DFI Blood Iron vcore mod

[Misko]

Because BI make vcore from vid + value in bios, max is ~1.9V for 45nm CPU, so I decide to vcore mod it, I think this is the right spot can someone approve this one? 10x in advance

[akin]

i think that's true..

this board is using the same IC PWM Controller with DFI DK P35 T2RS

[Misko]

i think that's true..

this board is using the same IC PWM Controller with DFI DK P35 T2RS

10x m8, I'll play these days with vcore and alternate soldering spots I'll post here results (maybe I try to figure out vdrop also)

[akin]

ouw nice,
we Will wait to your result...

[celemine1Gig]

Hi Misko,

you are right, the pin you marked with the blue line is the FB pin. Measure resistance between this pin and ground with the system completely disconnected from power.

And about the alternate soldering points, have a look at this:



Set your multimeter to continuity testing mode and try the green marked points. I'm not 100% sure, but judging by the visible traces, these points could be directly connected to the FB-pin.

[c1rcu1tburn3r]

I'm interested too...the fb pin was spotted...now we can work on the soldering point and the trimmer value
I don't have the mobo under my hands now, so i can't help you

[c1rcu1tburn3r]

Have you found somethings about m8?

[c1rcu1tburn3r]

Let's go... i was in trouble, but with ma crazy mind i risked it and... it work perfectly!

[Misko]

Have you found somethings about m8?

Sorry I was a little busy (start playing with commando) soon I'll do the same thing what happened with vdrop since u plug in vcore mod?

[c1rcu1tburn3r]

Sorry I was a little busy (start playing with commando) soon I'll do the same thing what happened with vdrop since u plug in vcore mod?

Nothing different...if you put around 1.98 you'll go under load to 1.965 and similar...maybe a little bit better than without vmod...anyway i've also vdrop to do, but i don't post it now couse i'm not sure

[Misko]

Nothing different...if you put around 1.98 you'll go under load to 1.965 and similar...maybe a little bit better than without vmod...anyway i've also vdrop to do, but i don't post it now couse i'm not sure

Why not m8, post it maybe someone good with electronics will see and correct/approve mod... I'll search for one too

[celemine1Gig]

Nothing different...if you put around 1.98 you'll go under load to 1.965 and similar...maybe a little bit better than without vmod...anyway i've also vdrop to do, but i don't post it now couse i'm not sure

I doubt that the VCore-Mod improved Vdroop in any way. In reality higher vcore should actually result in a higher droop voltage, because higher vcore means higher current. That's how Vdroop works.

[c1rcu1tburn3r]

Anyway for me it's a little bit reduced, dunno why but...
For vdrop you've to bypass the "upper right" resistance with green spot, a vr but i haven't tried how many kohm...I think that can work

[celemine1Gig]

Just had a look at the datasheet. You are actually most likely right that this vcore mod decreases vdroop. The reason is the close connection of the idroop pin (for vdroop; right next to the fb pin) to the fb pin. When doing the vcore mod, i.e. influencing Rfb, you are influencing Rll (the loadline resistance) with it. It's on page 34 of the datasheet.

The simplest and cleanest way of completely getting rid of Vdroop on this board is to disconnect the Idroop pin from the circuit. I don't recommend doing this, though. You'd most likely have horrific voltage overshoot, which could potentially kill your CPU. When completely eleiminating VDroop you'll have to take care of enough decoupling capacitance first.

[Misko]

c1rcu1tburn3r maybe u can try this one: set vcore in bios 1.3~1.5V than with trimmer set to 1.95 in win, just curious in that way how vdrop will react. (I presume that u set 1.9 in BIOS and than just a little touch the trimmer in win till 1.95~2V) I'll soon be in game with DFI