to move. If they could sell them at the same volume at double the cost (to you/me), they would - in the wink of an eye.

 

So back to my merom laptop. I put windows 7 back on it and RMClock really's isn't doing anything other than voltages. I set performance on demand and locked it to x12 and it still goes up and down with no regard to my setting in rmclock. It also lags between 1Ghz and 2Ghz or any transition lags a bit. Windows xp transistioned much smoother. Also noticing a constant 50% cpu usage from rmclock and windows task manager shows 0%. Rmclock is insanely buggy in windows 7. Any suggestions on the perferct configuration for windows 7 to not control anything and all rmclock 100% control or something?

 

Not sure what to suggest since for me RMclock works perfectly either under XP or Win7. Are you running RMclock as an administrator?

If you're planning to lock the CPU under one speed\multiplier you should consider using the "power saving" and "maximal performance" instead of "performance in demand".

My CPU is locked at x6 (600MHz) in the "power saving" profile. I do see occasionally the frequency fluctuate to 800Mhz. This is not suppose to happen but it sometimes does. I can't tell you if this happens because of a bug in RMclock or if it because of the architecture of the CPU.

 

I uninstalled RMClock on my core2duo laptop and run through the windows power options and it actually works beautifully. You set your min and max frequencies and your profile of passive or active which tells how quickly to jump up the frequency. However windows power options drops the voltage to the minimum just like rmclock would do but on windows 7, the integrated power options are extremely efficient. Rmclock isn't intended for windows 7 but since windows xp didn't manipulate the voltages as much RMClock was necessary. Anway I'm getting great battery life without rmclock getting in the way. Locked at x6 0 min 0 max on battery at 1Ghz gets me like 2.5 hours of battery life compared to before when windows and rmclock were constantly battling over control it would only get about 1.5 to 1.75 hours on battery.

Later

*by the way my new laptop with the gt 540m is very good and it does dx11 tesellation excellently at 720p and all high at playable framerates which thats all that matters. It's a powerful fermi based chip.

 

Good to hear that you fixed it

 

What an awesome guide! Congrats!
I own the notebook in sig. As you can see I've replaced the cpu with a X9100 and I'm now wondering if it'd be possible to downvolt it via pinmod. Infact the minimum voltage is way too high (1,05V).
I use Throttlestop and I've seen that with that voltage the cpu is orthos-stable @2,93GHz (that is just a half multiplier below maximum clock).
At maximum clock (3,06GHz) the cpu is stable @1,0875V, while at default the voltage at 3,06GHz is 1,2375V.
Now I'm wondering, what would be the best way to lower voltage by pinmod? Especially the minimum, since it can't be modified by Throttlestop. But if possible the maximum too, since I dual boot Linux and there's no throttlestop for linux!

 

Edited on the 22/03/2012:

Vcore 1.2375v, means VID6543210 = 0010101

You can try VID3 + Vcc

Before mod ------------- after mod
1.2125v to 1.2375v --> 1.1125v to 1.1375v
1.1125v to 1.2000v --> 1.1125v to 1.2000v (remains the same)
1.0500v to 1.1000v --> 0.9500v to 1.0000v

This should allow you to go as low as 0.95 while still being able to boot @ 3.06GHz. The only issue with this mod is you'll be losing access to all the Vcores between 1.0125 and 1.1000 including 1.0125 and 1.1000.

 

Hi naton!
I don't understand: isn't VID3 + VCC supposed to lower the voltage by 0,1V? And not by 0,2V?

Thank you.

 

Yes you're right. My bad. I fixed the previous post...

 

Thanks Naton! I'll try this ASAP and will let you know the results!

 

hey people... i'm from brazil and need some help... first, sorry by my poor english...

I have an x200m chipset with one T2080... I do the pinmod 133-200 and +0,02v, working i little bit, don't take more than 3 pass on "intel burntest"... So i need to improve more 0,01v or 0,02v, to put in 1,3v at least and get steable... Anyone can help on it??? Thanks in advanced...

 

What's the default max voltage of your T2080? Do you want to add 0.1v or 0.2 v?
You can read this value in RMclock.

You should consider getting a T5300. A T5300 is faster than a T2500 even at stock speed. I had the T5300 running at 2.6GHz with an FSB pin mod from 133 to 200 MHz. the default max voltage was 1.25v if I remember correctly and I increased it by +0.01v to solve stability issue. I tried the same mod with a T7200 and it didn't like it.

 

Hi... Thank's for answer... I try the OV what you do in your T5300 on the first page of this topic... But can't get stable on my t2080... The max Vcore of my T2080 is 1.2125v... I like to put more 0.2v or 0.25v on it at least...Do you think is to high??? thanks in advanced... (sorry my english again)...

May i take an T7400, but not now...

 

Try with a small increase first and if the CPU is still not stable then go higher.

from 1.2125v to 1.2250v ----> connect VID0 to Vss (+0.0125v)
from 1.2125v to 1.2375v ----> connect VID1 to Vss (+0.0250v)
from 1.2125v to 1.2650v ----> connect VID2 to Vss (+0.0500v)

0.1v to 0.2v for a laptop CPU is too much. It will increase the temperature by a lot and your cooling system/fan won't be able to cope with it. My T7200 with the BSEL mod of FSB = 200MHz and stock voltage was reaching 90c in summer.

 

I man... Thank you... I'll try this on next week... I put in VID1+VSS and don't get stable... I go on VID2 and we'll see what happend!!! Bye...

 

Has anyone tried if doing the EIST unlocking MOD(BIOS Mod) often used for dual IDA, also solves the Multiplier-stuck-at-lowest-value-issue with Intel chipsets?
From my experienc what happens when doing the BSEL mod with an Intel chipset is just that the EIST is disabled and the EIST lock BIT ist set.
What I mean is there any proof that EIST is really disabled by the chipset and not the BIOS?

 

That's a nice idea.

 

I'm not sure exactly what you mean. I have quite a few CPUs around and a badly damaged Acer laptop with a GL960 chipset. It powers up and works fine but I think the GPU is gone so even if the motherboard blows on me when I do a mod I'm fine with it.

So what mod are suggesting we should try? I mean what pins we should connect?

 

Did the 133->200Mhz BSEL mod on a 2510P which already had a EIST-unlocked bios. Result was it was locked at 200x6 @ 0.85V. Ie: lowest multiplier and lowest voltage.

So what happens is the chipset locks EIST when it detects the chipset-sampled BSEL is higher than the CPU-set BSEL. Sorry, no cookie.

 

I suggest we take a Laptop with Intel Chipset and dual IDA Bios mod(removal of EIST lock) and apply a BSEL-mod to a higher FSB.

My assumption is that the BSEL-mod = multiplier-locked-to-lowest-value-thing with Intel chipsets, is acctually caused by the Bios disabling EIST.
So if we we have a EIST-unlocked Bios we can reenable EIST in Windows and increase the multiplier -> et voila:
working BSEL mod with Intel chipset
(so much for the theory)

Edit:
Oh didn't see your post nando4 as it was on new page.
So it seems my theory has gone the way of all good theories

 

I have two older dell studio xps 1340, with a MCP79 chipset and 2.53 P9500 cpu

after nearly completing a successful egpu on my other machine (a lenovo t520) ... was looking for some overclocking help on this dell, and what can be achieved.

not sure entirely were to start, but have a feeling some kind of pinmod can be achieved since it uses a nvidia chipset. ( did a search but no one with this machine came up)

any hints or tips of where to start would be useful.

 

The NVidia chipset won't lock the CPU into lowest multipler if the chipset and CPU BSEL sampled signals differ. However, your P9500 is a 266Mhz CPU. Only place to go to overclock would be 333Mhz. I doubt your chipset supports that speed so a CPU BSEL pinmod is likely out and a PLL pinmod the way to go. See http://forum.notebookreview.com/har...pll-pinmod-overclocking-methods-examples.html for examples of PLL pinmods.

 

If you have an intel Chipset and an nVidia GPU (which I think you have) the pin mod will not work. Your CPU will lock to its lowest mutiplier.

 

Please

How to do a pin mod in a core 2 duo t6500?


It is possible?

Thanks!

 

What chipset do you have in your motherboard?

 

Only a NVidia or AMD chipset system allow BSEL pinmodding overclocks such as a a T6500 from 200 to 266Mhz.

When trying that on an Intel chipset it will lock the CPU in the lowest multiplier (x6) and run at lowest voltage with no options to change them.

So for Intel chipsets need to do a PLL pinmod to overclock a T6500. See http://forum.notebookreview.com/har...pll-pinmod-overclocking-methods-examples.html for details.

 

Is it possible to do a BSEL mod from 266MHz to 200MHz on Socket P? If yes - can anyone post a picture on how to do it?
Also, will the multiplier be stuck at x6 if I use an Intel chipset + Intel onboard VGA?

 

See http://forum.notebookreview.com/har...intel-gl960-gl40-useful-info-pll-modders.html for details on the 266Mhz to 200Mhz BSEL mod. No x6 multiplier lockdown is applied.

 

Hello, i am happy to to confirm that the BSEL mod worked on my Dell Vostro 1500 and installed a new Intel Core 2 Duo T9400 on it.

That is right, having the latest BIOS (A06) the computer recognized the new Penryn CPU and booted right away, which means that the Vostro 1500 is compatible with the first and second Penryn CPUs (around the time the Montevina chipset was released).

So, to all the Vostro 1500 users out there, you will be able to upgrade the CPU, making your laptop faster, cooler, or both.

Thanks to all the guys that made this possible by sharing their work with the rest of us.

 

I am trying to install a cpu Q9100 into a HP 8530w (which has a motherboard that officially doesn't support it).

If I put the CPU directly windows doesn't start.

If I use Voltmod from this guide to decrease volts by 0.10V I get to see windows for a few minutes before BSOD. (decreasing by 0.2V pc doesn't work or I did the voltmod incorrectly).

My aim is to make pc stable with this cpu.
What do you suggest?
Is there a fsb pin mod to DECREASE the speed?
I didn't try by using RMClock yet. do you think that would help?

I don't want to buy new cooling system or new hardware.
I just want to use it for a couple of months before buying a new machine and I NEED more cpu speed.
If the pc breaks I can live with it.
Any help will be appreciated!

 

you could try to lower the fsb to 200mhz. Check this thread out:
http://forum.notebookreview.com/hardware-components-aftermarket-upgrades/605383-fsb-downclock-mod-intel-gl960-gl40-useful-info-pll-modders.html

 

Please check the FSB downclock mod. The link for it is in my signature. This will allow you to run the Q9100 @ 1.7GHz by lowering the FSB to 200MHz.


To the moderator:
Why can't I put a link in my post without getting your approval first?
Did I break any rules? .... thanks

 

does a VID mod raise the voltage? also the VID mod pic for socket P are super tiny.

 

I don't get the socket M pinmods, the chipsets that support socket M cpus only support 166mhz and 133mhz bus. How do you get 200mhz on it?

 

i think the OP needs updated badly as half the pictures are just thumbnails.

 

This mod works great... I've got a t4500 using 0.975mv on all frequency (All the way to 2.3ghz). After doing this pinmod, I noticed cooler temps and better battery life so I know I am running at .875mv on all frequency

 

Hi, I have done a volt mod to decrease the Vcore by 0.1 but when I restarted the PC CPU-Z, and Throttle Stop showed me the same 1.25V for my CPU, I am using a T7500 with a Santa Rosa mobo on a Toshiba Satellite A205-S5804. The thing is before and after the Vcore is the same any thoughts?

 

i try undervolt t7500 and t4200, i use picture for socket P,
but any mod
connect wire vid4 to vcc, or trying other mod vid3 ti vcc, but
not affect on cpu voltage.. it still on same value for both cpu.

 

Hi there.
It is possible to work full cpu frequency?
I would like to reach 2.8 Ghz was upgraded from T4200 to T9600 using BSEL mod.
Acer Extensa 5630ez-ms2231 GL40 chipset.
Thank you for your help.

 

Hi I have an old Toshiba satellite with the same T2250 processor. I would like to pimp it up a little .As I understand this can be done by pin mod .Unfortunately I cannot visualize the posted photo with the hole connection..I do not know why...could someone post them again or just let me know which holes are to be jumped? I really would like to do this Tks

 

Sorry for the bump, but why can't I see the pictures?

 

I need them too, please!

 

Hi!
I try to pin mod a socket m o ati xpress200m on toshiba L30 to upgrade To T7200. Even 200m support 667, toshiba lock this facillity. Ho to do this pin mod? i need some picture to to this .I cant view the photo that are posted
Tank you!

 

I'm looking to squeeze a little more power out of an old laptop and was wondering if this works with the T7250. If possible I was wondering if someone could post this pin mod for
A) To overclock (if possible)
B) To lower a 1066Mhz FSB to 800Mhz
If it doesn't work with a T7250 I can just grab a cheap CPU that does work with it

 

Hello, I want to try pinmod 200 to 266 on my T4300, it should not work with my GL40 but I want to try it and the photo link's are dead at the first page, if someone can help me?
Thanks

 

Some of the pictures can be found using the WayBack Machine:
https://web.archive.org/web/2015030...od-on-a-socket-p-explained-with-photos.390696

However, none had the 200 -> 266 option. This is not documented on the T4300 series' specifications either; these only show the 200 MHz FSB. That being said, Intel appears to re-use the same table for various cpu series, so with a little common sense we can deduce two promising combinations that may do the trick:

Spoiler: BSEL on T4300 and others

So first choice would be L-L-L and secondary H-H-H. Since your current 200 equates to L-H-L that means pulling BSEL(1) to ground for the first mod and BSEL(0) plus BSEL(2) to high for the second:

Spoiler: T4300, BSEL 266 - option 1 (L-L-L)

Spoiler: T4300, BSEL 266 - option 2 (H-H-H)

For the second option you'll need two wires. I've chosen to link them together to one voltage pin since there's no VCC right next to BSEL(0) and you want to avoid having the wire cross other pins . The BSEL pins are just to id the fsb and require hardly any power, so a single VCC is perfectly fine.

Make sure to disconnect as many optional components as possible before powering on. At least, think this is still a synchronous fsb system, right? Hence the 'poorest' component will determine your fsb ceiling.

Good luck !

 

Yes 266MHz FSB config is L-L-L, You have to pull down BSEL1 to Vss, which is ground in socket. But it doesn't matter in this case, because chipset GL40 support max 200MHz FSB. You can't do mod 200-266MHz simply because it will not be recognise by chipset. it can be done but by altering BSEL signals directly on clock generator not in the cpu socket.

 

Might ClockGen (oh, the times ...) work on this model?

Spoiler: ClockGen v1.0.5.3

Code:

ClockGen Readme file
---------------------

Contact : [email protected]
Web page: http://www.cpuid.com/clockgen.php

ClockGen is a freeware and can be freely distributed.


Supported PLLs
--------------

Cypress CY28346
Cypress CY28551
ICS 94228
ICS 950403
ICS 950405
ICS 951402
ICS 951412
ICS 951416
ICS 951422
ICS 951446
ICS 951462
ICS 952505
ICS 952607
ICS 952618
ICS 954119
ICS 954123
ICS 954148
ICS 954519
ICS 9PLRS509
ICS 9PLRS587
IDT CV107
nVidia nForce2
nVidia nForce3 150
nVidia nForce3 250
nVidia nForce4
nVidia nForce4 SLI IE
nVidia Geforce 6100/6150
nVidia nForce 590
Realtek RTM 360-408
Realtek RTM 560-266
Realtek RTM 862-410
Realtek RTM 865-461
Winbond W83195-BG101


History
-------

---------------------------------------------------------------------------------------------
1.0.5.3 - February 2007
- Added support for Windows Vista 64.

---------------------------------------------------------------------------------------------
1.0.5.1 - December 2006
- Added nForce 650i/680i PLL.

---------------------------------------------------------------------------------------------
1.0.5.0 - October 2006
- Added Cypress CY28551 (Asus P5B Deluxe) PLL.
- Added Winbond W83195-BG101 (Asrock Conroe XFire-eSATA2) PLL.
- Added ATI RD600 memory PLL control.

---------------------------------------------------------------------------------------------
1.0.4.9 - May 2006
- Bug fix version.

---------------------------------------------------------------------------------------------
1.0.4.8 - May 2006
- Bug fix version.

---------------------------------------------------------------------------------------------
1.0.4.7 - July 2006
- Support for Asus P5WD2 (& Premium), P5WD2-E (& Premium), P5WDG2-WS, P5W DH (ICS 951423).
- Added ICS 954119, ICS 9PLRS509 (DFI 975X/G Infinity) and ICS 9PLRS587 (Gigabyte GA-965P-DQ6).
- Run at Windows startup feature added.
- Improved smbus stability on Intel chipsets.

---------------------------------------------------------------------------------------------
1.0.4.6 - May 2006
- Added ICS 954519.
- Added nVidia nForce 590 integrated PLL.
- Added Realtek RTM 865-461.
- SMBus support on Intel ICH8.

---------------------------------------------------------------------------------------------
1.0.4.5 - May 2006
- Bug fix version.

---------------------------------------------------------------------------------------------
1.0.4.4 - April 2006
- Bug fix version.

---------------------------------------------------------------------------------------------
1.0.4.3 - March 2006
- Bug fix version.

---------------------------------------------------------------------------------------------
1.0.4.2 - February 22, 2006
- Restored compatibility with Windows 98.
- Improved smbus support on Intel ICH6/7.
- Fix bug with RTM560-266.
- Fix AGP/PCI clocks for ICS 952618.
- Fix layout issue.

---------------------------------------------------------------------------------------------
1.0.4.1 - February 5, 2006
- Added memory buffer ICS 9P932.
- New icons.
- Snap improvement.

---------------------------------------------------------------------------------------------
1.0.4.0 - January 21, 2006
- Added ICS 951402, ICS 951462.
- Added RTM560-266.

---------------------------------------------------------------------------------------------
1.0.3.9 - January 9, 2006
- Use the cpuid global mutex.
- Fix smbus conflict with CPU-Z.

---------------------------------------------------------------------------------------------
1.0.3.8 - January 4, 2006
- Improvements in mutexes management.

---------------------------------------------------------------------------------------------
1.0.3.7 - January 2, 2006
- Added ICS 952505, ICS 951416, ICS 951422.
- Added vertical resizing to setup window.
- Fix bug in hexedit control.

---------------------------------------------------------------------------------------------
1.0.3.6 - December 12, 2005
- Added smbus support for ULi M1563/M1573/M1575.
- Close after dump save bug fixed (working directory is now the user's temporary directory).

---------------------------------------------------------------------------------------------
1.0.3.5 - December 6, 2005
- Added ICS 950403, ICS 950405.
- Added Realtek RTM 360-408.
- Added Cypress CY28346.

---------------------------------------------------------------------------------------------
1.0.3.4 - December 3, 2005
- Added RTM 862-410.
- Added ICS 952618.
- Fix "Reset Selection" bug.