ASUS P5W DH - Problems + Fixes Thread

[Grendel66]

fc /b says "FC: no differences encountered". Same binary.

[lawrywild]

Thought so

[nifty]

Hi Lawry, thanks for clarifying that for me

[bichi]

AJC Diff v1.90 File Compare Tool

- beta2004.rom *Dated: 03/07/2007
- P5W-DH-ASUS-Deluxe-2004.ROM *Dated: 03/07/2007

"no differences found"

[BenchZowner]

try comparing 1901 with 2004.
Changes:
version number change
other changes : none

typical asus

[MonkY]

Hello everyone. It's very nice to see such a great thread and so many good advices. Unfortunatelly for me, none of them helped. I have a P5W DH Deluxe, one of the early revisions (with max 1.65V mCH), and i have big big trouble with running memory even at 800Mhz. At first i've started with an E6600 which goes at 3.33Ghz 100% stable with 1.425V, and Corsair TWIN2X2048-6400C4 2048 MB DDR2 / 800 MHz / 4-4-4-12. Unfortunatelly, whatever i tried, those memory sticks couldn't go to anything else but 740Mhz in 4-4-4-12, even at 2.2V. I've tried with changing the ratio to the CPU, so i should be sure it's not the CPU's fault. The memory was a no go. Then, i started mounting some FAN's over the southbridge and northbridge, removed silly Asus paste with Arctic Silver 5, then tried again. No changes. Then, these days, i've decided to change the memory. I bought OCZ DDR2 PC2-8000 Titanium Alpha VX2, wich should go to 4-4-4-15 at 2.3V even in 1000Mhz. Well, with this RAM (2*1024), my sistem doesn't even boot at 920Mhz, everything else over 800Mhz just crashes instantly in Prime. So i am now sticked again to 370Mhz.

I don't know what else to do. I would like to go and change the motherboard, but unfortunatelly, i've done changes to the motherboard that ruins my warranty (those vents on chipsets).. Please, let me know what would be the next step i should do... or what would you do in my shoes. Thx in advance, and have a nice day !

[sesdave]

I have Corsair xms and have no ptoblems 1:1 at 400fsb but I do have later rev board. . Look at first post on this thread for settings advice - some people hit a wall at 370 FSB. Try boot with SPD settings and then use memset to tighen your timings. Disable anything your not using in the BIOS - memory should only need stock volts if you are not o/c it. I aslo recently posted a pencil mod if you want to up MCH volts. Make sure your CPU is getting enough juice - try it at 1.45. I need 1.5 for 400FSB and a lot of ther CPUs need more.

[AHolland]

Ho MonkY,

Your problem "feels" similar to mine in that certain settings that should work, do not. I can ramp up my memory to 900MHz from 800 and it's all stable. Then if I ramp up my CPU past 3.0 GHz, with any memory setting, including 667 MHz, it shows a memory error.

Perhaps someone more knowledgeable can help interpret what I am getting from ORTHOS. If I run just the CPU stress test, I can hit higher CPU values. My problem comes when I run the blended test. Under the blend test I can ramp up my memory past 900 as long as my cpu is at 3.0 GHz or less. As soon as I move the CPU FSB from 333 to say 340 it crashes and burns. Soemthing to the effect of "rounding error of 0.5, expecting error of 0.4", or something like that. I have tried a FSB of 350, 360, 370 just to see if anything would work at those levels. I'm retesting using a cpu voltage of 1.5 to see if that helps, but so far, no go. I'm going to also re-run the tests using "Crazyea"'s settings. Those settings are generally lower in most voltages to the chipsets. Everything is worth testing if other people are finding success with those settings.

From where I sit, the fact that the memory can be run quite a bit higher than what it is rated at, so it is not the memory causing my problem. Can anyone say with some confidence what my error shows? I could type it word for word if necessary, although I suspect my error type is fairly common.

Another quick question: What is a resonable temperature ceiling to use for the E6600. Someone somewhere suggested that the idle to full load temp diff should not be more than 26C....So in my case, my idle temp is 35C and my full load temp is 45C so my delta or difference is only 10C. I Could run as hot as 35 + 26 = 61C (gee mom, I did remember that math stuff after all). 61C sounds low from what I would expect. I would think something closer to 75C but I have not gone to the Intel web site to suck out the specification.

[lawrywild]

I got two kits of awesome ballistix it seems.. they do like 1000 4-4-3-4 with 2.3v easily but my dam chipset can't handle ddr2-1000 memtest will run without errors then just freeze due to bloody 975x...

At the moment I'm at 860mhz 4-4-4-4 2.00v 4x1GB, anyone got any tips for stabilising at ddr2-1000? I'm trying to get 333fsb 2:3 stable but mch 1.65v doesn't help from auto, nor does vfsb. God, just want to get a DFI 965 dark now..

[GFORCE100]

I got two kits of awesome ballistix it seems.. they do like 1000 4-4-3-4 with 2.3v easily but my dam chipset can't handle ddr2-1000 memtest will run without errors then just freeze due to bloody 975x...

At the moment I'm at 860mhz 4-4-4-4 2.00v 4x1GB, anyone got any tips for stabilising at ddr2-1000? I'm trying to get 333fsb 2:3 stable but mch 1.65v doesn't help from auto, nor does vfsb. God, just want to get a DFI 965 dark now..

Turn off HyperPath 3.

[lawrywild]

Turn off HyperPath 3.

Was that a joke?

[bichi]

Prompted by GFORCE100, tried x9 multi at FSB of 333, 330, etc.
- memory set at 1000, 1100 etc.
- XP SP2 booted successfully, but fails various memory and loading tests
- NB temp: 105f/40.5c MEM temp: 90.5f/32.5c, measured with Fluke Mini62 IR thermometer
- picture is of multi: x9, FSB: 333 attempt....

So it's back to multi: x10 @ 300 FSB, memory @ 1000/500, 3:5 for me

lawrywild,
- ever figure out what was causing your DL burn problem?

[GFORCE100]

Was that a joke?

No it's no joke. I've been running the Ballistix you have for months now at 1000MHz speeds at 4-4-4-4 now and you need HyperPath disabled if it's to be stable. I always keep it at 2.4V too. 4-4-3-4 will give you memtest errors sooner or later but 4-4-4-4 is stable.

[lawrywild]

No it's no joke. I've been running the Ballistix you have for months now at 1000MHz speeds at 4-4-4-4 now and you need HyperPath disabled if it's to be stable. I always keep it at 2.4V too. 4-4-3-4 will give you memtest errors sooner or later but 4-4-4-4 is stable.

Seen, but

FSB Wall at low fsb around 300 - Disable Hyperpath 3 in the Chipset section of the bios. [Credit - lawrywild]

haven't used Hyperpath 3 since the day I received this board.

@bichi:

Nope, and don't want to waste anymore discs...

edit: @ gforce again: do you have a fan over your sticks @ 2.40v? and what timing is your Write Latency Time at? 4-4-4-4-x

[GFORCE100]

Prompted by GFORCE100, tried x9 multi at FSB of 333, 330, etc.
- memory set at 1000, 1100 etc.
- XP SP2 booted successfully, but fails various memory and loading tests
- NB temp: 105f/40.5c MEM temp: 90.5f/32.5c, measured with Fluke Mini62 IR thermometer
- picture is of multi: x9, FSB: 333 attempt....

So it's back to multi: x10 @ 300 FSB, memory @ 1000/500, 3:5 for me

lawrywild,
- ever figure out what was causing your DL burn problem?

So you did turn off EIST didn't you? Did you give the northbridge 1.65V in the BIOS (not always necessary)?

[GFORCE100]

Seen, but

FSB Wall at low fsb around 300 - Disable Hyperpath 3 in the Chipset section of the bios. [Credit - lawrywild]

haven't used Hyperpath 3 since the day I received this board.

@bichi:

Nope, and don't want to waste anymore discs...

edit: @ gforce again: do you have a fan over your sticks @ 2.40v?

You may have not used it but if it's set to AUTO then that's your problem, it should be DISABLED.

[bichi]

GFORCE100,
- Duh! - EIST/C1E = off
- and just for that, Hyperpath = off too, (double DUH) - LOL!
- didn't try increasing NBvdc since summer is around the corner here...
- increase in memory performance was about 2%, so rather than risk heat, I'm opting for higher multi and slower FSB...
- was a bit of fun trying, though....

[AHolland]

Hi Bichi (thought it was Lawrywild but I was wrong),

What is that temp program you are running. It looks cool. I also noticed that it showed hard drive temps as well. I have a program called speedfan. It gives some good info, but not enough for my testing.

As a side note, does anyone know what acceptable HD temps are? I know that HD failures are closely linked to temps. My case (P180) allows me to separate my HD and MB from each other. I then run a fan through the hard drives (4 of them), and then through the power supply and out. I cannot make any changes to the power supply fan, but there is an inline fan that I can ramp up. I don't like to run allo my fans as fast as they can because I do some sound recording and you can hear the computer in the background if the fans are turned up. Basically I like quiet.

It's all a balance, fans, sound, temperatures etc.

[lawrywild]

You may have not used it but if it's set to AUTO then that's your problem, it should be DISABLED.

I guess what I'm trying to say is it is NOT hyperpath 3. Trust me on that.

[lawrywild]

Hi Lawrywild,

What is that temp program you are running. It looks cool. I also noticed that it showed hard drive temps as well. I have a program called speedfan. It gives some good info, but not enough for my testing.

As a side note, does anyone know what acceptable HD temps are? I know that HD failures are closely linked to temps. My case (P180) allows me to separate my HD and MB from each other. I then run a fan through the hard drives (4 of them), and then through the power supply and out. I cannot make any changes to the power supply fan, but there is an inline fan that I can ramp up. I don't like to run allo my fans as fast as they can because I do some sound recording and you can hear the computer in the background if the fans are turned up. Basically I like quiet.

It's all a balance, fans, sound, temperatures etc.

Hi, did you actually mean me or Bichi?

Bichi is using Motherboard Monitor 5 if you meant him. If you did mean me, what image are you referring to?

[M3kk]

Today i wanted to try my E6700 in P5W DH , rev 1.04g, but after the power on i received the beeps : biiiip ( long ) pause bip-bip-bip.

The system booted up fine ( i see from the keyboard ), but no signal on monitor.

I tried with 3 kits of rams ( with a single dims too ), with 2 vga`s ( 8800gtx, 7900gt ), vga in another PCI-E slot, but nothing .

Now ~week when the board makes this "beeps", after i put one of 8000UL dims in the firs black ram slot, the system stared up fine, but now this method with single DIM doesnt helped me .

Any suggestions?

PS : Sorry for my english, i hope you will understand.

[bichi]

guess AHolland needs new glasses, after staring at his BIOS settings for hours - LOL!

Linkage to MBM and required Core plugin and setup instructions:
http://www.thetechrepository.com/sho...rboard+monitor

[lawrywild]

@M3kk: Have you got a POST code analyzer card?

[AHolland]

Yes, I finally got directed to MM5. Thanks Guys.

Yes you do g odizzy from going back and forth in the bios trying to figure out how a system can be perfectly stable at a FSB of 333 and then die at 335. I was expecting a bit of a transition between totally stable and quite unstable. I still think I've missed something, particularily because I am also getting low overclockability...3.0 Ghz. I tried ramping the vcore from stock to 1.55 with no apparent difference. I also have dropped some of the other voltages to see where they affect stability. So far they look like they will be close to stock. It all takes a lot of time.

[M3kk]

@M3kk: Have you got a POST code analyzer card?

No .