ASUS P5W DH - Problems + Fixes Thread

[BigMonty]

First off, a 45nm CPU in a 1.02G board is stuck for now with AUTO or nothing. Even then, I have to cut back the FSB from 333 to 290 so my Q9450 doesn't crash within minutes of booting.

Do you CRASH at AUTO, or is it just that you'd prefer to set the voltages?...

When i set the FSB through the AI Overclocker app and leave it on AUTO or set to one of the lower available voltages the bios always reports 1.4V on reboot and crashes.

This seems like something Asus should fix because it reports one setting in the gui and applies a different one after reboot. I'm going to hammer on them from that angle, it seems dangerous to the chip and other hardware if it's not applying the setting it says it should.

[mikepaul]

This seems like something Asus should fix because it reports one setting in the gui and applies a different one after reboot. I'm going to hammer on them from that angle, it seems dangerous to the chip and other hardware if it's not applying the setting it says it should.

You could call 812-282-2787 and ask for help, but so far I've gotten eager techs who take my info and never get back to me. Not even to tell me they are making no progress...

[tomcug]

Older revisions of P5W DH simply can't set voltages lower than 1,4V with 45nm CPUs. So you can't use any software to set that voltage below 1,4V. But I've got one idea. You should leave that voltage on AUTO and use pencil voltage mod to get the voltage you want. I think someone has to try it. You can find some instructions here:

[Yolanda]

Better to set 1.4V than hardware mod. Such voltage isn't high for 45nm. Many work at 1.5V+.

[tomcug]

Yes, 1,4V is still safe votage for 45nm CPUs but many people are afraid of setting this voltage because they think it's too high. I just don't know why.

[mikepaul]

Yes, 1,4V is still safe votage for 45nm CPUs but many people are afraid of setting this voltage because they think it's too high. I just don't know why.

One reason: no $300 here to throw at another Q9450 if mine turns out to be touchy about high voltage...

[tomcug]

Mikepaul, it's impossible to kill 45nm CPU with 1,4V. By the way, my friend's fried his E7200 with too high voltage and he had no problems with RMA. I've already run my E7200 with 1,45V for few hours and it's still working . So I think you should try 1,4V for your CPU to see if it'll be stable.

[Yolanda]

One 45nm E8400 was used at 1.8V 24/7 and after a month its overclocking down from 4.3 to 4.1GHz. Later it was set to 1.6V and works about a year without problems.
1.8V is very high, but 1.55V is nominal for some 45nms in retail high end computers with factory overclock.

[crishan]

The funny thing is that even though I have a revision 1.04G board,
I do not get anything other than VCore set to auto in the 2801 Bios.

I have a E0-Stepping E8400 that only gets a Core Voltage of 1.168 volts,
which I find too low. According to Core Temp, my processor has a VID
of 1.2125 volts, which the board doesn't even set

So far, no matter how often I performed a CMOS clear, took out the
battery etc., I do not get the manual VCore option

[bodka]

I recently got a Quadcore QX9650 C0 step. I reinstalled Vista 64 after swapping it with a core 2 duo.

now the problem is that with only 2 cores enabled, Super PI completed 32M without errors. When I enable all 4 cores Super PI fails.

I am sure that my RAM does not have any issues. I am using the latest 2801 BIOS.

Btw, in windows XP without OS reinstall, I am getting BSODs randomly.

Any help?

[bodka]

I recently got a Quadcore QX9650 C0 step. I reinstalled Vista 64 after swapping it with a core 2 duo.

now the problem is that with only 2 cores enabled, Super PI completed 32M without errors. When I enable all 4 cores Super PI fails.

I am sure that my RAM does not have any issues. I am using the latest 2801 BIOS.

Btw, in windows XP without OS reinstall, I am getting BSODs randomly.

Any help?

Found the issue! It was to do with FSB speed. Reducing it by 10 worked.

But I have another issue.. I dont know if it is a issue.

HW monitor shows CPU temp as 17C for one of the cores! Can it be so low???
Max temp for was 25C for another core.

[Falkentyne]

What FSB were you running?

I found the QX's to clock the best if you leave the FSB close to 333 and tweak the multiplier with a little FSB changing. 380x10 (1.65 mch) was not prime stable even for 15 minutes of the small fft test even at 1.4v, though its worth noting that higher vcore did help stability slightly as i started with 1.3625v, while 333x12 passed a full pass of small fft prime (8k-64k-about 3.25 hours) with 1.3725v (1.368v shown in cpu-z), with the vdroop pencil mod. (200 mhz higher at lower vcore, take note of that).

to compare, 400x10 was a complete no POST at 1.4v. Oddly enough with both a 1.02G P5W (using a X6800 dual) and the 1.04G with qx9650, I couldn't reach 400 fsb, but I didn't try 1.75v mch. The 1.02G wouldnt even do 370x10 on the X6800 unless i pushed the vcore to 1.65v (!). But 340x11 posted at 1.55v.

Someone mentioned a long long time ago that they 'felt' that the board, when set to 370x10, was actually initially posting (before bios init finished) at the default multiplier (11), which meant the cpu was momentarily at 370x11 (4.05 ghz) then setting it to at the end of the BIOS init, which explained why 370x10 posted at 1.65v but not at 1.55v, while 340x11 posted at 1.55v. Possibly the CPU has to power up and accept signals at the FSB and its stock multiplier, before it can be told to lower it by the BIOS signal (?)

I never bothered to test something like this, but it would have been easy enough to set it to 370x11 to see if it still needed 1.65v for POST. I no longer have the x6800 or the 1.02 board though....

[Falkentyne]

The funny thing is that even though I have a revision 1.04G board,
I do not get anything other than VCore set to auto in the 2801 Bios.

I have a E0-Stepping E8400 that only gets a Core Voltage of 1.168 volts,
which I find too low. According to Core Temp, my processor has a VID
of 1.2125 volts, which the board doesn't even set

So far, no matter how often I performed a CMOS clear, took out the
battery etc., I do not get the manual VCore option

Does that Bios support E stepping CPU's?
Perhaps only the c0 stepping is supported?

and there's never any need to take out the battery. All that's needed after a flash, is (after u reboot, and stop the boot by going into the BIOS), is a power off, unplug power supply plug, then cmos clear.
An alternative for a TRUE pure cmos clear, is a FLOPPY/Flash drive award based bios flash, without using crashfree tool to do it.

I know when you first do a bios flash, you have to select some settings and "save and exit", then when you go back in, all the voltages are unlocked. If you've done this in the past, you would know if that's simply not working anymore or not....

[bodka]

What FSB were you running?

.

I was running FSB at 333 with Auto volatage for MCH. Also my Vcore starts only at 1.5V

What max MCH voltage is safe on this mobo?

[bodka]

Does that Bios support E stepping CPU's?
Perhaps only the c0 stepping is supported?

and there's never any need to take out the battery. All that's needed after a flash, is (after u reboot, and stop the boot by going into the BIOS), is a power off, unplug power supply plug, then cmos clear.
An alternative for a TRUE pure cmos clear, is a FLOPPY/Flash drive award based bios flash, without using crashfree tool to do it.

I know when you first do a bios flash, you have to select some settings and "save and exit", then when you go back in, all the voltages are unlocked. If you've done this in the past, you would know if that's simply not working anymore or not....

How can a stepping difference cause issues with Mobo?

according to Asus web site only C1 stepping of Qx9650 is supported. But I use C0 and it seems to be OK ( with a tad bit lower FSB ).

[tomcug]

First of all, you're lucky because you've got a CPU with unlocked multiplier. You should try to make 333MHz stable. I think increasing FSB Termination Voltage to 1,3V and MCH Voltage to 1,65V should help you. After that you can overclock your CPU with multiplier. By the way, in your motherboard voltage really starts at 1,5V? I heard that in older revisions voltage starts on 1,4V but not on 1,5V.

[Stealth187]

I've had several cpus in my 1.02g board (e6700, e6750, e2180, e2220) and could never obtain a high fsb. 300-310 is about the most I can get stable...believe me I've exhausted all possible cpu/chipset/jumperfree setting combinations to no avail.

Now I have an e8500, and its the same thing. It will only post at its nominal speed (3.16ghz with 333fsb) if I leave the 'AI Overclocking' option to auto, which is undesirable as I like to manually set everything like voltages and ram speed etc.

If I set 'AI Overclocking' to manual, put the fsb to 333 and leave all other jumperfree options still on auto, it refuses to post. So how the hell is the auto setting making 333 fsb stable? Anyone experienced a similar issue?

Also, once I change 'AI Overclocking' from auto to manual, I have to use the 'load default settings' option in bios to post again - simply changing the option back to auto doesn't work.

[bodka]

I've had several cpus in my 1.02g board (e6700, e6750, e2180, e2220) and could never obtain a high fsb. 300-310 is about the most I can get stable...believe me I've exhausted all possible cpu/chipset/jumperfree setting combinations to no avail.
.

Same here. With my mobo, all I can do right now is 320 Max. Beyond that if I disable 2 cores, I am able to go past 333. What a trade off

I tried various combinations of FSB voltage / MCH v , no avail.

Having a unlocked multiplier is one thing, but getting the CPU with that totally free is...

[Stealth187]

Just an update. Switching ram to higher speed Crucial and thus being able to use a higher fsb:dram ratio (4:5) has allowed me to set almost everything manually.

If I disable speedstep and thermal control, the cpu only runs at 300-400mhz, but I can still adjust vcore with it on

Now I can do 368 fsb by raising FSB Termination and MCH voltages up a notch. anything further and memtest freezes about 10mins in. Enabling SPD solves this, but its setting weird timings of 5-2-2-18 and its only good for another 10mhz so not much point.

[bodka]

Another finding..

Initially I had Vcore in BIOS starting from 1.5 . I then used AI booster to set a voltage below 1.5v. After this, BIOS showed voltages starting from 1.2 volts.

[mikepaul]

Well, AI Booster displays 1.19V but the vCore setting I picked of 1.375V is still there.

I'll wait for the crash before abandoning it...

[G40RoCkEt]

Well people.

My old E8500 run stable at 370FSB with 1.8v in memories and all voltages in AUTO. I can get 3.5ghz.

Now, i change the cpu to a E8600, get 400FSB (3.8ghz).

I have to get more test. I guess that when i select 1.40v in vcore cpu can get 100% stability.

Tomorrow i run some more exaust tests.

[mikepaul]

Well, AI Booster displays 1.19V but the vCore setting I picked of 1.375V is still there.

I'll wait for the crash before abandoning it...

Oddly, it didn't crash, but even though Orthos ran for 3+ hours OK I experienced some corrupted RAR file downloads that only became uncorrupted when I downloaded them again after going back to my original settings. One corrupted file occasionally I can understand, but not three that didn't happen again.

And the BIOS settings looked a touch odd. While I set AI Booster to 1.375V, I saw lower values displayed by CPU-Z but the BIOS was set to 1.45V since that was the lowest value available. If I get brave later I'll retry 1.45V with no Booster just to see what kind of results I get. I still don't WANT to just leave things at 1.45V though...

[Falkentyne]

Just wondering something:

Anyone know if the C1 stepping of QX9650 is supported (like the C0 steppping is?) Or the QX9770...

[secretworld]

Oddly, it didn't crash, but even though Orthos ran for 3+ hours OK I experienced some corrupted RAR file downloads that only became uncorrupted when I downloaded them again after going back to my original settings. One corrupted file occasionally I can understand, but not three that didn't happen again.

And the BIOS settings looked a touch odd. While I set AI Booster to 1.375V, I saw lower values displayed by CPU-Z but the BIOS was set to 1.45V since that was the lowest value available. If I get brave later I'll retry 1.45V with no Booster just to see what kind of results I get. I still don't WANT to just leave things at 1.45V though...

Corrupted rar files are in my experience symptoms of the on-board network not liking the overclock. Here I can overclock my q6600 to fsb 350 but lage rar files over the local network are unusable. Around 340 things are much better. I tried everything I could think of raising all kind of voltages, but never found anything that improved matters.