Socket P undervolting - hardware...

Hello all, it's been a while since I last visited. I'm quite ashamed of myself.

Have just upgraded my socket P Vostro 1700 from a merom-core Celeron 2ghz to a T5800 - still a merom, but a non-crippled one. All is fine, but I'd desperately like to bump the voltage to the CPU down. My problem with software - I don't use Windows very often at all, but flutter from Linux to FreeBSD and back again frequently, so the Windows-based solutions are of little use.

The machine is stable at 2ghz with the CPU running at 1v. Couldn't give a hoot about lower voltages - that's the lowest I can get at using the Windows software, and it's stable for several hours of Prime95.

http://forum.notebookreview.com/showthread.php?t=390696

If I were to do the second socket P pinmod, would it drop all voltages requested by the CPU by 0.2v? It worries me somewhat about hooking up vid[4] to Vcc, as the holes are small and I fear catching vid[5] with the wire.

Could someone tell me whether I can bend vid[5] outwards, removing the risk? If that pin is open circuit by default, then it'll make no odds, right?

I'm technically competent - but years out of date. Whilst I've had no issue in years gone by of messing with bsel or static vid's, I've not done enough digging to understand how they change with recent setups, as the voltage can be changed on-the-fly these days...

Thanks.

 

no not all of them .

Post in my thread (the one showing in your post) the vcores for all your multipliers, and the maximum/default Vcore of your CPU (e.g. 1,2v or 1.25v), and I will advise you if you're better off doing VID3 mod or VID4 mod. I'll also help you with setting RMclock after the mod.

For the wire, use varnish coated wire. You can find this inside electic engines, speakers and so one.

 

PHC for Linux is used to undervolt CPUs in the same way as rmclock/crystalcpuid is in Windows.

 

Thanks, will power it up now and see if I can suss them.