ASUS P5W DH - Problems + Fixes Thread

[Falkentyne]

Yeah vdroop is the voltage difference between idle and full load.
Vdrop is usually considered to be the difference between BIOS setting and BIOS readout, although this is not accurate because the BIOS actually puts a small load on the cpu. (you can tell by comparing BIOS temps to idle temps).
So Vdrop is the difference between BIOS voltage setting and idle voltage.

Whether the poster's chip can boot at 0.86v depends on its default vid, I think...it's possible a 1.15v vid chip can load windows at .86v, but I sure wouldn't think it would pass prime.

As Yolanda said, a VERY good test is to check load temps.

[marzliak]

With 1.7V setting, you can see the real readout on the voltmeter... any suggestions?
I have this board since 2006 and thus far i've installed more than 30 CPUs to test out... that very problem with the Vcore happens only with the E7000....

FU** I can´t even see the message "To adjust Vcore, please disable blablabla".

Bios 2901 @ V1.02g board.

I used a trick to mantain vcore@1,4v even in "vcore auto only" option. If u guys want to know how, please tell me.

[SOS_DK]

I have a P5W DH Deluxe (1.04G / 2901 BIOS) with Core 2 Duo (6400) CPU.
4x1Gb Kingston PC2-5300 RAM (CL5). BIOS default configuration.

I would like to upgrade to a Quad core because I need the extra cores, and preferably also higher frequency.

The compatibility matrix for the board clearly states that several 45nm Quad cores CPU's are supported, but we all know how much help I'm going to get from ASUS if something doesn't work :-)


So I'm looking for someone to confirm that some of the CPU's I'm looking to upgrade to are not only supported but also work! I don't Overclock so standard performance (with standard cooler) is fine by me.
Also since I don't understand much of FSB (board is 1066MHz max I think, anjd Quad cores CPU's are 1333MHz, does it just clock down or will BIOS now support ?)
(http://support.asus.com/faq/faq.aspx...luxe&product=6)


Does anyone have any experience/problems/suggestions, perhaps these (which I can buy cheap)?
- Core 2 Quad Q9550 12 MB (Intel Boxed)
- Core 2 Quad Q6600/Q6700
- Something you had working ?

Both are on the ASUS compatibility list, and the Q9550 seems like the best choice (and cheaper than Q6600), but the Q6600 as I understand is 1066MHz/non-45nm and might be a better match?
(http://support.asus.com/cpusupport/c...Language=en-us)


I am looking for the most stable upgrade (that works with my RAM), speed is secondary :-)

PS, I have been looking through this thread, and there are a LOT of posts and all of you are OC'ing some of these CPU's, but its difficult for me to see if I can do this without a complex OC setup/cooling/BIOS config.

[MarcusTaz]

I will have to find what page I posted my bios settings I think on page 226, but I ran my Q6600 GO step @ 3.4 all day long with mem @ just about 800mhz. It is a good board but the vdrop and heat issues with the 975x will give you that FSB wall. I am not an expert as some are but I feel it is a great board, still run one with a a newer E5400 as a backup gaming rig for my nephew. I did test out a Q8200 (crap quad) for a short period before I sold it and it ran fine but I did not check frequencys in terms of FSB etc... Hope this helps.

[secretworld]

Yes Q6600 running 24/7 @ 3ghz with all volts stock. Only set fsb to333 and ram to corsair timings and volts.
Good luck
Vincent

[Falkentyne]

With a 1.04g, the only cpu's you will have problems with are possibly vcore settings on the E7x00 and Q8x00 chips; all others should work fine.
You're just not going to get much higher than 370-380 fsb with a quad, and that's with a lot of voltage going into the chipset; a dual should reach 400 fsb or a bit higher. Lack of PL (TRD) and TRFC settings and GTL Refs is what holds you back (plus the board seems to be missing GTL reference data paths (on the PCB itself) for cores 2 and 3, which causes it to fall back on the corresponding address values for cores 0 and 1, which obviously don't help matters).

[SOS_DK]

I will have to find what page I posted my bios settings I think on page 226, but I ran my Q6600 GO step @ 3.4 all day long with mem @ just about 800mhz. It is a good board but the vdrop and heat issues with the 975x will give you that FSB wall. I am not an expert as some are but I feel it is a great board, still run one with a a newer E5400 as a backup gaming rig for my nephew. I did test out a Q8200 (crap quad) for a short period before I sold it and it ran fine but I did not check frequencys in terms of FSB etc... Hope this helps.

But does that mean my 667MHz DDR2 is a problem or only if I OverClock it?

With a 1.04g, the only cpu's you will have problems with are possibly vcore settings on the E7x00 and Q8x00 chips; all others should work fine.
You're just not going to get much higher than 370-380 fsb with a quad, and that's with a lot of voltage going into the chipset; a dual should reach 400 fsb or a bit higher. Lack of PL (TRD) and TRFC settings and GTL Refs is what holds you back (plus the board seems to be missing GTL reference data paths (on the PCB itself) for cores 2 and 3, which causes it to fall back on the corresponding address values for cores 0 and 1, which obviously don't help matters).

:-) being ignorant as I am to OverClocking I dont understand half of what you just wrote, but I hear you and the others are saying, is that it will support almost all CPUs if I keep it at stock settings... which is what I intend to do.... :-)

[Yolanda]

I think Q6600 will be a good choice for you, because Q9550 costs more and overclocks at this board to 3.0-3.2GHz, with Q6600 you will get the same 3.2GHz without problems and even can reach 3.4-3.6GHz.

Or maybe just overclock your dual core to ~3.2GHz and see new performance.

Also new Core i5 570, starts shipping today, with new Socket 1156 board will be very nice for quad core computing.

[SOS_DK]

I think Q6600 will be a good choice for you, because Q9550 costs more and overclocks at this board to 3.0-3.2GHz, with Q6600 you will get the same 3.2GHz without problems and even can reach 3.4-3.6GHz.

Or maybe just overclock your dual core to ~3.2GHz and see new performance.

Also new Core i5 570, starts shipping today, with new Socket 1156 board will be very nice for quad core computing.

Actually the Q9550 is 40$ cheaper in Denmark than the Q6600 :-) which is why I asked about that one specifically. And I will not do any OverClocking, I dont have the time to play around with it so I will just plug'n'play the new CPU...

[Yolanda]

In this case Q9550 will be better, however overclocking both CPUs to 3GHz is very easy.

[G40RoCkEt]

Maybe someone says that the Q6600 perhaps is better because is 65nm and FSB1066 what match perfectly with P5W DH native supports. And that's right!

Otherwise, the Q9550 as Yolanda said maybe overclock at maximum at 3.2Ghz. But 3Ghz is beat for sure with Vcore CPU at AUTO. And is 1333FSB CPU with 12MB wich is good for large apps.

My guess is: your best choice now is Q9550.

Reasons:

- Lower temps
- Good overclocker
- Powerfull processor

So, tell news after.

[Yolanda]

Read an article, how to make working SLI on all Intel chipsets, even without modified NVidia drivers (only one Windows DLL cracked and some data added to Windows registry). It was tested on X48 and i965P, so I think our board now is SLI capable also =) However my GTX 295 is a single slot card, so I can't test SLI on P5W myself.

[Falkentyne]

Can you PM me the link to that article?

[Supa Chiney]

Quote:
Quote:
Originally Posted by Falkentyne View Post
Remove one stick of memory and see if the problem goes away. Or switch the remaining stick. Memory that is dying (like micron D9's) will cause this to happen.

Raise the memory voltage to 2.1v .....does the problem go away then?
Quote:
Originally Posted by Supa Chiney View Post
this sticks are brand new, i did that when i first installed them, i forgot to mention that this problem started when i added a pair of the same type i had before (on my sig) but those were different version, so i though that was the problem, that's why i bought the new ram.

I cold booted twice this morning without issues will, report back in a couple days or later if problem stops/continues

is it safe to run them at 2.1v even if they are rated to run at 1.8v ?
the only way around i found to "fix" the problem is to switch the PSU everytime i'm gonna shutdown and it boots everytime i turn it on

Update.... i found out what the problem was.. it was my PSU that is faulty I tried my old ULTRA PSU that I had on my bro's PC and it instantly booted, i ordered a new PSU right away

[SOS_DK]

So I bought it, installed it, at first boot, the bios said 2,2xGHz....???... I was forced to CMOS, saved & rebooted and it went dead.

Reset the CMOS with jumper, got into bios, everything set at auto by default, and it set everything right, FSB to 1333MHz and CPU 2,83GHz, happy day. I saved, rebooted -> board went dead. Tried that several times....

Finally I set the BIOS JumperLess part to manual (changed nothing else), started with Bus Speed of 233MHz, worked fine, worked my way up to 266/300 and finally 333MHz, and still it works fine. Now everything is the way it should be, but still set to manual, changed nothing else. According to CPU-Z it multiplies with 8.5 (which someone said on a ASUS forum was only possible in auto).

I am unable to compare what other changes might be set when using "auto" since i cannot boot up to use CPU-Z or even go to CMOS for comparison.... but at least it works :-) Now comes a few days of stability testing...

[pc1x1]

What settings are you guys using for a Q9550 on a 1.02g board?
Bios: 2901

Also computer is a tad noisy, even with a fan controller, any ideas?

Ps. Also my Plexor DVD Burner, can't boot DVDs at all, but it can boot CD's? Think its a problem with the burner or mobo? Seems burner? Looking for different media to reburn DVD.

[pc1x1]

Update:

Yea it was the media, strange it wouldn't work with RiData DVD-RW, whereas all my other burners worked, did work on a spare DVD-RW I had though.

Anyhow off to install Windows 7. Will be here for some settings configs later , would like to get the CPU to 2.8ghz, or 3.0 if possible.

[MikeR397]

Sorry for my bad english.......

Auto = 1.040V
1.400V=0,0992V
1.700V=1.27V (for my best overclock)

Test all Bios, clear CMOS. Never readout Vcore more 1,27V with Е7200 and E7300. E6600 working good!
My be revision is old for this CPU

I am having the exact same problem as DJBamby. Please see here for a starting reference: http://www.xtremesystems.org/forums/...postcount=8311. Then see the conversation after the post (#8307 by Falkentyne) on this page: http://www.xtremesystems.org/forums/...10193&page=333

I have a Asus P5W DH Deluxe Rev 1.02g with latest Bios 2901 (I've tried 2801 with same problem). I just upgraded from a E6600 to a E8600 processor and now am having screwy Vcore issues (which I understand is standard for the 1.02g Rev of this board, but there has to be some fix). I am getting the exact same actual Vcore readouts as DjBamby....

1. Auto Vcore - 1.07v
2. Anything under 1.525v Bios vcore setting will not give enough juice to allow the system to post.
3. 1.55vcore is really about .9v and allows post and windows startup. I've verified the actual is .9v in both bios and multiple CPUZ versions. I do not have a voltmeter, but DjBamby did and confirmed the actual voltages this way too.
4. 1.70v is actually 1.26v. I KNOW THIS WOULD FRY THE 8600 (45n) at really 1.70v, but its not getting 1.70, it is getting 1.26v. I can confirm this with the temps also, which are about 50c at 420 FSB (4.2ghz is about my max stable overclock in Prime95 with this voltage, boots fine at 4.35ghz but not stable).

So, chosing 1.70v in Bios gives actual 1.26v to the E8600 and allows me to hit 4.2ghz (420FSB) prime stable. I WANT to take vcore up to 1.40v actually, and I'm sure I could hit 4.4-4.5ghz plus with that voltage (on air with well ventilated case, Zalman CNPS9500 heatsink). I just got this chip for only $187 (less $40 in various cashback) new on Ebay, and I don't care that I'm slightly risking its life with 1.40v. I usually upgrade processors every 2-3 years anyway and have never burned one out.

Any help or suggestions would be much appreciated. I saw someone say they found a way to run at 1.4v with this problem, and would like those details please (djbamby?)! I've tried calling Asus, but they don't help. I understand a Rev 1.04g of this board would solve this Vcore issue, but I really don't want to put in a new motherboard and go through that hassel, even if I could get them to exchange my board (3 year warranty).

Thanks so much!

[mikepaul]

I don't know what to say. My 1.02G board sucks, and despite calls to ASUS where they SAID they'd try to recreate my 40nm issues at their shop using a board they had, no such thing ever happened. You'd have to send the board back and if they prove that it isn't good, then MAYBE you'd get a working board, but they seem obstinate enough about 1.02G not being 'bad' that they might send you another of those.

I'm having a hard time implementing my boycott of ASUS boards for my i7 project. ASUS seems to make the best boards, but after they screwed with me for weeks promising stuff that ended up never happening, I want to live ASUS-free from now on. Good thing I have no immediate plans, so maybe next year I can get going on it...

[MikeR397]

Does it sound like this is a bios version issue? I'm pretty sure you need 2801 or 2901 in order to support the e8x00 series 45n chips, and I've tried both of those. Both have the same vcore problem, but actually the 2901 bios gives more voltage than the 2801. With 2801 bios 1.70vcore gives about 1.19v actual, while 2901 bios 1.70v gives 1.26v. Who knows when/if they will release a newer bios that might help a little more. All I want is to be able to get vcore up to 1.40 actual (and while setting it to 1.70v is unnerving on paper, I don't have a problem if it gets me 1.40v acutal).

I thought I read that the problem with 1.02g boards was that they could only set at 1.45v core or higher. I'd consider, hesitantly, setting 1.45v on the e8600 if that was an option. However, like DJBamby, my 1.02g board has a significant undervoltage problem.

I meant to call Asus today (before 5) for thier technical support, but didn't get a chance. Thier 24 hour customer service support is anything but support.

[mikepaul]

My Q9450 has never been stable unless I underclocked it. I even tried the lowest voltage 2801 allows as a manual setting (1.45V I guess) and that also rebooted on me. I guess it's not insufficient voltage like I thought. Since the latest BIOS hasn't been reported to make a difference I haven't bothered with it.

I still see ASUS pulling a too-many-defects-deny-everything on the 1.02G problem. To tell me that NOBODY ever sent back a 1.02G with the kinds of complaints we have is hard to believe. I just toughed it out rather than be without a PC for a month, but I can't see everybody being that willing to put up with junk. Must be they ran out of bad boards before the hard-to-please types got theirs, or another reason for the returns was going on...

[Falkentyne]

I am having the exact same problem as DJBamby. Please see here for a starting reference: http://www.xtremesystems.org/forums/...postcount=8311. Then see the conversation after the post (#8307 by Falkentyne) on this page: http://www.xtremesystems.org/forums/...10193&page=333

I have a Asus P5W DH Deluxe Rev 1.02g with latest Bios 2901 (I've tried 2801 with same problem). I just upgraded from a E6600 to a E8600 processor and now am having screwy Vcore issues (which I understand is standard for the 1.02g Rev of this board, but there has to be some fix). I am getting the exact same actual Vcore readouts as DjBamby....

1. Auto Vcore - 1.07v
2. Anything under 1.525v Bios vcore setting will not give enough juice to allow the system to post.
3. 1.55vcore is really about .9v and allows post and windows startup. I've verified the actual is .9v in both bios and multiple CPUZ versions. I do not have a voltmeter, but DjBamby did and confirmed the actual voltages this way too.
4. 1.70v is actually 1.26v. I KNOW THIS WOULD FRY THE 8600 (45n) at really 1.70v, but its not getting 1.70, it is getting 1.26v. I can confirm this with the temps also, which are about 50c at 420 FSB (4.2ghz is about my max stable overclock in Prime95 with this voltage, boots fine at 4.35ghz but not stable).

So, chosing 1.70v in Bios gives actual 1.26v to the E8600 and allows me to hit 4.2ghz (420FSB) prime stable. I WANT to take vcore up to 1.40v actually, and I'm sure I could hit 4.4-4.5ghz plus with that voltage (on air with well ventilated case, Zalman CNPS9500 heatsink). I just got this chip for only $187 (less $40 in various cashback) new on Ebay, and I don't care that I'm slightly risking its life with 1.40v. I usually upgrade processors every 2-3 years anyway and have never burned one out.

Any help or suggestions would be much appreciated. I saw someone say they found a way to run at 1.4v with this problem, and would like those details please (djbamby?)! I've tried calling Asus, but they don't help. I understand a Rev 1.04g of this board would solve this Vcore issue, but I really don't want to put in a new motherboard and go through that hassel, even if I could get them to exchange my board (3 year warranty).

Thanks so much!

I can't help with this. I don't have a 1.02 motherboard. Mine is 1.04. Your only solution is to contact Asus and ask for a BIOS that can fix this; and it may be possible that that revision simply is unfixable. A more permanent fix is to spend the $100 and grab a gigabyte EP45-ud3p. Some things in life are simply not worth the stress in worrying about; and I don't know your money situation, but if you had enough money to get a Wolfdale, surely you can find a way to save up something for a UD3P motherboard, which will let you run your CPU at proper settings?

[MikeR397]

I can't help with this. I don't have a 1.02 motherboard. Mine is 1.04. Your only solution is to contact Asus and ask for a BIOS that can fix this; and it may be possible that that revision simply is unfixable. A more permanent fix is to spend the $100 and grab a gigabyte EP45-ud3p. Some things in life are simply not worth the stress in worrying about; and I don't know your money situation, but if you had enough money to get a Wolfdale, surely you can find a way to save up something for a UD3P motherboard, which will let you run your CPU at proper settings?

Thanks for the response, I'll keep the EP45-ud3p in mind. It's not really about the money, but the fact that I'm really not crazy about gutting my computer to install a new MB in conjunction with spending the money. Also, given I can already squeeze 4.2ghz with the crazy 1.70=1.26v configuration, there is not a lot of incentive for the money and time together just to try for the 4.4-4.45ghz I expect I could get on air with 1.40v actual. In all honesty, I'm not sure I notice too much a difference going from my OC'ed e6600 at 3.35ghz to the e8600 at 4.2ghz. I certainly doubt I'd notice a .2ghz gain. The only reason I upgraded is that my E6600 sold after 2 months on Amazon for $100 (crazy in my opinion), and it only cost me about $50 after the e660 sale and cashback to get the E8600 on ebay.

I'm going to keep tinkering with my existing setup and talk to Asus again on the phone. I'm certainly not letting it stress me out. My computer, and building/tweaking it, is just a hobby. I enjoy attempting to maximize the value out of my components, both new and old. I could afford to buy bleeding edge if I wanted, but I get more satisfaction when buying "at the sweet spot" and overclocking for maximum value. Not always the easiest, but its where I get the most satisfacation.

I'll let you know if I make and progress or find a solution. Please let me know if anyone else has any recommendations for the screwy 1.70v=1.26v issue.

[MikeR397]

Just to add a datapoint, I tested all compatable bios versions from 2901 backwards (to about 25xx I think) and all bios have the same vCore problem too.

For some reason, Asus update utility now fails the bios flash everytime, leaving the board without a bios. Fortuntely, they have a nifty feature that lets you add a bios file to a usb and when you restart (the second time, the first is a no post), Asus looks for a file called "P5WDH.ROM" to flash automatically (you do need to rename the bios to that). What a good feature (in addition to the bios default setup reversion when you overclock too fall).

Anyway, I'm done trying different bios, and am sticking with 2901 and 1.70v = 1.26v I guess.

[Pawelek]

Hello !

I'd like to buy Xeon 3220 - (i guess it is an equivalent of Q6600). Still I have some doubts whether it fits my motherboard. I'm going to update my BIOS to the latest version but I'm not quite convinced everything will be fine then. Can someone give me some support?

Thanks
Pawelek