NF7 rev2 VTT mod need it so bad...

[uwackme]

So the feedback resistor is R750. It is located between DIMM1 and DIMM2 near the winbond chip end. It appears to be a 120ohm resistor, but my eyes even with the magnifier arent that good. It measures 120. The pad on R750 nearest the DIMM levers is Vtt, the end nearer the middle of the DIMM is attached to the winbond chip pin2. So this is the feedback path. There is a missing R751 that goes between Pin2 (feedback) and GND, but that would be the opposite of what we want.

Ultimate change would be to replace R750 with a pot, but that would be VERY hard based on location. Our mod so far...adding resistence into the feedback path....cut trace, reconnect through small resistence. Even my EC3001 soldering iron cant get in there without harming the DIMM sockets.

So at least we have a place on the top of the board to measure other than the DIMM. R750 end near levers.

[enduracell]

Originally posted by Malves

Start trimmer at max? No. According to the IC specs there's no feedback divider. So, we have to "create" one, to trick the controller in feeding more VBT (Vbt=Vtt). You start at 0 omhs and raise the resistance to get more volts out of the controller. [/B]

Yes u are right Malve. This is wrong when u do the mod with inline sense trimmer, but not if u do it with one point on the Vbt sense and the other to the ground. (like the moboFAQ Vdimm way for example) Of course I prefer ur way, I just thought be more practical and safer to first just lower the V of the trace with grounding to see how reacts.
Congratulations guys I think U find it @last
Edit: To see if the cutting trace gives the same volts with the Vbt mesaure the legs D1 of the FDS9630A on the side of dimm3.
I have mesaure it many times and pin 2 of the Winbond and the 2 legs of the FDS fet mach in all 4 selections of BIOS (2.9 gives 3.08Vdimm that gives 1.54Vbt @least in my mobo)

[Endre]

Just did the vdimm mod and it works great.
I recommend this mod to others too!

And I follow the progress of the vtt-mod with excitement. Really hope it helps!

[bachus_anonym]

gentelman, we got ourselves a WINNER !!!!!!!!!!!!!!!!!!

1.6 vtt

pics to follow. !!!!!!!!!

[enduracell]

This is great news!!!!!!!!
bachus_anonym u take all the glory

[uwackme]

Easy place to measure Vtt... there are a pair of 56ohm resistors for each channel. Channel A (DIMM1-2) is in area between DIMM2 and DIMM3 between a couple resistor packs (the multipin black resistor banks all over the place). Channel B's is on outside of DIMM3 about 1" from lever.

So Vtt at DIMM socket pin is result of Vtt passing through a 56ohm resistor.

[bachus_anonym]

i'm so excited !!!! guys, u r geniuses !!!









if only it will go so smooth on my good board :-)

i need to calm down coz my hand are shaky now :-)




Guys, take on it and post results...

good luck

[Dani]

Gonna do that tomorrow

[bachus_anonym]

DFI Infinity, here we come !!! Hopefully

[Bugs]

very nice, good job

[daos]

who figured this mod out? i amv ery curious who finally figured this board out.

we need to give credit where it is due...

[Dani]

How much is default Vdimm, 2.6V? (started mobo without CPU, Memorys and VGA)
I just measured my VBT and it showed 1.45V...not VBT modded

[bachus_anonym]

default is 2.6v... but on my test-mobo even when i clear CMOS and power it up it still shows 2.96vdimm and 1.6vtt... i guess i would need to get to BIOS and "save and exit" with default settings to get right readings at 2.6v...
which i can't do coz it's "dead"

[Dani]

Hmm, I measured before I cleared CMOS VBT=1.45V VDIMM=3.45V, after I cleared CMOS it was still VBT=1.45V VDIMM=3.45...
I have write down that 2.6V should be 3.14V, weird...

[bachus_anonym]

ok, it's time to take on my precious good board :-) hopefully it will be as easy as test one...

[uwackme]

Mine at default 2.6V has Vtt of 1.37V, at 2.9V it's 1.54V... pre mod.

Do VDimm mod, VDimm sails off and Vtt stays 1.54V

Interesting thing, before I do the mod, I using Kingston PC2100 value ram as ginuea pig. At stock 2.8V setting 200Mhz 11,3,3,2.5 CPUINT enb. If I turn up the VMod so Vdimm is 3.2V... errors it's ASS off at any speed. I have a feeling Vtt mod will wake this stuff up HARDCORE. It's very sensitive to the Vdimm Vtt disparity.

Ultimately, we'll just come up with a fixed resistor value that sets Vtt correctly for a higher Vdimm range, most people with CH5/BH5/BH6 chips wont need to bother with a pot. Just cut etch, and bridge with a small resistence...BUT... you will be tied to keeping VDimm in higher range, otherwise with Vdimm low you end up with the exact OPPOSITE situation, with Vtt too high vs Vdimm and unhappy rams.

The key here is that the DDR data bus is a transmission line, and the terminating resistors.."PULL MIDDLE" the lines. IE: they pull each signal to a point 1/2 way between a "0" and a "1" which for Vdimm 3.0V would be 1.5V. If Vtt is offset, then it is easier to make a "0" on the bus than it is to make a "1"... conversely it is HARDER to make a "1" and here at high speeds is where the ERROR comes from.

Once Vtt is dead center, you truly will find the REAL maximum speed your rams will operate at....at any given Vdimm. If Vtt is offset, you wont see the true maximum because "0"'s are easier to make than "1"'s....I hope that helps people understand WHY we need this.

[enduracell]

Originally posted by Dani
Hmm, I measured before I cleared CMOS VBT=1.45V VDIMM=3.45V, after I cleared CMOS it was still VBT=1.45V VDIMM=3.45...
I have write down that 2.6V should be 3.14V, weird...

When U clear CMOS the fail safe settings always sets the Vdimm at 2.8 in BIOS. The default Vdimm in this value should be around 2.9v (if I remember correctly) and the Vbt is 1.45v. If u done only the Vdimm mod the Vbt isn't effected and repents only from BIOS selection. So ur Vbt readings is correct, and ur Vdimm is wrong if u haven't done the mod (prompably u are measuring the 3.3 rail upon the fet.)

[Dani]

Ok, Im not sure what I measured but my +3.3 is around +3.85V
I have "normal" Vdimm mod...

[enduracell]

Ok I got it then
U need definitely 1.7v Vbt if u are indeed around 3.4 Vdimm

[uwackme]

The mod works correctly.

My Valueram doesn't like Vdimm over 2.9ish, but now Vtt is correct.

3.2Vdimm and 1.6Vtt

Ill go test the CH-5 ram now.

[STEvil]

Good going guys

[muzz]

WTG Fellas!!

I knew the guys at Xtreme would somehow get together on this and figure it out, dead boards help too it seems (B.A... )

Excellent, and some spare 2100 for guinea pig, heheheh

Great job guys, need data....... need data..........heheh
:thumbsup:

OT: Marlborough and Andover huh?
Hmmmmm I live in Easton, I was in BOTH Marlborough and Andover TODAY!!

Go Sox!!!

[saaya]

very interesting

so the dfi nf2 just stop increasing vtt after 1.45v as well?

[STEvil]

good question saaya.

[muzz]

I THOUGHT I saw a post where someone else said the VTT on the DFI board tracked correctly(up to 1.65/3.3v or something?).
I wasn't aware it was an issue with them.