IS7-E vdd mod HERE

[Learn]

hi


i' ve found the vdd mod for my is7-e

don't forgot, WOID WARRANTY

instruction:
you need to remove the R115 resistor (near the cpu fan connector, bottom and left)

use a 20k POT or 50k POT, set to 10k for start (it's very important )

increase resistance, increase voltage

measure the voltage from the mosfet near the agp slot, (below the PLL, right leg)

some measure:
without removed the resistor, r115 = 4.52 K ohms
unsoldered R115 = 10.0 K ohms

anyone can post a pic?
thx

Learn

[HKPolice]

Can anyone please confirm this?

Thanks

[helsyeah]

i happen to have a couple spare is7's layin around and plan on trying it to one of them this evening, ill post results..

[Learn]

hi,

i' ve finished the vagp mod, like the hipro5 way, it work OK, but the monitoring won't see the change.

[HKPolice]

Originally posted by helsyeah
i happen to have a couple spare is7's layin around and plan on trying it to one of them this evening, ill post results..

Any results yet?

[helsyeah]

sorry about the delay, senior projects are picking up and my last few nights have been shot to hell..

i performed the vnorth mod as described by learn, only i initally set my 20k pot to 4.35k which matched the resistance of R115. everything worked flawlessly. as directed, just increasing the resistance increases the vnorth voltage.

I will try to borrow one of my buddies digital cameras so i can post some picks, overall the mod is about the same level of difficulty as the vdimm mod. R115 is a bit of a pain to get to, but it is accessable.

As a side note, Learn, where is the trace you cut to do the vagp mod? i have hunted around quite a bit and have yet to find anyone else who has done the vagp mod on an Is7...

[HKPolice]

Originally posted by helsyeah
sorry about the delay, senior projects are picking up and my last few nights have been shot to hell..

i performed the vnorth mod as described by learn, only i initally set my 20k pot to 4.35k which matched the resistance of R115. everything worked flawlessly. as directed, just increasing the resistance increases the vnorth voltage.

I will try to borrow one of my buddies digital cameras so i can post some picks, overall the mod is about the same level of difficulty as the vdimm mod. R115 is a bit of a pain to get to, but it is accessable.

As a side note, Learn, where is the trace you cut to do the vagp mod? i have hunted around quite a bit and have yet to find anyone else who has done the vagp mod on an Is7...

Awesome, how about OC results? I don't see why Vagp would need to be modded, since max voltage is already 1.65v, what is the point of increasing that unless you're OCing the agp?

[Learn]

Originally posted by helsyeah
sorry about the delay, senior projects are picking up and my last few nights have been shot to hell..

i performed the vnorth mod as described by learn, only i initally set my 20k pot to 4.35k which matched the resistance of R115. everything worked flawlessly. as directed, just increasing the resistance increases the vnorth voltage.

I will try to borrow one of my buddies digital cameras so i can post some picks, overall the mod is about the same level of difficulty as the vdimm mod. R115 is a bit of a pain to get to, but it is accessable.

As a side note, Learn, where is the trace you cut to do the vagp mod? i have hunted around quite a bit and have yet to find anyone else who has done the vagp mod on an Is7...

if you can post a pic on the back of the winbond w83303D, i can help you.
this trace is between the winbond regulator and the southbridge,and at the bottom of the vddr mosfets.
edit: on the back of the motherboard
or follow the trace from this leg
http://www.thelab.gr/images/reviews/...gp/Vagp_12.jpg
http://www.thelab.gr/reviews/ic7mods_en5.php

it's a very small trace, it's not easy to do.

perhaps, it's possible to find an easy way.


vagp help for higher fsb (with the 1:1 mod) and also for higher fsb with the 5:4 and 3:2 ratio, and also for an highest 3D stability.

sorry for my bad english, it limit me... a lot :S

[helsyeah]

actually vagp usually helps a ton on memory stability and increases the maximum stable fsb oc. I have not run across any concrete examples, but i know people run 1.95 or more vagp without issue. I dont know about a "safe" voltage as far as going over 1.65 volts, but i imagine if we're volt-modding our mobos we're not necessarily doing the "safest" thing .

as for results, I havnt seen how far she goes with the 3:2 divider yet, but i will tonight or tomorrow.

The paticular is7-g that i performed the mod on has the issue where it cant boot when going past 280fsb with the 5:4 divide no matter what kind of voltage is thrown at the memory. Increasing the vnorth did not resolve that issue. I experenced the same thing with my IC7-Max3 as well. No amount of vnorth or vagp would let me use the 5:4 divider over 280 mhz.

I also have an is7-e that does not have any problem doing over 292fsb with the 5:4 divider with the memory that i have so its not my memory having problems with the raw speed. Actually, the memory is capable of 245mhz with 3.2v and 2-5-2-2 timings, at least at 1:1 mode. There is more stress on the memory with the 5:4 divider however. I currently am stuck at 235ish with the 5:4 divider, but im hoping that the vagp mod lets me get beyond that. Especially since i know my proc is capable of whats in my sig

[helsyeah]

learn, ill definately track down a digital camera then this weekend and get some pics posted... maybe if we can get some other mod-gurus in here we might find a better way to perform the vagp mod

[Learn]

ok, good idea

i have a problem over 280 5:4 with the pat enabled with my is7e, no problem without the pat.
bios problem ? (abit)

[HKPolice]

Thanks for the info guys. I tried pushing my Vagp to 1.65v (1.6v actual) and it helped prolong Prime's stability by a few minutes @ 300FSB on a 2.4C, 3:2 ratio, 2x 512MB Corsair PC3500 BH-5 chips, 3.01-3.09vdimm, 1.54VTT.

I believe my main problem is the PSU, the default rails are too low, 3.18-3.22v, 4.89v, and 11.68-11.80v.

Gonna do mod on all 3 rails when my Antec TP550 comes.

I'm concerned about my vcore, its 1.67v idle, but varies from 1.60-1.62v underload, is that normal? A full 0.07v sway from load to idle.
Vdimm idles at 3.09v, but sways from 3.01 to 3.04v underload... Is this all normal? Vcore is set to 1.7v in BIOS, but only reads max of 1.67v...

295FSB is completely stable so far, been doing 2 instances of prime95 for an hour or so.

[HKPolice]

Is there a "droop mod" for IS7? Seems like it helps the Asus boards a lot, the Vcore only varies by like 0.04v after the mod

[helsyeah]

i havnt run across a specific droop mod yet. there is a mod just posted today for vcore for the is7. However it is unclear to me why the is7 would need a vcore mod due to the insane vcores we have available... i would be interested in seeing if this specific mod did happen to deal with fairly drastic drop in vcore during load.

[HKPolice]

Originally posted by helsyeah
i havnt run across a specific droop mod yet. there is a mod just posted today for vcore for the is7. However it is unclear to me why the is7 would need a vcore mod due to the insane vcores we have available... i would be interested in seeing if this specific mod did happen to deal with fairly drastic drop in vcore during load.

You get the same kind of variance in vcore and vdimm while underload as well?

[helsyeah]

Originally posted by HKPolice
Vdimm idles at 3.09v, but sways from 3.01 to 3.04v underload...

this is not normal, my vdimm maybe varies from 3.08 down to 3.07 at full load... so it would definately seem you have a weak PS... the true power should help a bit

my vcore does drop at least .05 though with full load, and more when i crank the proc over 3.6ghz

[HKPolice]

Originally posted by helsyeah
this is not normal, my vdimm maybe varies from 3.08 down to 3.07 at full load... so it would definately seem you have a weak PS... the true power should help a bit

my vcore does drop at least .05 though with full load, and more when i crank the proc over 3.6ghz

I'm currently using an Antec TP430, but i'm selling it to my friend once my TP550 arrives. Problem is that the 3.3v rail varies from 3.18 to 3.22, what PSU do you use and how much does yours vary?

I did the Vdimm and VTT mods posted on AkibaPC's page here: http://www.akiba-pc.com/article.php?38.0 except that I used a 10K pot, and I think my resistance is higher than 1K.. but it works, when I set vdimm to 2.5v in bios, it drops down to 2.75v actual.

Running 295FSB @ 1.55vcore load, stable so far

[HKPolice]

One more thing, I see you're running an IC7, with 3:2 divider with over 300FSB, do you have FULL PAT enabled?? I am unable to select full PAT with 3:2 divider, no matter what FSB. Can't get full PAT beyond around 260FSB with 5:4 divider... and I have the same RAM as you

[helsyeah]

im actually using an antec smartpower 400 and it holds up pretty well... i did do the mod to get my 3.3 rail up to 3.5v. The only time i see my rails start to drop at all is when i go over 3.65ish ghz on the proc, then my 3.3 rail drops from 3.5 down to 3.4ish other than that, the 12v and 5 v rails stay within .02v of their idle values of 11.98v and 5.02v

[HKPolice]

Originally posted by helsyeah
im actually using an antec smartpower 400 and it holds up pretty well... i did do the mod to get my 3.3 rail up to 3.5v. The only time i see my rails start to drop at all is when i go over 3.65ish ghz on the proc, then my 3.3 rail drops from 3.5 down to 3.4ish other than that, the 12v and 5 v rails stay within .02v of their idle values of 11.98v and 5.02v

Hmm, either my winbond monitor chip is old or inaccurate, or something.... I did monitor my 12v line with a multimeter before and it didn't vary more than 0.03v when booting up and prime. It was 12.14 - 12.17v, even tho winbond hardware doctor says 11.74v idle.

[helsyeah]

Originally posted by HKPolice
One more thing, I see you're running an IC7, with 3:2 divider with over 300FSB, do you have FULL PAT enabled?? I am unable to select full PAT with 3:2 divider, no matter what FSB. Can't get full PAT beyond around 260FSB with 5:4 divider... and I have the same RAM as you

actually im running my is7-e atm, the ic7 was an experiment to see how far i could push my proc

as to PAT, it was fully enabled when i was running @ 308fsb and i had pat all the way up to 275 using the 5:4 divider, but at 280fsb i hit the invisible wall where i could not post with the 5:4 divider.

on my is7 I actually have my GAT settings set to turbo, normal, auto, disabled, disabled. memtest and cpuz do not see pat as enabled, but i get over 150 mbps more unbuffered bandwidth on my memory by just having turbo enabled in GAT vs. having it set to auto.

[HKPolice]

Originally posted by helsyeah
actually im running my is7-e atm, the ic7 was an experiment to see how far i could push my proc

as to PAT, it was fully enabled when i was running @ 308fsb and i had pat all the way up to 275 using the 5:4 divider, but at 280fsb i hit the invisible wall where i could not post with the 5:4 divider.

on my is7 I actually have my GAT settings set to turbo, normal, auto, disabled, disabled. memtest and cpuz do not see pat as enabled, but i get over 150 mbps more unbuffered bandwidth on my memory by just having turbo enabled in GAT vs. having it set to auto.

So on your IS7-E PAT was fully enabled upto 308FSB with 3:2 divider? But that's not possible, cuz if you set the NB strap to 800PSB, it only does partial PAT unless you select F1 setting.

If you select PSB667, FULL PAT is enabled all the time, even if you select auto for GAT, but it only allows 1:1 and 5:4 dividers.

CPU-Z shows PAT as disabled if it's partially enabled. You need CTIAW to truly see the state of your PAT.

[helsyeah]

oops, sorry about the confusion, i meant that on my ic7 pat was fully enabled @ 308fsb...

[HKPolice]

Originally posted by helsyeah
oops, sorry about the confusion, i meant that on my ic7 pat was fully enabled @ 308fsb...

308FSB with 3:2 divider? :o

[helsyeah]

lol, yes with the 3:2 divider