Asus Striker Extreme vDroop Mod

[celemine1Gig]

...Kobalt, everytime I make the pencil marks to the resistors, should the number (lets say 129.5) start going up or down in the multimeter?

Answer me this question and Ill be ready to go!
...

Well, I'm not Kobalt, but I'll still answer you that question. The resistance will go down because that's the only possible consequence. What you do when penciling a line next to the resistor (or whatever part) is creating a resistance in parallel. The rule for parallel resistance is that the total resistance can never exceed the resistance of the smaller (in value, not in size) resistor. That's to say for example, when connecting a 10 Ohm and a 10000 Ohm resistor in parallel, you'll always get a total resistance lower than 10 Ohm. To be exact, it'd be around 9.99 Ohm in the example.

Hope the answer will help you. Good luck with your mod.

BTW, the formula is:

R(parallel)= (R1 * R2) / (R1 + R2)

Increasing the resistance is only possible when adding a resistor in series. In practice this would mean cutting the trace and reconnecting it with a resistor.

[slim142]

Thanks man, I appreciate your help too.

Gonan try this as soon as I can, I really hate this vdroop!

[ineedaname]

Well i was gone for the weekend so i couldn't really answer your questions. But it seems like you're pretty much set now.

One thing I do have to mention though is that having a lower B pencil does two things. First its less conductive so you need more and second is that it tends to be less sticky so it falls off as soon as you rub it on the resistor. This is cuz the surface of the resistor isn't exactly ideal for marking on.

I definitely suggest using the multimeter but just to give you an idea, when i drew on the resistor it pretty much looked covered with dark pencil lead.

[slim142]

Even when I do like 10 marks, the resistors look shiny but they dont seem to get fixed. I can mark them 10 more times but they would not get fixed neither. I think I will have to get the multimeter just to make sure im not messing up anything.

Also there is something weird that is happening to me... please read the following topic

http://www.xtremesystems.org/forums/...d.php?t=137655

[Kobalt]

If you're doing a pencil mod without a multimeter, you're just asking for problems.

[slim142]

yes I just noticed that

[JPxM0Dz]

slim142 , here's a couple helpful links on how to get familiar with a digital multimeter if and when you decide to use one to complete this mod

http://www.floyd.k12.va.us/itrt/pwrp...#256,1,Digital
http://www.umd.umich.edu/casl/natsci.../DigiMeter.htm

[ineedaname]

Man if the resistor looks like its covered and you're still not getting the right results I really wonder if you're marking the right one?

[Kobalt]

Man if the resistor looks like its covered and you're still not getting the right results I really wonder if you're marking the right one?

I doubt it...I took my resistor down from 130k to 9.75k and I still get .02vdroop (although it is down from .07vdroop).

[ineedaname]

I doubt it...I took my resistor down from 130k to 9.75k and I still get .02vdroop (although it is down from .07vdroop).

LOL at that point i guess you really could use the Circuit Pen on it.

[Kobalt]

yea...i give up. If this was for my computer it would be going straight back to Asucks. Its for a friend so I don't care that much although im kind of a perfectionist.

[ineedaname]

yea i know wut u mean. I couldn't get my buddy's e4300 to run over 3.3ghz and i was so pissed off. I was so adamant that it would hit 3.6 but it was running too hot to get there. In the end I just left it at 3.3ghz.

BTW slim you might wanna try taking off the vCore mod and leaving the vdroop. If you say they're affecting each other then stick with the vdroop cuz you can always just enter a higher voltage into the bios instead of a vcore mod.

[xs64]

Hi guys , i have a P5N32E SLI , i measured Vcore and saw that real Vcore is just below 0.02v than Vcore was set in Bios. Ex : I set 1.425v in Bios and real Vcore is 1.40v in DMM

And i have no Vdroop at all. I dont care what monitoring in Asus Probe.
So i wonder how do you guys know real Vcore ?

I measure at this red point in the picture

[slim142]

ineedaname, I have the DMM now, I will first read the resistence and show them here to see how they look.

[slim142]

ineedaname

http://www.xtremesystems.org/forums/...d.php?t=137655

[BenArcher]

What is an acceptable amount of VDroop in you minds?

Also is having a negative VDroop bad? As in voltage goes up under load

Anyways I had a VDroop between 0.06 & 0.07 Volts so that seemed pretty stupid. I did the VDroop Mod & reversed teh problem so i was actually gaining 0.02V under load. So I rubbed some off & now I'm getting 0.02 VDroop atm.

The reason i ask about the negative VDroop is because it would seem like a good thing to me if i can run VCore at 3.5 V which is stable until I run stress stuff. But the CPU takes up to 1.425 before temp becomes a prob.

[ineedaname]

I have a tiny bit of negative vdroop depending on what voltage I'm at. There's nothing wrong with negative vdroop as long as its not ridiculously high.

[slim142]

Benarcher what exactly did you do? how and what did you use?

[xs64]

ineedaname, thx for the guide.
I did the Vdroop mod and it works perfect !

[theteamaqua]

wahts the vcore measure point??? i have P5N32-E SLI PLUS btw .. but it seems like its the same PCB as teh striker

[ineedaname]

if u look up like 7 posts it shows the measure point.

Its any of those coils

[pygothian]

Before I was requiring 1.425V in bios to get 1.38V in probe II

Tried the vcore mod 1st expecting if I reduced the resistence I would get a closer vcore to what was set in the bios - the exact opposite occured when i reduced the ohms with multimeter and 4B. So I used ethanol to get it off.

I then tried the vdroop mod, as the 1.38V would drop to 1.36 with 100% load.

Dropped the resistence 100, (which was 3 strokes with 4B) and amazingly, now I can lower the bios vcore to 1.3825, probe shows 1.38 and it stays at 1.38 on 100% load. Managed to drop about 5C off the processor by reducing the vcore. Which I will now use to hopefully squeeze a few more hundred Mhz out before idling at 43C again and load to 50-51C.

Thanks for the useful info

[celemine1Gig]

Hehe, 5°C off the CPU, but how much more on the vcore regulation?
Did you by chance measure the regulator temps. Especially the mosfets should get a little warmer now.

[ineedaname]

Its nice 2 hear those great results pygothian.

Hmmmm seems like there's something wrong with the Vcore mod. There are a few ppl who tried it who said it didn't work. If that's the case then I'm gonna take it off the site. For that mod I followed what Nicke put on that post but it seems like its in question.

I'm gonna take it off for now and review the mod again.

That mod doesn't seem that useful though since we got more than enough voltage settings to fry your cpu's anyway.

[pygothian]

Yeah. I've been monitoring my temps for the past month. To get a stable overclock of 3375Mhz on my E6600 I needed these setting (Orthos >12hrs)

SPP - MCP 200
Unlinked
FSB 1500
DDR 875 4-3-3-9 RRD 5 RC 18 WR 6 WTR 8 RFC 38

Vcore 1.425 ( Probe II measured 1.38)
Mem 2.075 (2.13)
1.2V 1.3 (1.36)
NB 1.4 (1.44)
SB 1.55 (1.58)
CPU VTT 1.4 (1.34)

With the Vmod Dropping the resistence down at 10 at a time, the Vcore setting in the bios and probe actually decreased further so at settings 1.425 and with 50 off the resistence the computer was unstable and when got to windows probe measured it at 1.28V (well below the setting of 1.425)

I removed the vmod with ethanol, and tried the vdroop - which I didnt really think I needed. It was about 0.02 which I thought wasnt too bad.

Dropped 100 off the resistence, and found when the vcore in the bios was 1.425 that was actually delivered in probeII. So I dropped the bios setting to 1.39 which I had found was the previous stable settings. And it came out at 1.39 in probe II also. The 0.02 vdroop had also resolved.
I dropped it to vcore to 1.36 then - and found it was still stable with orthos for >12hrs.

I have since reduced
1.2V from 1.3 to 1.2
NB from 1.4 to 1.35
SB from 1.55 to 1.5
and CPU VTT to auto

Also tightened the subtimings from the great guide on this site called
680i ram timing guide - And it now runs stable, and 5C cooler.

I also got my memory read above the 10000MB/sec for the 1st time )

Tonight I am going to attempt increasing up to FSB of 400x4 (1600) as previously I have never got a stable system at 3600Mhz and to get it stable I was scared that I needed too high vcore and idle temps were in high 40's with stress gettings close to 60C

I also monitor with the striker optional themistors bluetac'ed in place

1 ram heatsink temp (OPT1) 37C
2 NB heatsink temp (OPT2) 43C
3 Middle HDD in raid) (OPT3) 27C

Core temp currently is 39C with ambient temp 23C (idle)
In probeII Mobo temp 31C
CPU temp 27C

Anyone know a way of better cooling the northbridge - or does that elaborate copper heat pipe have to entirely come off?

Thanks for all the great info...this site is by far the most informative I've read

Cheers