Need help with a Asus P5WD2

Arthalen

Arthalen

Arthalen

Associate
Joined
20 Apr 2006
Posts
2,037
Location
Leeds, UK
Not an -E board, or a premium, its just the normal P5WD2.

Specs in my Sig.

Clocked upto 4Ghz quite happily on a low voltage, but past 24 hours has started to show a little instability.

Tried to raise the Vcore to get more volts into it, but I've had a very weird pattern happen here.

Btw, I've used the 'pencil trick' on the board to stop the board from Vdrooping.

At 200FSB, max Volts I can spec is 1.275 (1.375 actually supplied).
At 195FSB, max Volts I can spec is 1.300 (1.400 actually supplied).
At 190FSB, max Volts I can spec is 1.325 (1.425 actually supplied).
At 180FSB, max Volts I can spec is 1.350 (1.450 actually supplied).

Weird, no ?

If I spec a higher voltage, then the board trips out after trying to POST. I then have to disconnect power from the board, reconnect and power back on. Crash-Free bios then posts up and allows me to change settings.

1) I've tried going 'over-the-top' with the voltage to 1.45+ actual voltage at 200FSB and this has not stopped the issue.

2) I've also tried lowering the memory right down to DDR2-400 at stock volts and this has made no difference.

3) I've also tried the FSB Term Voltsage , MCH Voltage and ICH Voltage from Lowest and then Highest settings - neither made any difference.

This is very wierd and doing my nut in as I can SuperPi Bench and play games at 4ghz at 1.375Volts actual for about an hour before it trips up. I need to get more volts to the CPU to get it stable....

Help me Ocuk, Your my only hope ;)
 
Check for Throttling - ASUS boards are notorious for under-reading the temperature at the die. Also - check the PWM temperature isn't tripping it out. Generally, the chip will start to throttle rather than cause the system to fail, but the PWM temperature will shut the system down very fast if it's too high. Because you have dual cores, the PWM is working VERY hard indeed on a overclocked 805 and usually runs at least as hot as the chip itself, if not hotter. This very high temperature can cause the thermal pad (they always seem to be fixed on with a thermal pad) to break up or loosen and I've found that stripping that off and reseating the heat sink with AS5 works well, as does fitting an active cooler like the Thermalright one.

There is one further possibility - you've damaged the chip running it hot over a prolonged period - and it's starting to exhibit symptoms of imminent death.
 
1) Definitly not throttling, because when I boot up at the Vcore available for 4 Ghz it does not throttle at max temp - just occasionally cuts out on stability with a hard shutdown.

2) This seems like the likely culprit. There is no airflwo to the fanless heatsink on this area and also, the retention bracket that holds my waterblock actually sits over one of the push pins on the heatsink for this area. This could be pushing it back and thus the heatsink is making poor contact. However if poor contact is to blame I would have expected poor stability at any OC speed, not just above 3.7Ghz.

3) I don't think the cpu/motherboard is failing as I've backed off to a OC of 3.7 ghz and my previous settings and everything is hunky dory fine.

I am going to back off from my 4Ghz overclock at the moment as I'm currently doing some research into the voltage requirements/max voltages that the CPU/Chipset/Memory can handle.

I also need to look into cooling possibly the Northbridge and the PWM area on the board. since swapping over to water the PWM area no longer gets an air flow supply, and sits there with a fanless heatsink - which Is why I believe that problem 2 could be the cause.

*Sigh* :rolleyes: Will need to look into some more active cooling for the board methinks.

Can anyone else post their thoughts on this issue ? Or anyone have any ideas for actively cooling the PWM/Northbridge.

Oh, and what the hell is FSB Termination Voltage all about ????????
 
The FSB termination voltage lets you run chips with variable Vcores in the same motherboard. This was very handy when you wanted to run a S478 Pentium 4 Mobile chip in a desktop motherboard. The chips would run at one voltage is power saving mode and another voltage in performance mode. The FSB Termination voltage marks the point at which the CPU is throttled back to it's lower power saving mode (it did this by halving the FSB!). If this option was not present then the mobile processors would run at half-speed all the time. I can't see any use for it in your application.

If money is no object that I can thoroughly recommend the Thermalright Active chipset cooler - but it is mentally expensive at almost £25.

It completely sorted out a PWM issue I was having on an Abit LG-81. It is quite tall and narrow, which is a good thing if you're using an unpright CPU cooler, but a bad thing if you're using a flower type (Zalman/Blue Orb). With your water cooling I don't suppose it will make any difference.
 
Could you give me the exact name of the Thermalright PWM kit you recommended ?

I don't think OcUK sell it.

I'm very interested in getting the VR & PWM area's cooled on my board to help with the stability.
 
It's the Thermalright NB-01 although they have a new one new (even more expensive) called the the HR-05.
 
You must log in or register to reply here.
Back
Top Bottom