Droop Mod or similar for NF7-S

[-Sonic-]

Does any mod exist that can stop the vCore fluctuating so wildly on the NF7-S V2?

At 1.9V BIOS, it shifts from 1.81V to 1.87 :eek:

[sislianli]

Does any mod exist that can stop the vCore fluctuating so wildly on the NF7-S V2?

At 1.9V BIOS, it shifts from 1.81V to 1.87 :eek:

get larger capacitors on the motherboard.
replace the large ones.

[-Sonic-]

Isnt there a simpler way to stablize the voltage? Allready tried 2 different PSU's and running nothing but a old GeForce2 with 1 HDD

[bobbobson]

The most simple way would be to replace your caps in your voltage regulation circut. It isn't very difficult.

- Michael.

[persivore]

Replacing caps won't help much, and it isn't worth the effort/cost for the minimal gain you will probably get IMO. There is a droop mod which needs to be done with a pencil posted here (http://xtremesystems.org/forums/showthread.php?t=48006&highlight=NF7). You might be able to use a variable resistor instead of a pencil. If you can wait until Monday, I will do the mod myself on my NF7 and find the value variable resistor that you need :)

[-Sonic-]

persivore, thats perfect!

I have one concern however, it seems people who have done this mod cant seem to be able to select more than 1.85V in the BIOS, how can i overcome this?

[-Sonic-]

hmmm well ive done the mod and im pleased to say it works :D

vCore drops 0.01V under load if that at all!

A NOTE: you have to pencil ALL 3 resistors for this mod to work! DONT pencil too much on resistor 3 otherwise the vcore will be jammed open (as i found out :eek: )

If you put too much graphite on resistor 3, you will not be able to select vCore greater than 1.85V in BIOS, trial and error really.

Previous to this mod, anything above 1.9V selection in BIOS = no boot, now i can select up to 2.225V with no issues and VERY little (0.01V) fluctuation if ANY.

[LorD_Kvadd]

Would you mind measuring the resistance on those resistors?

If I ressurect my dead nf7-s then I would love to do that mod but with resistors instead of graphite :P

[-Sonic-]

LOL TBH mate im too scared to point a multi-meter anywhere near something that small

[LorD_Kvadd]

okey? In fear of what?

You don't have to have the board running to measure resistance you know ;)

Please, it would help alot :)

[-Sonic-]

That part about board not having to run means i'll do it :)

Can you tel me exactly what points you need me to measure and how i should go about doing it?

[Monster]

That part about board not having to run means i'll do it :)
Can you tell me exactly what points you need me to measure and how i should go about doing it?
When will you measure these 3 resistances ? Could you please measure ?
It's very important for me (and not only for me ;)) !
Thanx in advance ! :)

[HaMMuS]

Hey -Sonic-,

The three resistors you need to measure are the ones you pencilled for the mod, shown in the pic below.

To measure the resistance, set your MM to ohms, and put a probe at either end of one resistor (may be fiddly), note the reading, and repeat for the other two resistors.

Hope This helps

Mikey

[Monster]

Hey -Sonic-,

The three resistors you need to measure are the ones you pencilled for the mod, shown in the pic below.

To measure the resistance, set your MM to ohms, and put a probe at either end of one resistor (may be fiddly), note the reading, and repeat for the other two resistors.

Hope This helps

Mikey

As I see you Vdrooped your mobo!
Could you please measure these 3 resistances too ? ;)

[HaMMuS]

As I see you Vdrooped your mobo!

Errrrmmmm, actually it was more of a pre-emptive vdroop, havent got round to doing the mod yet :slapass:

[-Sonic-]

I'll do it this weekend :)

Bit tied up with work atm :stick:

[Monster]

I'll do it this weekend :)
That's great ! Waiting for results (resistances) ! ;)

[-Sonic-]

Sorry about the delay, will do it tomorrow.

Got too much on atm :(

[Monster]

Oh! :(

[HaMMuS]

Any news on those resistor values -Sonic-? :stick: :)

[-Sonic-]

Apologies all, aint had time to get around to it yet.

I'll be pretty busy for the next few weeks but i'll try my best to make some time. Think some of the graphite is coming loose from the fan blowing over the mofsets :-\

[jumanji969]

My vcore hardly fluctuates at all, just overvolts slightly.

[persivore]

I think I've got a new Vdroop mod for the NF7. It involves replacing the 3 resistors which sense the currect from each phase of the Vcore regulator. Its based on the Vdroop mod for the DFI infinity. I'll do the mod and test it later, and post it if it works as expected :)

[persivore]

Sorry it took me so long to post the mod, but here it is.

The 3 resistors circled are the current sense resistors for each phase of the Vcore regulator. They have an origional value of about 3k. I've found that replacing them with 4.7k resistors will cause the Vcore to increase by 0.004v, from 1.895v to 1.899v under prime95 load. This was tested using a duron @ 1350mhz, so the voltage change with a higher load will probably be slightly less.

Note: Do not power up the board with the currect sense resistors removed as this could damage the board and possibly the CPU and powersupply.

[HaMMuS]

So your saying that with your new mod, under load the Vcore actually increases rather than droops, persivore?

[persivore]

So your saying that with your new mod, under load the Vcore actually increases rather than droops, persivore?
Yea, before the mod was done, the Vcore would drop by 0.002v under load, and now it increases by 0.004v :) The same thing happens when the DFI Nforce2 series of boards are Vdroop modded.
If more current is needed by the CPU (like when running high voltage/high clock speeds) the droop/voltage increase will probably be close to zero.

[giorgioprimo]

Yea, before the mod was done, the Vcore would drop by 0.002v under load, and now it increases by 0.004v :) The same thing happens when the DFI Nforce2 series of boards are Vdroop modded.
If more current is needed by the CPU (like when running high voltage/high clock speeds) the droop/voltage increase will probably be close to zero.

may you post a pictures of your vmod?

thanks

[persivore]

may you post a pictures of your vmod?

thanks
If I can get a digital camera from somewhere, I'll try to get pics, but it just looks exactly like an unmodded board, but with different numbers on top of the 3 current sense resistors ;)

[Monster]

Sorry it took me so long to post the mod, but here it is.

The 3 resistors circled are the current sense resistors for each phase of the Vcore regulator. They have an origional value of about 3k. I've found that replacing them with 4.7k resistors will cause the Vcore to increase by 0.004v, from 1.895v to 1.899v under prime95 load. This was tested using a duron @ 1350mhz, so the voltage change with a higher load will probably be slightly less.

Note: Do not power up the board with the currect sense resistors removed as this could damage the board and possibly the CPU and powersupply.
Didn't you change other 3 resistors are circled (in the left part of attached pic)? If I got it, circled resistors in left part of the pic are nominal (native), but other 2 resistors in right part of pic changed from 3K to 4.7K?
And the point of your voltmod is changing/replacing 2 resistors (right part of the pic)? Am I right? :)

[persivore]

Didn't you change other 3 resistors are circled (in the left part of attached pic)? If I got it, circled resistors in left part of the pic are nominal (native), but other 2 resistors in right part of pic changed from 3K to 4.7K?
And the point of your voltmod is changing/replacing 2 resistors (right part of the pic)? Am I right? :)
You are correct, but there are 3 3k resistors which need replacing. Two run vertically on the picture, and the last one is above the 2 vertical ones ;). These 3 are the only ones which need replacing.
The other 3 resistors on the left (numbered 1, 2 and 3) have not been changed.

The origional 'pencil droop mod' would probably be easier than this mod, but doing it this way will still give you control of votlages over 1.85v if you use the wrong amount of pencil :)

[giorgioprimo]

You are correct, but there are 3 3k resistors which need replacing. Two run vertically on the picture, and the last one is above the 2 vertical ones ;). These 3 are the only ones which need replacing.
The other 3 resistors on the left (numbered 1, 2 and 3) have not been changed.

The origional 'pencil droop mod' would probably be easier than this mod, but doing it this way will still give you control of votlages over 1.85v if you use the wrong amount of pencil :)


so i think thsat is possible to solder other resistor (talking about normal ones, i CANNOT FIND AND SLODER smd RESISTOR... in parallel with the 3 mentioned just to obtein a final value of 4,7k right?

[persivore]

so i think thsat is possible to solder other resistor (talking about normal ones, i CANNOT FIND AND SLODER smd RESISTOR... in parallel with the 3 mentioned just to obtein a final value of 4,7k right?
To do the new mod, you need to increase the resistance of the 3 resistors from 3k to 4.7k, and soldering resistors in parallel will decrease the resistance, and will probably make the droop worse.

If you were talking about the origional pencil mod, it would be possible to do the mod by soldering resistors in parallel with the existing resistors (thats all that the pencil is doing afterall)

[giorgioprimo]

To do the new mod, you need to increase the resistance of the 3 resistors from 3k to 4.7k, and soldering resistors in parallel will decrease the resistance, and will probably make the droop worse.

If you were talking about the origional pencil mod, it would be possible to do the mod by soldering resistors in parallel with the existing resistors (thats all that the pencil is doing afterall)

thanks...so the only chance is to de-solder the old resistor and replace them.... :( I dn't fill myself able to do that....

[persivore]

thanks...so the only chance is to de-solder the old resistor and replace them.... :( I dn't fill myself able to do that....
I'll look into other ways to do the mod, but I think that removing the resistors will be the only way. If you don't want to remove resistors, the best way to mod the board will be to do the 'pencil droop mod'. You could use variable resistors instead of penciling if you want to make the mod more permanent.

[giorgioprimo]

I'll look into other ways to do the mod, but I think that removing the resistors will be the only way. If you don't want to remove resistors, the best way to mod the board will be to do the 'pencil droop mod'. You could use variable resistors instead of penciling if you want to make the mod more permanent.

You looks like quite confidential with the vmod..I've abother question: do yuo have any suggestion to eliminate the vcore protection and get the vcore go over 2,1 volt? (real mesured with a tester, not from the bios...)..
Thanks

[persivore]

You looks like quite confidential with the vmod..I've abother question: do yuo have any suggestion to eliminate the vcore protection and get the vcore go over 2,1 volt? (real mesured with a tester, not from the bios...)..
Thanks
If the Vcore is limited to 2.1v becasue of OVP, there is a (simple) mod which can be done to raise the OVP level. I'll try to find a link to it :)

[gsan]

i'm waiting for it :)

[HaMMuS]

@Persivore:

Any tips on whereabouts i can find some 4.7K resistors on the board? I have a couple of dead NF7s lying around that i can scrounge from and theres a lot of resistors on em! :D

[Rookie]

i've been searching everywhere for a vdroop mod, been to many forums, i was told 2 things.

1. that it doesnt exist
2. its a pencil mod(but noone could come up with a pick, and noone had done it, they just heard about it)>

instead of the pencil mod i'd like to do persivore's mod with resistors. please can you post specs on the resistor and part#?

also a detailed pick of how you did it. this is great :) my vcore with an 400w fortron droops .3 - .4 volts.

[Batt[o]]

I have quite the same problem, 1.875 by bios and it drops to 1.808-1.825 :S
Can you Explain it quite better, and add some pics? please I'd like to try it too.

[persivore]

I don't have a digicam, but the board looks just like an unmodded board. The only way that you would see a difference would be to compare the numbers on the 3 resistors to those on an unmodded board. If somone can get a close up picture of the Vcore regulator, I'll mark the resistors on a picture :).

The resistors I used were 3 4.7k SMD resistors which I 'borrowed' from a dead MSI P4 board, so I don't know what size code they were. 4.7k SMD resistors should have 472 written on the top of them. If you use resistors taken from another board, you need to check that they are the correct value because if their value is too high, or are damaged it will probably kill the board, PSU and CPU.

[Batt[o]]

I'll take a close up of my board...please mark wich resistors should be changed...

http://img191.imageshack.us/img191/5129/nf7scloseup0my.th.jpg (http://img191.imageshack.us/my.php?image=nf7scloseup0my.jpg)

And I need to look for 3x4,7k resistors...and a good hand to sold them :P

[persivore]

I've circeld the 3 resistors which need to be replaced. It would also be possible to replace them with 3 5/10k variable resistors set to 4.7k, but it might not be as tidy as directly replacing the resistors

[Batt[o]]

Sorry I just uplodeaded the wrong pic...

http://www.xtremesystems.org/forums/attachment.php?attachmentid=36846&stc=1

[persivore]

yea, thats right (the ones inside the smaller circle)

[Batt[o]]

Okey now i know what i have to change...
need some tools and some 4,7k capacitors...i'll see were i can get them.

[giorgioprimo]

If the Vcore is limited to 2.1v becasue of OVP, there is a (simple) mod which can be done to raise the OVP level. I'll try to find a link to it :)

Any news about that ?


Thanks

[KODE]

I just did persivore´s mod, works like a charm! :toast:

[Rookie]

']Okey now i know what i have to change...
need some tools and some 4,7k capacitors...i'll see were i can get them.


i'm an EXTREME NEWB, please can you post pics when you've done the mod? i have all the tools necessary. i know persivores directions are clear for the experienced and intelligent, i somewhat lack both.

pics would be greatfully appreciated.

:)
eddie

[<Wouter>]

Nice mod persivore. I'll preform it on my nf7. :)

[Monster]

I've circeld the 3 resistors which need to be replaced. It would also be possible to replace them with 3 5/10k variable resistors set to 4.7k, but it might not be as tidy as directly replacing the resistors
Could U please write your Vcore params after this mod:
1) Vcore is selected in BIOS (for example my Vcore is selected in BIOS = 1.675V).
2) Vcore fluctuation in idle and in load (my values are 1.632V & 1.584-1.568V).
Thanx for your simple mod! ;) :toast:

[Monster]

And what power of resistors are necessary (0.125, 0.25... ???)?

[Monster]

I've done this mod. I changed 3 resistors to 3 others (4.7KOhm) and my new Vcore params are:
1) Vcore is selected in BIOS = 1.65V.
2) Vcore fluctuation in idle and in load are 1.616-1.632V & 1.600-1.584V.
Before voltmodding I though that Vcore would be constant! :(
Now I think that resistances of these ones should be more than 4.7KOhm (5-6KOhm?).

[persivore]

I was getting some droop when I tried running a 2500 barton in the modded NF7, so it might be worth trying some larger value resistors. I would suggest trying 5.6k or 6.8k resistors, but the value that you need would depend on how heavily you are loading the Vcore regulator.

[Monster]

I was getting some droop when I tried running a 2500 barton in the modded NF7, so it might be worth trying some larger value resistors. I would suggest trying 5.6k or 6.8k resistors, but the value that you need would depend on how heavily you are loading the Vcore regulator.
Could U please try 5.6KOhm & 6.8KOhm resistors and write here your droop results ?
BTW my CPU is 2600+ (166x11.5, 1.50V) overclocked to 3300+ (225x10.0, 1.60V in real, but in BIOS I set 1.65V).
P.S. The man who resoldered my mobo yesterday told that he think these 3 resistors are soldered sequentially! Is he right?

[persivore]

Could U please try 5.6KOhm & 6.8KOhm resistors and write here your droop results ?
BTW my CPU is 2600+ (166x11.5, 1.50V) overclocked to 3300+ (225x10.0, 1.60V in real, but in BIOS I set 1.65V).
P.S. The man who resoldered my mobo yesterday told that he think these 3
resistors are soldered sequentially! Is he right?
I can't try anything out on my NF7 atm becasue its currently mounted in a case. When I take it out again, I'll try replacing the resistors with higher values :)

Each of the resistors are connected between Vcore 'phase' and a current sense pin on the Vcore regulator. Each resistor relates to one of the 3 phases of the Vcore regulator. Vcore 'phase' is the junction between the Vcore MOSFET and the inductor. The Vcore 'phase' point is common to each resistor, but the current sensing pin on the Vcore regulator is different for each resistor.

[Monster]

I can't try anything out on my NF7 atm becasue its currently mounted in a case. When I take it out again, I'll try replacing the resistors with higher values :)

Each of the resistors are connected between Vcore 'phase' and a current sense pin on the Vcore regulator. Each resistor relates to one of the 3 phases of the Vcore regulator. Vcore 'phase' is the junction between the Vcore MOSFET and the inductor. The Vcore 'phase' point is common to each resistor, but the current sensing pin on the Vcore regulator is different for each resistor.

I've got nothing about sense pins and etc...
Just write here please when you'll resolder these ones! ;) OK?

[persivore]

I'll probably try replacing the fixed resistors with variable resistors next weekend so that I can try differnet values :)

[Monster]

Is it really to do that Vcore(by BIOS) = Vcore (in real) = Vcore (in idle) = Vcore (in full load) ?!
As FULL LOAD I mean S&M utility!

[Monster]

I'll probably try replacing the fixed resistors with variable resistors next weekend so that I can try differnet values :)
WOW! That's GREAT! :toast:

[Monster]

I'm look into datasheet of HIP6301.
Typically, IC HIP6301 is working in 4 phase mode.
I think that problem with non stable Vcore under heavy load related to non existed 4th gate.
If you make upper resistance in feed-back loop by current it's possible to have two problem:
1) IC can trigger into "overload by current" because this resistors related also to detection of an over-current condition
2) Output MOSFET can be damaged because IC doesn't know about its real current.
Another possible problem is that under heavy load temperature of motherboard can be too high and internal CLK generator of IC changes own frequency and output power of MOSFET gate can change...

[persivore]

I'm look into datasheet of HIP6301.
Typically, IC HIP6301 is working in 4 phase mode.
I think that problem with non stable Vcore under heavy load related to non existed 4th gate.
If you make upper resistance in feed-back loop by current it's possible to have two problem:
1) IC can trigger into "overload by current" because this resistors related also to detection of an over-current condition
2) Output MOSFET can be damaged because IC doesn't know about its real current.
Another possible problem is that under heavy load temperature of motherboard can be too high and internal CLK generator of IC changes own frequency and output power of MOSFET gate can change...
The 4th phase of the Vcore regulator is disabled by pulling its 4th PWM output high/low (I can't remember which from memory).
The overcurrent threshold should be raised by increasing the value of the Isense resistors, becasue you are making the regulator think that the CPU is drawing less current, so it gives less Vdroop.
The Vcore regulator on the NF7 is capable of supplying up to 2.3v without any mods, so the Vcore MOSFETs should be able to cope with the load. The droop function of the Vcore regulator is designed to allow motherboard maufacturers to use cheaper capacitors/fewer capacitors in the Vcore supply. The MOSFETs on the board should be able to cope with the load, but the capacitors on the board might have a slightly shortened life (although they shouldn't be effected too much).

[Monster]

Just wait for the news about this ! ;)

[giorgioprimo]

The overcurrent threshold should be raised by increasing the value of the Isense resistors, becasue you are making the regulator think that the CPU is drawing less current, so it gives less Vdroop.

If you will find the solution, you are the man..... :stick:

The MOSFETs on the board should be able to cope with the load, but the capacitors on the board might have a slightly shortened life (although they shouldn't be effected too much).

I don't care....BTW I should also change the capacitors with bigger\beter ones...

[persivore]

BTW I should also change the capacitors with bigger\beter ones...
You shouldn't need to change the capacitors, unless you are running a vary high Vcore (2.1v+) with a lot of load on the Vcore regulator :)

Edit: I might be able to re-do the mod tonight with variable resistors. I've just got my powerstream back from RMA, so I will have to take the board out to put it in anyway :)

[Monster]

You shouldn't need to change the capacitors, unless you are running a vary high Vcore (2.1v+) with a lot of load on the Vcore regulator :)

Edit: I might be able to re-do the mod tonight with variable resistors. I've just got my powerstream back from RMA, so I will have to take the board out to put it in anyway :)
WOW? TONIGHT?! You're GREAT man! :toast:

[persivore]

I've got the board modded with 3 seperate 10k variable resistors connected instead of the 4.7k fixed resistors, and the board is working. I've got to re-mount it in a case and get the cooling sorted on it so that I can push some volts through the chip, so I'll probably start testing tomorrow :)

[Monster]

I've got the board modded with 3 seperate 10k variable resistors connected instead of the 4.7k fixed resistors, and the board is working. I've got to re-mount it in a case and get the cooling sorted on it so that I can push some volts through the chip, so I'll probably start testing tomorrow :)
Amazing! WOW! Simply the best (Tina Turner (c))! :)
P.S. Could you please overclock your Barton 2500+ up to 2300-2400MHz @ Vcore = 1.7-1.8V (in real) ? Also it would be nice if:
Vcore (in BIOS) = Vcore (in REAL) = Vcore (in IDLE) = Vcore (in FULL LOAD) !!! ;)

[persivore]

Amazing! WOW! Simply the best (Tina Turner (c))! :)
P.S. Could you please overclock your Barton 2500+ up to 2300-2400MHz @ Vcore = 1.7-1.8V (in real) ? Also it would be nice if:
Vcore (in BIOS) = Vcore (in REAL) = Vcore (in IDLE) = Vcore (in FULL LOAD) !!! ;)
I can run my 2500 at 2300mhz on 1.8v, so I'll try running at that speed. I don't know weather I will be able to make Vcore in BIOS = true Vcore because that will depend on feedback components which set the final output voltage for the Vcore regulator. I'll try to make Vcore idle = Vcore load tho ;)

[Monster]

I can run my 2500 at 2300mhz on 1.8v, so I'll try running at that speed. I don't know weather I will be able to make Vcore in BIOS = true Vcore because that will depend on feedback components which set the final output voltage for the Vcore regulator.
Is it unreal ? :(
I'll try to make Vcore idle = Vcore load tho ;)
I think it will be enough... ;) But to make Vcore in BIOS = true Vcore is more than enough ! :)

[persivore]

I've just finished testing the mod :)

All tests were done with the CPU at 200*11.5=2300mhz on 1.775v set in BIOS using prime95 under priority 4 to load the CPU. The idle voltage was taken at the windows desktop with no other applications running.

When using 3 8.2k resistors for the Isense resistors, I get a droop of 0.003v, from 1.799 idle to 1.796 under load.
With 3 10k Isense resistors I get 0.002v droop from 1.802v idle to 1.800v under load.
When the value of the Isense resistors were increased, the Vcore overvolted more, but there was less droop. With Isense resistors with a value of 4.5k, I get an unloaded voltage of 1.782. With Isense resistors with a value of 10k, I get 1.800v idle, which is a 0.025v overvolt from the Vcore set in BIOS.

I would suggest using 3 8.2k fixed resistors for the mod. Using higher values does not give much of a gain, and the droop given with these values should be enough to give some protection to the Vcore capacitors.

[Monster]

I've just finished testing the mod :)

All tests were done with the CPU at 200*11.5=2300mhz on 1.775v set in BIOS using prime95 under priority 4 to load the CPU. The idle voltage was taken at the windows desktop with no other applications running.

When using 3 8.2k resistors for the Isense resistors, I get a droop of 0.003v, from 1.799 idle to 1.796 under load.
With 3 10k Isense resistors I get 0.002v droop from 1.802v idle to 1.800v under load.
When the value of the Isense resistors were increased, the Vcore overvolted more, but there was less droop. With Isense resistors with a value of 4.5k, I get an unloaded voltage of 1.782. With Isense resistors with a value of 10k, I get 1.800v idle, which is a 0.025v overvolt from the Vcore set in BIOS.

I would suggest using 3 8.2k fixed resistors for the mod. Using higher values does not give much of a gain, and the droop given with these values should be enough to give some protection to the Vcore capacitors.

Could you please define which resistor is used to overvolt Vcore ?
And what pair of resistors are used to stop fluctuation ?
Thanx for your job!

[persivore]

Could you please define which resistor is used to overvolt Vcore ?
And what pair of resistors are used to stop fluctuation ?
Thanx for your job!
The 3 resistors circled here (http://xtremesystems.org/forums/attachment.php?attachmentid=36846) are the 3 Isense resistors, and are the resistors which were changed in the mod. All 3 of them are used to change the ammount of Vdroop, but they also have the effect of indirecly changing the default Vcore slightly

[Monster]

The 3 resistors circled here (http://xtremesystems.org/forums/attachment.php?attachmentid=36846) are the 3 Isense resistors, and are the resistors which were changed in the mod. All 3 of them are used to change the ammount of Vdroop, but they also have the effect of indirecly changing the default Vcore slightly
I got it!
Resuming you think that optimal value of these ones are 8.2KOhm?
Could you please tell what Vcore overvolting & Vcore limits (in idle & in load) will I have if I use resistors with nominal about 6.8-7 KOhm ?
Thanx a lot!

[persivore]

I got it!
Resuming you think that optimal value of these ones are 8.2KOhm?
Could you please tell what Vcore overvolting & Vcore limits (in idle & in load) will I have if I use resistors with nominal about 6/8-7 KOhm ?
Thanx a lot!
Anything over 8.2k didn't give much of an improvement in droop, but I only tested up to 10k, because that was the limit of the VR's I used.
I've got values for the resistors from between 4.5 to 10k in about 1k steps, so I'll let you know how much of an overvolt/droop you get at 6/7/8k when I find the piece of paper I wrote everything on ;)

[Monster]

Anything over 8.2k didn't give much of an improvement in droop, but I only tested up to 10k, because that was the limit of the VR's I used.
I've got values for the resistors from between 4.5 to 10k in about 1k steps, so I'll let you know how much of an overvolt/droop you get at 6/7/8k when I find the piece of paper I wrote everything on ;)
OK! I'm waiting for this data! ;)
And let me suggest you utility (http://testmem.nm.ru/snm.zip) that burns better than your soft! ;)
When I use this one the temperature is highest!

[persivore]

Here's the full list of results:
http://xtremesystems.org/forums/attachment.php?attachmentid=37091&stc=1

I'll try out that app sometime to see what the Vdroop is like.

[Monster]

Here's the full list of results:
http://xtremesystems.org/forums/attachment.php?attachmentid=37091&stc=1
I'll try out that app sometime to see what the Vdroop is like.
THANX! I'll resolder these ones to 8KOhm this week ! ;)

[Rookie]

hello guys,

please can you upload of a pic with this mod done? i'd rather not use the pencil trick permanently. i will do it to verify that it gains me stability and then i'll do it with trimmers.


i know for the experienced, your posts are sufficient but i'm new to v-modding :)

[Monster]

hello guys,

please can you upload of a pic with this mod done? i'd rather not use the pencil trick permanently. i will do it to verify that it gains me stability and then i'll do it with trimmers.


i know for the experienced, your posts are sufficient but i'm new to v-modding :)
There is no difference between mobo after & before mod! ;)

[Monster]

I've replaced these 3 resistors (4.7KOhm) by 3 others (10KOhm)!!!
Vcore is selected in BIOS = 1.600V
During testing my voltages were:
1) Vcore in load (FULL LOAD by S&M v.1.7.3) = 1.568-1.584V
Vcore in idle = 1.584-1.600V
(1.568V & 1.600V were very rare values)
2) Vmem = 3.15/3.26V (load / idle)
3) +3.30V = 3.36-3.38V
4) +5.00V = 5.00-5.05V
5) +12.0V = 11.61-11.86V
You would like to notice that Vcore drooping about ~1% (1.584V) concerning selected in BIOS (1.600V) specified by drooping of +12V voltage ~1% (11.8V)!
Now I'm VERY happy that my Vcore is rock stable!

Also I've remade Vdd mod!
I replaced 5.57KOhm resistor by one another with nominal = 2.4KOhm!
Now I can set next values of Vdd (BIOS - real values measured by MM):
1) 1.4V => 1.73V (not interested now)
2) 1.5V => 1.88V (in use now)
3) 1.6V => 1.99V (in my planes for usually work :))
4) 1.7V => 2.05V (for extremal testing using my copper Titan CUV2AB, I hope that I'll reach FSB > 230-250MHz ;))
All of these mods were done during 15-20 minutes!!!

[giorgioprimo]

Here's the full list of results:
http://xtremesystems.org/forums/attachment.php?attachmentid=37091&stc=1

I'll try out that app sometime to see what the Vdroop is like.

still anything about the overcuurent protectio?

Wich is the max vcore thsat you can reach on your mobo?

[Rookie]

monster,

i just did the vdd and vdimm mod per motherboardfaq site. i have the vtt and this v-droop mod let. i finally found pics on how to do the vtt mod, but i still cant figure out how to add resistors for this v-droop mod.

i'm an XXXXXXXXXXXtreme newbie but i've so far performed the 2 v-mods above with success. my board will eventually be w/c'd with a v-modded 5950u card.

thanks and i know i'm being a pest but the technical understanding involved with out seeing any pics is beyond my intelligence.

:)

[Rookie]

guys i did the pencil vdroop mod on my nf7-s v2 and now i'm 2.7ghz stable on air. since the mod is putting resistance in between both ends of the pins. i can just solder to each side of the pin with the resistor in the middle.

is this correct? pins 1&2 regulate the vcore fluctuation by adding resistance, on pin 3 adding resistance increases the vcore.

so for pins 1&2 on each pin, put a vr in between, same for pin 3?

is this correct, even though i did this mod with a pencil, i want something more permanent, i want to leave my machine folding 24/7, i feel that using resistors is more reliable.

help please :) a picture would be nice