***OFFICIAL Asus P5K-E/WiFi Thread

[Gen]

can those bug finding please be confirmed ? because i am using all my options on auto almost and if those reading are true ??? wtf ?

I just confirmed the bogus voltages - as syl said this is a big deal, especially the Vpll - mine is at 1.84v when set to either Auto or 1.55v. I also have a Vmem bug where Auto = 2.05v (my lowly XMS ddr667 was really impressing me making it up to 950 with allegedly only 1.8v - now I get why when I would change Vmem from Auto to 1.85 or 1.9v I would BSOD on bootup).

the "voltbug", there is on more of p5k's family: p5k, p5k-e, p5k pro....

it was confirmed by Plester a few posts up and by BarboneNet many pages back in this thread.

[TheVaLVe]

1013 bios also cured the x2 pci-e syndrome on my mobo, even though it is not 100% cured. But i also suspect that this syndrome come from the quad core that is very burdening on mobo and it needs more voltage than 1.7v on NB. Well overall pretty happy since i can run 6GB RAM now with all slots occupied.

[syl]

well.. maybe not the steadiest voltages ever..

then again boards do differ but only a bit

Steady As She Goes ...the story may change if I plug in a Quad of course:

[xwards]

on my p5k-e wifi with 1012 bios i measured Vpll:
and at 1.5V set in bios, i get 1.56V measured

[syl]

on my p5k-e wifi with 1012 bios i measured Vpll:
and at 1.5V set in bios, i get 1.56V measured

That's interesting...I see the volt-bug with 1012 and 1013 bios versions on my P5K-E. What processor are you running and which revision is your board?

[Gen]

1013 bios also cured the x2 pci-e syndrome on my mobo, even though it is not 100% cured. But i also suspect that this syndrome come from the quad core that is very burdening on mobo and it needs more voltage than 1.7v on NB. Well overall pretty happy since i can run 6GB RAM now with all slots occupied.

what settings are you using that need higher than 1.7v NB?

with 4 dimms populated i only need 1.5v for 433x9 - 3.9g, ram 1:1 on a Q6600. (granted that's not a high mem clock)

[TheVaLVe]

what settings are you using that need higher than 1.7v NB?

with 4 dimms populated i only need 1.5v for 433x9 - 3.9g, ram 1:1 on a Q6600. (granted that's not a high mem clock)

With 1.7v i get 484x8, 4dimms populated with ram 12:10 at 580MHz, and perfomance level 3. Below 1.7v, it isnt stable on Portal...

[xwards]

That's interesting...I see the volt-bug with 1012 and 1013 bios versions on my P5K-E. What processor are you running and which revision is your board?

rev. 1.02g
cpu is E4500

[CrazyNutz]

Except 1.5V PLL in BIOS is NOT 1.5V actual:



All P5K-E owners (maybe other Asus M/B's as well) should keep this in mind because they may be running much higher voltage levels than they think they are. I have taken DMM measurements to confirm the above findings.

Yep, you are correct. I just measured mine, I got 1.86v for PLL, and it was set to 1.50v in the bios. I have it set to 1.6v in bios now, and
I get 1.65v, so I guess that as low as it will go

On 1013 bios right now. Prob. same with other bios' as well.

[Gen]

With 1.7v i get 484x8, 4dimms populated with ram 12:10 at 580MHz, and perfomance level 3. Below 1.7v, it isnt stable on Portal...

that explains it then

[Stijndp]

Could anyone confirm if the P5K-E can boot perfectly without a graphic card inserted?

Thanks to the person doing the trouble of a double reboot

[holwill]

Hi Guys, I have a friend who gets his P5k-e wifi next week and I have a couple of questions for the guys running them. Firstly is the best bios, still as per the front page "0602" for a Q6600? He is looking for somewhere around 3.6ghz if he is lucky or are the later 1012-1013 a better bet? Secondly on the upper mosfets which have a copper sink on my P5K-Premium, are the mounting holes there for adding sinks or water blocks? It is quite hard to tell from the pic's I have seen.Many thanks in advance guys.

[Zero939]

Hi Guys, I have a friend who gets his P5k-e wifi next week and I have a couple of questions for the guys running them. Firstly is the best bios, still as per the front page "0602" for a Q6600? He is looking for somewhere around 3.6ghz if he is lucky or are the later 1012-1013 a better bet? Secondly on the upper mosfets which have a copper sink on my P5K-Premium, are the mounting holes there for adding sinks or water blocks? It is quite hard to tell from the pic's I have seen.Many thanks in advance guys.

602 or 503.

[GoGoGadget]

I wonder why they completely removed bios 1004 and 1006 from the bios section on the Asus site. That's kinda weird considering all the older bios is still listed...

[CrazyNutz]

I wonder why they completely removed bios 1004 and 1006 from the bios section on the Asus site. That's kinda weird considering all the older bios is still listed...

Prolly had some bug in them that was silently, and/or slowly killing our hardware, that they are not willing to confess to I honestly would not doubt it.

[munim]

Could the people that have confirmed the voltage bug also tell us which revision of the board they have? I ask because the person that had 1.56v vpll when setting it to 1.5v in bios had the newer 1.02g revision board.

[CrazyNutz]

Could the people that have confirmed the voltage bug also tell us which revision of the board they have? I ask because the person that had 1.56v vpll when setting it to 1.5v in bios had the newer 1.02g revision board.

I have rev. 1.02g and my PLL was set in BIOS to 1.50v, however when tested
with a DMM the reading showed 1.86v, yes very shocking!

This is on BIOS 1013, and I dont know if this was occuring on other bios revisions.

I currently have it set to 1.60v in bios which gives me 1.66v when read with
my DMM, so at this point 1.66v is the lowest PLL voltage the board will
supply to my CPU.

[munim]

I'm pretty new to overclocking so I have a possibly silly question. my rig:
E6300
p5k-e 1.02g 1012 bios
ocz reaper 6400

The only things I really mess with when overclocking are vcore, vNB, and vmem. Question is, my cpu/board seems to have a sweetspot of 3ghz with only 1.23v ( I think this is good) but for instance 3.15ghz requires 1.28v ( load voltage). Is this a significant jump? Can it be that such a high vcore is required to mask the instabilities arising from the other screwy voltages as posted above? Or does it not work that way? BTW, all other voltages etc. are set to auto except the spread spectrums and NB is at 1.4. Anything other than 1.4 causes pci-express to go to x1. I've also disable dram static control and transaction booster, relaxed to 3rd level.

[syl]

on my p5k-e wifi with 1012 bios i measured Vpll:
and at 1.5V set in bios, i get 1.56V measured

Could the people that have confirmed the voltage bug also tell us which revision of the board they have? I ask because the person that had 1.56v vpll when setting it to 1.5v in bios had the newer 1.02g revision board.

Xwards, your measurement is absolutely correct but you may still have the volt-bug Here's why:

Barbonenet had mentioned earlier that setting Vpll, Vtt/Vfsb, and Vmch to their lowest level is equivalent to setting them to Auto. What was not mentioned was that the Auto settings are FSB dependent, meaning the voltage levels are mapped to certain FSB frequencies.

Here's what I have discovered with my trusty DMM:

Vpll set to 1.5v or Auto
FSB-----Vpll Measured
333-----1.550
367-----1.654
400-----1.757
434-----1.861

Vtt/Vfsb set to 1.10 or Auto
FSB-----Vtt/Vfsb Measured
333-----1.111
423-----1.204
459-----1.293
495-----1.387
*Note: Vtt/Vfsb measurements for non-45nm chip may be higher.

Vmch set to 1.25 or Auto
FSB-----Vmch Measured
333-----1.261
351-----1.412
400-----1.564

These measurements were taken from my Rev 1.02g board with 1013 bios and 3110 Wolfdale. I suspect this volt-bug is present on all board revisions and bios versions, but please perform your own measurements and share your results.

Hope this is useful information to those perplexed by the strange behavior of these voltage adjustments.

[CrazyNutz]

Xwards, your measurement is absolutely correct but you may still have the volt-bug Here's why:

Barbonenet had mentioned earlier that setting Vpll, Vtt/Vfsb, and Vmch to their lowest level is equivalent to setting them to Auto. What was not mentioned was that the Auto settings are FSB dependent, meaning the voltage levels are mapped to certain FSB frequencies.

Here's what I have discovered with my trusty DMM:

Vpll set to 1.5v or Auto
FSB-----Vpll Measured
333-----1.550
367-----1.654
400-----1.757
434-----1.861

Vtt/Vfsb set to 1.10 or Auto
FSB-----Vtt/Vfsb Measured
333-----1.111
423-----1.204
459-----1.293
495-----1.387
*Note: Vtt/Vfsb measurements for non-45nm chip may be higher.

Vmch set to 1.25 or Auto
FSB-----Vmch Measured
333-----1.261
351-----1.412
400-----1.564

These measurements were taken from my Rev 1.02g board with 1013 bios and 3110 Wolfdale. I suspect this volt-bug is present on all board revisions and bios versions, but please perform your own measurements and share your results.

Hope this is useful information to those perplexed by the strange behavior of these voltage adjustments.

Nice, thanks for creating this chart! I have my FSB set @ 440Mhz so I was
getting the 1.86v PLL, That is really disturbing to me! But now I have PLL
set to 1.6 in bios, and VTT set to 1.2, that gives me PLL 1.66v(DMM)
and VTT 1.2v(DMM). These voltages Im ok with.

[krulin_m]

Eeww... So from what I am reading, with the voltages I have had set, I might be cooking things.. o.O

[TheVaLVe]

Can someone help me, Is there any VNB voltmod for this board?

[xwards]

syl: you are right, at 420FSB with Vpll set to 1.5v or Auto i get 1.87V so thats not good

[Plester]

freaky that the voltages ramp with fsb - this seems less a 'bug' than a dirty trick on asus' part.

[CrazyNutz]

freaky that the voltages ramp with fsb - this seems less a 'bug' than a dirty trick on asus' part.

I Agree! Possibly to make ASUS boards "appear" to OC better @ lower voltages than the competing boards.
Or It could just be a bug, but I like the conspiracy theory better