BSEL Mod on a socket P explained with photos

Thank you for reply.

So 1,050~1,2625 V is the correct range of VIDs? OK. But isn't here any HW way of measutement the real VID? I read something about measurement on Vcc_sense × Vss_sense pins on the socket... Should I try it?

Yes, performance of this CPU is good, but there is too high power consumption (default: > 60 W, now: ~ 50 W in full load), so I'm trying to get this CPU running at 2 GHz@~0,9x V.

I'm not sure, what do you want to hear, but with 1,050 V set in ThrottleStop it's rockstable.
With mods -0,4; +0,2 V and ThrotleStop set 1,1625 V it's OK too and based on measurement of power consumption it was successful. (~10 W less)

No, this CPU wasn't running at 2,6 GHz, it was desktop C2Q Q8400. (Yes, I know, that each piece of CPU has different limits etc., but...)
Max. on this CPU was 2,35 GHz (276×8,5, resp. FSB+10 MHz (MSI "Turbo mode") × IDA-multiplier).


Regards,

JK

 

Hm, it seems to be "problem solved"
Probably my C2Q9000 can't run at less voltage, than 0,9125 V. But it has strange behaviour

Pinmods:
VID5+Vcc
VID4+Vss

==> VIDs are in the range decreased about 0,2 V.

When I have in ThrottleStop set 1,1625 V, everything is OK. I can go lower and lower. Everything is in full load of all cores stable even on 1,1125 V. When I run singlecore load, it makes errors. And when there is no load, system crashes (BSoD: MACHINE_CHECK_EXCEPTION).
Strange, I would except opposite behavior - crash while the system is under full load... :/

Any idea? Or I am on the limit of this CPU?


Regards,

JK

 

Hi JK

Thanks for the answering my questions.

I didn't try it myself by one member of the forum connected one end of two wires to Vcc_sense and Vss_sense socket holes and the other end to a voltmeter. Doing so he was able to read the voltage directly from the CPU.

1.2625v is VID6543210 = 0010011

With VID4 + Vcc:
1.1125v to 1.2625v ---> remains the same
1.0500v to 1.1000v ---> becomes 0.8500v to 0.9000v
Try this mod and set the VID in RMclock or ThrottleStop of multiplier x7.5 to 1.1000v and see if your CPU is stable. That's a real voltage of 0.9000v.

With VID3 + Vcc:
1.2125v to 1.2625v ---> becomes 1.1125v to 1.1625v
1.1125v to 1.2000v ---> remains the same
1.0500v to 1.1000v ---> becomes 0.9500v to 1.0000v
If VID4 fails VID3 might work since the minimum accessible Vcore is 0.9500 instead of 0.9125v.

With VID5 + Vcc and VID4 + Vss:
1.1125v to 1.2625v ---> becomes 0.9125v to 1.0625v
1.0500v to 1.1000v ---> remains the same

Does the Q9000 support IDA? I've noticed in my laptop that I had to set the voltage about 0.05 to 0.1v higher in IDA mod to avoid BSOD. It is a bit odd. IDA is supposed to kick in when only one core is in use, and since the other core is disable the power consumption of the CPU should be lower. Hence, the voltage in IDA mod should be lower. But in reality it is not the case.

I like the VID3 mod since you can get as low as 0.95v using only one wire

 

Hello.

Yes, this CPU supports IDA, but it works strange (it switches to 8,5× multi in singlethreaded load, but only for a few seconds). Dual/Quad-IDA via ThrottleStop doesn't work. I'll try to do some flashes of BIOS (older BIOSes had more advanced settings), maybe this will help...
Anyway, I have set in TS multi 7,5× fixed and since that time there hasn't been any crash.
I don't know, why has the IDA-multi higher voltage. Maybe it's just "for all cases", but it's weird...

=> I will try to do this.

And yes, VID3 + Vcc is easy-to-do, but not low enough


Regards,


JK

 

Hi!

I did the Pinmod (both, 133->166; 133->200) with a Celeron M 530 1,73ghz exactly like in the first post and the laptop started normally.
But nothing was changed. No Program shows any difference, but it's getting very hot.

The chipset is an SIS 672+968 in a FSC V5535
The core of the Celeron is a Conroe-L (like CPU-Z and some other progs show), or maybe a Merom-L (like RM CPU Clock and some others show?!).

Does anyone have an idea why it doesn't work, but is getting hotter then before?

 

If you did the mod properly CPU-Z will show 2.6GHz when going from FSB 133 to 200.

It is likely that one of the wires is shorting out. Just try the mod again.

The SIS chipset is an excellent chipset since it doesn't lock CPUs in their lower multiplier after mod. I pin modded a Celeron T1400 to 2.16GHz (FSB 133 to 166) and a T2310 to 2.2GHz (FSB 133 to 200) and a T5550 to 2.2GHz (FSB 166 to 200) in a laptop with a SIS chipset. If you're consider using a dual core CPU you have to use DDR2 PC4300 RAM instead of PC5300 RAM or flash your RAM.

 

Hello,

I tried to do this:

The mod is done correctly, but the effect is very strange. I run CPU load and the power consumption was > 80 W with SW VID = 1.100 V. With SW 1.050 V it was about 62 W (and consumption like this was on CPU default).

I'm turning back to previous mod (VID5+Vcc; VID4+Vss) - that was working without problems.


Regards,


JK

 

Hello again!

naton, I don't know, which change I did (only change the thermal compound to Arctic Cooling MX-4 and cleaning inside of the ntb), but now it's really great!

Mod is VID4+Vss, VID5+Vcc, BIOS - EIST off, ThrottleStop VID 1,1125 V.

Now I have power consumption in full CPU load only about 47 W (original consumption was over the 65 W) and temperatures only about 60°C (originally about 80°C). Rockstable, of course. Great, great, great!

Big thanks for support and +rep!


Regards,


JK

 

Could it be that before applying MX-4, the CPU temperature was high and the fan was spinning none stop? That could explain the high power consumption!

 

Maybe there was something bad with the mod (when I changed the thermal compound, I reworked that mod) , who knows... Anyway I won't touch it, when it works great now
Fan speed depends on CPU temperatures:
First step: 10~50°C
2nd: 50~55°C
3rd: 55~60°C
etc.

Unfortunately the fan spins nonstop, but at the lowest speed it's acceptable by my ears.
The power consumption of fan is insignificant (max. 2,5 W between off/max_speed states).

 

yours spins non stop and the fan in my friend's Toshiba laptop doesn't spin enough. As a result his CPU was reaching 93c before applying MX-3 and undervolting the CPU. Now the max temps are around 80c.

 

I'm stuck with a "not stable" overclock over time. I was looking into how to increase CPU voltage, but I think it is more my RAM now. Is there anyway to incrase voltage for my RAM? Or what do you think?

http://forum.notebookreview.com/7796135-post257.html

 

I have no idea about how to increase the voltage but I think I know how you can get your laptop to work without reverting the mod.

The FSB/DRAM divider your laptop is selecting for your RAM is too high. What I could suggest (if you have a flashable RAM) is to decrease the RAM default speed. For instance, I pin modded a T2310 from FSB 533 to 800 with a PC5300 RAM. My FSB/DRAM divider was 1/2 when the FSB was 200 so my RAM was set at 800MHz. I used SPDtool (you can download it from here) to make the laptop believes that the RAM was a PC4200 (266MHz) instead of a PC5300 (333MHz). With this little trick the laptop selected a FSB/DRAM of 3/5 and thus the RAM was operating at 667 instead of 800MHz.

The laptop I used for the mod had only one memory slot. If yours has two, instead of using SPDtool install a slower RAM in the second slot. This will force your laptop to select a lower memory divider

 

Thank you so much for your answer. I have alot of 2GB modules (1066Mhz). At this point, I think I want to flash my G.SKILL (1333Mhz) modules. Then I don't need to buy new set of modules either. I have not looked into SPDtool yet, but can I perma-flash my modules with it? If not, is there any other tool you can suggest for DDR3 RAM? There is a lot of old therads about this, however, mainly for DDR2.

EDIT: I have seen a bit around with SPDTool, confused. So, I want to set my sticks at 800 Mhz? That wil give me room to overclock.. But again, there is no easy way for this, any clue about which line I should edit?

EDIT2: I've found your therad:
http://forum.notebookreview.com/har...ades/515956-spdtool-experiments-findings.html
And can't belive not a single one have replied. World is not a fair place, however, with this, I can say that my SDRAM Cycle Time at Maximum Supported CAS is at 5.20 ns (192Mhz). (Picture)

 

For DDR3 RAM, I think thaiphoon burner is best.
A working trial version is attached at the end of this post:
http://forum.notebookreview.com/har...pll-pinmod-overclocking-methods-examples.html

 

The changes does not stick with SPDTool, as far as not CPU-Z can read however. I'll try this, thank you.

EDIT: with I get a error message:
Could not install system driver!

SPDTool, for sure does not save my settings. Also tried overvolting one of my sticks, did not work. thaiphoon burner needs 32-bit OS (?). I'm back where I started. I now only have my 8GB G.SKILL RAM modules (1333mhz) insted of 4GB Samsung sticks. However, it downclocks to 1066mhz and CL7.

Any ideas?

 

Thaiphoon burner sometimes has trouble with x64 windows, you might have to troubleshoot it. First thing I would try is to disable driver signing.

 

Nope, tried that.

 

I tried everything i could find on the net but still can't get Thaiphoon Burner to work under Win 7 64bit.

tuηay did you restart your computer after reflashing the RAM with SDPTool?
CPUZ (and the rest of the laptop I believe) requires reboot to show and detect the new values.

EDIT:
Isn't CAS at 5.20 ns too low for your RAM?

So you are able to go from 1333 to 1066MHz. Is your laptop stable after setting the RAM to 1066MHz?
What is the FSB/VRAM divider shown in CPUZ after reboot and after setting the RAM to 1066MHz?

 

Yes. I did a restart everytime after *write*.

That was defult for the 1066Mhz Samsung sticks, atleast SPDTool said so. My 8GB kit from G.SKILL is orginally 1333Mhz stick modules, so they automaticlly downclock to 1066Mhz (My chipset/cpu mhz). But, I can't overclock them up to 1333Mhz again with SetFSB. My whole system crash/freeze.

 

The PM45 will run the RAM at 1066MHz if the CPU FSB is 1066MHz.

Can you try to lower the frequency of the G.SKILL to 800MHz (400MHz)? or is 1066MHz the limit? this would force the BIOS in your laptop to select a different FSB/VRAM divider instead of 1:2.

I'm not an expert overclocker, but might be you're already reached the maximum capacity of your CPU at 3.49GHz. The T9300 and T9500 according to various posts seems to hit a wall at around 3.5 - 3.6GHz.

The picture in the link shows the voltage as 1.15v. Did you try to increase it to 1.2 or 1.25v? Also is you system stable if IDA is dis-activated?

 

What line in SPDTool is that?

The thing is, about my CPU, I'm stuck at v1.15 @ 11.5x. Other users, such as corssfire_gti (here and down this therad: http://forum.notebookreview.com/ace...spire-5739g-owners-lounge-24.html#post7716116) can get 1.175. I can't, they have it defualt at 11.5x. BUT, sometimes, random, my CPU kicks to 12(!)x and raise the voltage, by itselfs all the way to 1.125(!). DUAL IDA is not enabled and I can't lock the x on TS (thorttlestop) althrough I can select it.. Only locks to 11.5x which is stock.

I have modded BIOS, there is a lot of settings there, but mostly C4, C5, deep sleep etc. settings. Wish I could change the mhz there

I know not all the CPU's are the same, but 3.5Ghz is seen by crossfire again,
http://forum.notebookreview.com/ace...spire-5739g-owners-lounge-25.html#post7736490

 

I even got it on screenshot here, nothing intence, just skype convo with my gf. It still kicks to 12x and voltage to 1.250, if I just could get it to stay there....


EDIT: Also noticed that, it can kick up to 12x even on 0% off usage You can see what's running down there...

 

Do you want to try to up the voltage via pin mod?

I can help you with that.

 

With my G.SKILL 8GB RAM, I can't even overclock to 3.31Ghz (333 fsb). So, I think, I also have to do something with my RAM sticks..

EDIT:
Can anyone confirm that thaiphoon burner works with 32bit? I can just install 32bit OS on my other HDD and flash my sticks from there. SPDTool still raports 5.20 ns.

 

I'm having some problems with the pin mod for undervolting my socket p intel core 2 duo t5750. First I did VID3 to VCC, then I tried VID4 to VCC, and finally I did VID5 to VSS and VID4 to VCC because someone here posted they dropped their power consumption from I think 65W to 47W under load. That's what I'm looking for. However after modding my battery life remaining time doesn't really change nor does the cpu temperature. It doesn't run hot anymore since I used IC Diamond on the GM965 chipset and the cpu core. It runs pretty cool. Idling at 54C in a 76F ambient room at the moment. I use AIDA64 to monitor voltages and to dial RMClock into AIDA, I have to set RMClock to desktop cpu mode. Minimum voltage of 1.0625V @ x6 multi and I have x12 multi at 1.100V. Is there something different about my setup? Does the pin modding for under volting work on GM965 chipset? Also shouldn't I see a change in voltage? It seems like nothing changes honestly. I almost think it may be running hotter now than before. Info and help would be highly appreciated. The max voltage is 1.3625V. I want to get the lowest possible voltage for the best battery life. I almost think it's not even changing since I'm using Windows 7 x86 and RMClock. I may have to install windows xp to really see some voltage changes. My battery life is erratic. I'm on a 4400mah battery which came factory. Sometimes it'll say 3 hours then I load something and it'll drop to an hour and 15 minutes. Maybe I'll just pick up a 7200mah battery which I plan to do anyway and more if they're available.

Thanks brotha's

 

So what your saying is in order to adjust the clock speed of the intel chip you have to either scratch out the circuit link between the pll clock generator and the north bridge in order to sustain the max multiplier on an intel chipset? Also say I just to undervolt my intel chipset. Are you sure the same under volt mods work on our intel chipsets? Right now windows 7 is driving me nuts and no management software is working at all. Windows always wants to have complete control. I hate hate how this chipset is locked down so tight. So say I set the pin to 166 to 200. Boot up into windows. Can you remove the lock on the multiplier within windows? Can you confirm that VID3 to VCC, VID4 to VCC, and VID5 to VCC all really do undervolt the processor?


THANKS

 

If the chipset is an intel chipset the BSEL mod will only work with a Celeron and a Dual Celeron CPU. Win a Pentium dual core or a Core 2 Duo, the CPU will lock at x6 multiplier.

The BSEL mod is a hardware overclocking mod. The alternative is the software overclock with tools such as SetFSB. The PLL mod is required if the clock generator is locked to prevent overclock. The PLL mod basically consists on removing the lock so that the FSB of the system can be increased with SetFSB.

No one knows how to remove the multiplier lock on laptop with an intel chipset .

VID + Vcc will decrease the CPU voltage when used in conjunction of a tool such as RMclock or ThrottleStop. There is a bit of math involved in selection the Vcores in RMclock or similar software after a VID mod. Tell me more about what you want to do with the VID mod so I can help you .

 

There is also a second type of PLL pin mod, which is hardware overclocking. It's the FSLx mod.

It's method 3 here:
http://forum.notebookreview.com/har...pll-pinmod-overclocking-methods-examples.html

The PLLs FSLx pins match up with the CPUs BSEL pins.

 

question #1
Ok, so firstly I want to ask about overclocking. I have set the BSEL for 166 to 200 before and witnessed the lockdown. However bios reported 2400mhz compared to 2000mhz. I booted into windows and the multiplier is locked. So if I want to overclock, can the multiplier be unclocked within the OS manipulating the PLL Clock Generator? Also do you think a VID3 to VCC plus the BSEL 166 to 200 with a pll multiplier unlock likely possible?

question#2
Regarding undervolting. So your saying when i set say VID3 to VCC or VID4 to VCC, that in RMClock it will still show as the same voltages as stock however the "real" voltage will be different? From my experience with vista or windows 7 you cannot manipulate the voltages hardly at all. Windows wants to control everything or it's just the god damn intel chipset wanting to control everything. What do you think? One thing that confused me is AIDA64 reported RMClock's desktop mode voltages, however CPUz and CoreTemp reported Rmclock's mobile voltages. Another thing I noticed is not no matter what I did to adjust the voltages and manipulate the multiplier in RMclock. AIDA would most of the time not reflect the voltage changes. However ocassionally it would but temps wouldn't change that much. So what I have done is installed XP Pro finally after 12 hours of creating a usb flash drive with the ICH8M ahci drivers integrated which was a pain. Anyway now I'm in XP, and I have yet to start modding again. I will update on my new management experiences with the mod in xp later. I want to be clear on what I want to accomplish and what needs to be done to accomplish that goal. My first priority is to extend battery life and lower my temps across all multipliers. I want to get atleast a 100mv reduction at x6 multiplier. I'll explain how I set up my RMClock. I use performance on demand on startup and current for both profiles. I set the x6 multi to the minimum allowed voltage (mobile mode) which shows 0.950V. I then usually set my x12 voltage to 1.100V and have been able to reach 1.087V under load at 2Ghz as well. I would also like to comment on the fact that it's impossible to pass Intel burn test on a laptop. Any experiences with intel burn test on a mobile chip would be interesting to hear. Anyway what I would prefer is voltages as low as say 0.650V at x6 in RMClock and with no need for voltages above 1.200V on x12 multi.

THANK YOU VERY MUCH!

btw - loving this laptop modding especially the intel chipset since it's like playing god or something since intel provides extremely restrictive chipsets which tend to upset us "Enthusiasts". (Felt like it was funny I believe it auto censored the word "p--i--s--s" Funny

One more thought. What about a modded bios for my Gateway M-6827 that unclocks the pll multiplier and possibly allows for a broader range of voltages by stock? Another thought is what about custom modding the ACPI specifications? I saw in Notebook hardware control had a huge range of custom ACPI adjustments. As well as a voltage table that went outside the range RMClock showed but seemed completely 100% useless since applying those voltages in Notebook hardware had absolutely zero affect on actual operating voltages. I think software has an extremely difficult time getting a strap on the inner workings on the intel gm965 chipset.

Later brotha's!

 

After you do the BSEL mod, there is no way to unlock the multi (afaik).

 

It is weird that you've problems with RMclock and Win7. The laptop I'm using to write this reply has Win7 64bit, a GL40, and a T7300 CPU and I don't have any problems with RMclock.

Can I have the minimum voltages for x6, x7,..., x12. Also I need to know your CPU maximal voltage @ x12. You'll find it in the "CPU info" tab in RMclock.

 

x6 @ 0.950, x12 @ 1.250. You tell me what voltages I need to set to get 0.4 reduction in voltages. Right now I bridged VID4 to VCC and I'm not seeing any changes. I'm in windows running fine.

THANKS

 

I just did the VID4 to VCC and I'm not seeing any difference from stock. I just don't think they make any difference at all. I'm running a Gateway M-6827 with an intel mobile 965 chipset and an intel core 2 duo t5750 @ 2Ghz. I now have xp pro sp3 installed and have more control over voltages and multipliers than I did in windows 7. I agree it is unusal however for my specific laptop I'm having much more success in windows xp.

Thanks

 

So now I installed RMclock and I can run all multipliers on the 0.9500V and they're are all hot as hell and all run perfectly stable at the minimum voltage allowed. Sitting at 60*C at x6 and 60+*C at x12. So I think that no one understands the socket P and the GM965 chipset. By the way the picture of the socket p on page 1 is not the same as my socket. Which is Socket p. I forgot the designation on the socket. It says like mpga478 or something. I'll confirm latter. So i think VID4 to VCC on my socket is actually increasing the voltage since all multipliers are perfectly stable and very hot at the minimum allowed voltage in RMClock. I'm going to remove the pin mod and confirm the pga socket type.

 

I did a lot of research. My socket is indeed a socket P socket. The socket is mPGA478MN. As far as I have seen. There are no pin out identifications available for this socket. Nor does anyone no a single thing about it. The socket p you have listed is mPGA478M. Not the MN in my system. So I'm out of luck and damn lucky I didn't fry my board and cpu. If anyone can find info specifically regarding the difference between mPGA478M and mPGA478MN please let me know. If you can find an intel technical engineering article regarding the mPGA478MN please let me know.

THANKS

 

Ok, so I removed the pin mod. Now I'm running about 15*C less than with it. Here's a pic of the socket that's in my laptop.

 

Correct me if I'm wrong, but it doesn't matter what your socket says, the t5750 has only one pinout.

Look for the pinout for the t5750, not for the socket.

 

Your right possibly.... I'll try and find a t5750 pin out. Good idea. I also can't even believe I was able to find this... May help with a hard overclock or something... This is the schematic for my specific hardware in my laptop. May help with my PLL Clock generator lock on intel chips. WHat do you think?

 

Looks like your pll is one of these 2:

RTM875t-606:
http://forum.notebookreview.com/5967340-post85.html
http://forum.notebookreview.com/har...t-606-info-gathering-ocing-9.html#post6707491
http://forum.notebookreview.com/har...clocking-methods-examples-23.html#post5457881

Or ics9lprs365:
http://forum.notebookreview.com/har...clocking-methods-examples-43.html#post6282117
http://forum.notebookreview.com/har...clocking-methods-examples-60.html#post6804435
http://forum.notebookreview.com/har...clocking-methods-examples-15.html#post5353672
http://forum.notebookreview.com/del...21279-inspiron-1525-successful-overclock.html
Has anyone successfully overclocked p7805u CPU? - Page 22
http://forum.notebookreview.com/acer/438415-1810t-tme-mod.html#post5565888
http://forum.notebookreview.com/acer/438415-1810t-tme-mod-3.html#post6021417

Both very common and have been modded before.

 

Thanks for the above info. Anyone know of a way to overclock the integrated video X3100? I'm getting BSOD before entering Need for speed hot pursuit (11') at the moment. Happened with standard and modded drivers. Anyway I did a lot of research and the datasheet from intel for the socket pin identification should be correct. However, I think VID5 to VCC for me what outside of my cpu's useful range and forced higher than standard voltages. So say I want to reduce my minimum of 1.0625V to say below 1.0000V. I know that VID3 to VCC may work but about VID 3 to VCC plus VID2 to VCC. Would that combine voltage reduction? Or maybe VID4 to VCC and VID3 to VCC to give 0.3500V reduction. I'm happy with my xp setup right now. I just really think the minimum is way to high. I can run stable at 2Ghz very close to the minimum so I guarantee I could run stable at 0.8500V or so at 1Ghz (x6) Multi. I just don't understand the voltage tables. I understand that taking a VID to VCC forces it to a 1 I believe? The voltage tables show 0's and a couple 1's. Does that mean you bridge several vids to get the desired range? Please tell me how to just get my minimum of 1.0625V reduced quite a bit while retaining anything up to 1.2000V at x12 multi. So I want my new range to be say 0.7500V-1.2000V. Please help. Thanks

 

I doubt that the latest NFS hot pursuit will work with the x3100. NFS Prostreet which is 4 years old runs like a dog on the X3100. The X3100 and newer Intel GPUs cannot be overclocked. There are no tools to overclock them. Intel is not given away enough details about its GPUs to allow the development of overclocking tools.

VID5 + Vcc wont increase the voltage. It will decrease some values and keep the others at their initial value. Basically any Vcore with a VID5 value of 1 will remain unchanged after mod and any Vcore with a VID5 of 0 will drop by 0.2000v. For example:

Vcore of 1.2000v with VID5 + Vcc = 1.0000v (drops by 0.2v).
Vcore of 1.1000v with VID5 + Vcc = 1.1000v (remains the same).

<s>I like to help you but you need to answer my questions first. Answering the following will help me work out the best mod for you:

1- Assuming you're using RMclock, please use orthos (a cpu stressing program) and find out the minimum voltages for all your multipliers between x6 and x12.

2- I also need the default max voltage and minimum voltage for your CPU.

3- What CPU do you have by the way?</s>

 

Okay. Below is a summary of your previous posts:
- CPU T5750/ x6 .95v/ x12 1.25v
- All multipliers stable @ 0.95v
- temps around 60c and above for multiplier x6 @ 0.95v

Have you ever opened your laptop, cleaned the heatsink, and applied a new thin layer of thermal past? Your heat issue is likely to be cause by the fact that the heatsink is blocked by a layer of dust. Cleaning the heat sink and applying a new layer of thermal past will reduce your temperature considerably.

The T5750 @ x6 - 0.95v should be around 40 - 45c max when idle, and no more than 50c under load.

Since all multipliers are stable @ 0.95, your best mod is indeed VID4 + Vcc. This should extend your Vcore range to 0.75v. Note though that your CPU will lose access to all Vcore in the range of 0.9500v to 1.1000v. After mod set the following Vcore and stress test the CPU:

x06 - 0.9500v (real: 0.7500v)
x07 - 0.9875v (real: 0.7875v)
x08 - 1.0125v (real: 0.8125v)
x09 - 1.0500v (real: 0.8500v)
x10 - 1.0875v (real: 0.8875v)
x11 - 1.1125v (real: 1.1125v)
x12 - 1.1125v (real: 1.1125v)

As shown above your x6 to x10 should now be below 0.95v but x11 and x12 will have to be 1.1125v. You might be able to go lower for x7 to x11. I mean you can try to see if x11 for instance is stable at 1.1000v (real: 0.9000v). You should stress each multiplier individually to figure out your stable minimum voltages.

N.B.
All the sockets used for Core Duo and Core 2 Duo CPU are the same with 478 holes. The difference is on the connections.

 

My cpu fan just failed. I had removed it like 20 times cleaning lapping and reapplying IC Diamond thermal paste. My laptop runs hot. I hate it. So maybe the fan was dying. I even reoiled it before it "died". It sounded fine. Anyway. There aren't any merom's that run at 0.950V. I got support from intel and indeed the voltage range is 1.0625V minimum and 1.3625V maximum. The mobile voltages are incorrect. That would explain why coretemp and cpuz were displaying incorrect voltages. AIDA64 was displaying my correct voltages as well as RMClock's desktop mode. Also wiki states all the T series merom nearly all run 1.0V minimum. Not 0.950V. Anyway. So at x6 i can only go down to 1.0625V. Idling at 50-55. I run my x12 without a problem at 1.1125V. Under load the cpu hits 70C no problem. Probably more. I applied thermalpaste correctly and cleaned everything as well as did a mild lap. I use 90% Alcohol to clean surfaces. I know my stock heatsink piping looks very weak and it looks like a poor design. I want to upgrade to a higher cfm DC5V fan. Anway AIDA64 reports my voltage directly from the internal voltage regulator. It displays 1.0625V at X6 and a Max of 1.3625V. However at x12 I run at 1.1125V stable. That 1.0625V minimum is so freaking high for 1Ghz (x6). So say I apply VID3 to VCC would that lower my 1.0625V? I need to get some nail polish to insult the wire to not short it. My laptop is out of comission. I really hate it after all the crap I've been doing. I ran NFS pro street pretty playable at minimum graphics. However I hate intel GMA graphics. I also hate these high minimum voltages. I may just buy a new laptop. Say core i3 nvidia gpu. Maybe core i5, not sure. I want a laptop with overclocking/ voltage options in the bios. Know of any?

Thanks bud

 

Overclocking and voltage options in the bios, that would be mostly available in gaming notebooks, alienware/clevo/msi.

BTW I have a T7300 (mobile socket P merom CPU), min voltage was 0.85V and max was 1.1625V.

 

How are you monitoring your voltages? CPU-Z and coretemp are incorrect. According to intel most standard merom chips run above 1v at their minimum multiplier. AIDA64 displays the correct voltages via the voltage regulator. According to wikipedia the t7300's range is 1.0375–1.3 V. My T5750 is 1.0625V-1.3625V. According to wikipedia it's the same as the T7300. I would say wikipedia is 95% correct, with the other 5% being that tiny .0125V deviation. Rely on AIDA64 and not cpuz or coretemp. An intel support engineer told me merom's are reported incorrectly on such programs. He recommended AIDA64 Engineering edition since it's what they use to monitor their hardware. That in mind, will give you a new outlook on what to expect from your pin mod. I think you should download AIDA64 and then monitor your voltages to confirm intel, wikipedia, and my suspicion. If it is indeed below 1v, please post a screenshot from AIDA64 sensor page. By the way, what are your temps like? CPU, and Graphics Chip?

Thanks

 

I'm sorry but I can't check it with AIDA64 because I sold my graphics card...

My temps were really bad, because I had 1 fan cooling a CPU and an FX 3700m, 90C +.

I had my T7300 in a different notebook (toshiba tecra a9, intel x3100 gpu) for a while, here are the temps:
http://forum.notebookreview.com/del...k-powersaving-whining-stop-5.html#post5304106

 

Yeah all I know is core 2 duo laptops are pretty fast. They're just way too hot and inefficient for me. I need power and efficiency. So i've decided to not continue screwing with it and sell it. I'm going with a core i3 or i5 acer laptop with Nvidia GT 520M 1GB dedicated graphics. These new core laptop's have the same 4400mAh batteries and are able to pull like 4-6 hours off. Which is damn cool. Plus the new core chips are able to reach extremely low voltages at idle. Plus the damn near base model core i3's even outperform the old generation Core 2 Extreme X9000's while using way less power. These laptop's only run around $500. I'm looking to get a steal on one. I'm very excited! Intel Core generation with a damn good Nvidia gpu. Plus the core chips have the intel integrated gpu within the cpu core itself. So it uses the Intel integrated while browsing etc so save battery and the nvidia when under load. Also the new intel graphics chips integrated blow away even the best last generation GMA chips. Pretty impressive. I'm really excited. Just completely fed up with the core 2 duo and the x3100's. I like to game it up sometimes and I come find out, I need more power and less heat. Intel core chips kick butt. Plus the core i3's are really really amazing processors. They're actually amazing.

 

Same here my T7300, min voltage is 0.85 but the max is 1.25v and 1.2875 for IDA. However I've seen few Pentium Dual Cores with a minimum voltage higher than 0.95v. It is usually in the region of 1v. I think those are faulty CPU that requires higher voltage to run stable. I think the OP T5750 is faulty. I think his T5750 started its life as a T9*00 or a P8*00 CPU, but was downgraded to a T5750 because a portion of the cache L2 was faulty.

From experience I know that RMclock reports the voltages correctly. The latest version of CPUz does a good job too. Realtemp also reports the voltage correctly.

Chris because the voltage values vary from a CPU to another I can't rely on Wikipedia or Intel's website to help you with the mod. Please confirm your min and max voltages as read in RMclock.

By the way, RMclock sometimes does not identify laptop CPUs correctly. Please go to the "Advanced CPU Settings" tab and check that "CPU type selection" is set to "Mobile". The reading of the max and min voltage will be incorrect if it is set to "desktop".

70c at full speed and under stress is a good temperature. Only a few i3 and i5 laptops can run as cool as 70c. Some run around 80c but most reach the 90-100c barrier.

 

Just an update on my research in finding the perfect laptop. I'm very serious in considering a laptop from Eurocom.com which I find to be extremely impressive. the fox 2.0 has a 15.6" led lcd that pulls 1920x1080, 4gb of ddr3, a minimum of a devastating GT 540M, or 555M (I'm getting the 555m ). I'm stilling waiting on monday to call and ask about how well the cooling is on their laptops with the core i7 chip. Since I've heard they run hot (45W TDP). Core i3's/i5's run way cooler. So i'm considering the i7 with the GT 555M and it's less than a grand. I'm gonna get 10% off since my friends a college graduate. I'm going to pay $850 for a 1080p led lcd, core i7 2630QM, Geforce gt 555M 2Gb, 250gb etc etc which is an amazing deal. Nothing out there even comes close to the rates they offer on their unreal laptops for sale. The plus side is they're an enthusiast custom engineered custom laptop building corporation. They report to be the world leader in laptop engineering. Anyway your guy's thoughts on Eurocom laptops would be great! By the way my gateway T5750 laptop is still out of comission . The cpu fan died and i accidentally bought one from Asia on ebay (gotta wait 2-3 weeks ). Later