Can anyone help me Vmod a DFI NF4?

[MAValpha]

Angry Games over at DFI-Street posted that the VR circuitry in the NF4 boards has two stages: stage one runs 3.2/4.1V (depending on jumper setting), followed by stage two fine-tuning down to the BIOS selection. IOW, modding a BIOS to allow for more than 4.0V won't accomplish anything, because there won't BE more than 4V available- thanks to TicTac for pointing this out to me.

As some of you know, I'm trying to get some extra volts running through said board, so that I can get some BH5-type stuff to DDR600 and perhaps above. I'd like to feed 4.3-4.5V through the DIMM slots to accomplish that, which requires a massive voltmod. Can anyone help me figure out how to Vmod the first stage of the regulators to supply 4.6V (or even *gasp* 5V) to the second stage?

As always, proper credit will be given to anyone who helps me accomplish this.

[bachus_anonym]

I don't think that 4.3-4.5v would bring anything more to the table than 4v that DFI NF4 allows. Maybe death of sticks, the most
Do not get me wrong, but I've been playing with a lot of BH-5 and giving it up to 4v (on Neo2/DDR Booster and DFI NF4). BH-5 seriously stops scaling up with voltage past 3.6-3.7v. For marginal gains above 270Mhz you will need more and more voltage than when going from 250 to 260. Then, after 3.9v it just stops gaining and only what you get is more errors. So, even If someone comes up with a vmod, I do not think that it will give you stable 300Mhz. We have seen 300 on DFI Nf4 by Mardok, but he cannot stabilize it at that speed. Mardok has probably the best BH-5 I've ever seen in my life and more that 3.8-3.9v does not help him to stabilize it.

[justwOo]

I think it'd actually be more beneficial to find avolt mod, where we could leave our 3.3v rail jumpers, but still get more from a voltmod, so we wouldnt haveto use the 5v jumper which seems to be causing lots of "mishaps"

[persivore]

I think that Malves might have posted a Vdimm mod for the board.
The easiest way IMO would be to use a DDR booster (or Hipro's maximiser when its relased), with the Vdimm jumper set to 3.3v (might work if you just unplug the jumper as well).
If you manage to mod the first stage of the linear regulator, you might find that MOSFETs start to overheat. There is also a possibility that the first linear regulator will drop too much voltage for the mod to be of any use.

[Malves]

I think that Malves might have posted a Vdimm mod for the board.

Nah, I was gonna work it out if someone was willing to be a guinea pig.

Vcore is pin 10 to ground via 50k. For vdimm we could start with that LM358 in the pic.

I think we could take it from here.

[MAValpha]

Here's a really dumb thought...
What if I de-soldered the DIMM sockets from the power circuitry altogether, and wired in a PSU line directly. The RAM will still have the traces for data obviously, but effectively cut the power lines on the board. After this, just use a pot (or two, perhaps) to fine-tune the voltage.

It probably wouldn't be stable enough for the RAM... would it?

Oh, and I should elaborate. I've got tons of RAM, ranging from BH-5 to CH-5 to Micron-B. All three of these are known for responding well to voltage.
OPB got his to ~DDR595, but I think he was limited by the voltage on his mobo. I'm trying to pick up where he left off.

Re: DDR Booster/Maximiser
Is the Booster compatible with the NF4? My NF3 had the OVP kick in at 3.7V and my CPU was limiting me anyway, so I have a Booster already.
Problem: the Booster stops at 4V also. Could I override this by modding my PSU, or the Booster itself?