PLL Pinmod Overclocking Methods and Examples

[nando4]

Originally Posted by moral hazard

, have you tried the "trial and error" method with setfsb?

Can work through the PLLs faster if you can load a minimal XP install by doing a microXP+setfsb install. Some PLLs freeze the system requiring load/cold-boot cycles through the PLLs.

[sean473]

could i have some screenshots to show exactly where in the info you si sandra gives u the pll would be in?

[moral hazard]

here is a screenshot:

[sean473]

Nope still can't find it. Any other software to find PLL?

[moral hazard]

Great news, I have found how to make spdtool and Thaiphoon Burner work for me.

For some reason they work after I press "getfsb" in setfsb.
I guess they find it hard to read my smbus, but after setfsb does it they are fine.

I don't really understand what's happening but I'm happy that I can now flash my ram.

@kaltmond, how do you "hot flash" the ram?
will the system boot with 1 bad stick and 1 good stick?

another question:
If I wan't to increase the cas latency from 5 to 6, should I change "cas latencies supported" from "4,5,6" to "6"?

[aleks nice]

anyone did 200 -> 266 hardware pinmod working ? = multiplier s not locking with 1066 fsb.
i own toshiba x205 s9349
didnt disassembled it yet but google says it has ics pll
so i want to oc it a little
post pls whis details if u had sucess in this

[emike09]

Amazing guide. A good help for me to continue my research overclocking my X9100 to greater heights. Thanks!

[naton]

Originally Posted by aleks nice

anyone did 200 -> 266 hardware pinmod working ? = multiplier s not locking with 1066 fsb.
i own toshiba x205 s9349
didnt disassembled it yet but google says it has ics pll
so i want to oc it a little
post pls whis details if u had sucess in this

I did 133 -> 166, 166 -> 200 and 133 to 200 via pin mod (tested on both socket M and P).
I tested the above with two laptops; one with an ATi Chipset (Socket M), and one with an Intel chipset (Socket P). I didn't had any problems with the ATI based laptop. With the intel one, the CPU was locked at it lowest multiplier.

The 200 -> 266 consists on connecting the two holes A23 and B23 on the CPU socket with a copper wire, or B23 and B24 with a copper wire. I haven't tried this myself (I don't have the right laptop to test), but it is likely to work.



since this is a 33% overclock, increasing the voltage might be necessary.

[nando4]

Originally Posted by naton

The 200 -> 266 consists on connecting the two holes A23 and B23 on the CPU socket with a copper wire, or B23 and B24 with a copper wire. I haven't tried this myself (I don't have the right laptop to test), but it is likely to work.

The PM965 valid FSB shown on the first post, duplicated on the right, suggest that is an invalid BSEL. FSB Capability: These values are determined by the BSEL[2:0] frequency straps. Any unsupported straps will render the GMCH host interface inoperable. 01: (G)MCH capable of up to FSB 800 Mhz 10-11: (G)MCH capable of up to FSB 667 Mhz

PM45 valid FSB

PM965 valid FSB

I think there would need to be a different set of BSEL signals being seen by the chipset/PLL and the CPU for the system to boot *and* potentially work at 1066Mhz FSB as method 3 describes.

[naton]

Originally Posted by nando4

I think there would need to be a different set of BSEL signals being seen by the chipset/PLL and the CPU for the system to boot *and* potentially work at 1066Mhz FSB as method 3 describes.

I know that the PM965 doesn't support 266FSB . The mod will work only with an intel PM975 / x4500