• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

few questions on a p5wdh deluxe

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

jblindy

Registered
Joined
Feb 6, 2005
Is there any way that I can get the northbridge temperature on a p5wdh deluxe?

Right now I've got a e6300 oc'ed to 3.15ghz at a 450 fsb, fsb/memory ratio is 4:3 at 337.5 mhz. To get stable at this speed I had to max out the mch which is 1.85v.

I know that I've got more headroom in the cpu because vcore is only at 1.36. I should have more left in the memory as well at a 4:3 divider. Running 1:1, I can only get to 410 fsb no matter what I do.

Any recomendations or comments? I really want to know how to find the temp for the northbridge.
 
Short of zapping it with a temp gun, I do not believe there is a sensor onboard to measure the NB temp. I am not sure that you need that kind of voltage for your clocks though....
 
For a CPU clock of 450, you shouldn't need anywhere near 1.85v to the NB (like Stilletto mentioned)... 1.6-1.65v should be plenty up to about 465. If you have a setting for SB Core / CPU PLL Voltage, you'll need to raise that to about 1.75v... and no more for long term use. If you also have a setting for CPU VTT Voltage, that should also be set to about 1.6v... and no more for 450. So play w/ your voltages (and drop the MCH voltage), RAM timings, and different ratios until it's stable w/ only 1.6-1.65v to the NB.
 
If I could get it to post or remain stable with lower voltage on the mch I would be tickled pink.

I'll post some screenies of my settings.

If I set the dram to 1:1, I post but crash as windows is loading. If I set it to auto it won't even post. I've set memory voltage to 2.2 which does nothing and I'm afraid to set it any higher.

If I try to manually set memory timings it won't post.





I want to add that I could run 3DMark 06 all the way through with the mch at 1.75v, but about two hours into orthos the computer froze and I had to hard boot. I ran orthos for about four hours at mch 1.85v with no problems.

Also, the max voltage available for the ich is 1.20. I tried setting the ich at 1.20v and lowered the mch to 1.65 and it hung on post. I upped the mch to 1.75 which is where I'm at right now.
 
Last edited:
What memory do you have installed, how much, and what are it's timings? To start I'd probably bump the Vcore to 1.4-1.45v, FSB Termination Voltage to 1.5v, lower the MCH to 1.6v, ICH Voltage to 1.070-1.12v, and set Dram Timing by SPD to Disabled (set your RAM timings manually).

Edit: Set PEG Buffer Length to Short, Link Latency to Slow, and PEG Link Mode to Slow (start at Slow, and move to Faster).
 
Last edited:
What memory do you have installed, how much, and what are it's timings? To start I'd probably bump the Vcore to 1.4-1.45v, FSB Termination Voltage to 1.5v, lower the MCH to 1.6v, ICH Voltage to 1.070-1.12v, and set Dram Timing by SPD to Disabled (set your RAM timings manually).

Edit: Set PEG Buffer Length to Short, Link Latency to Slow, and PEG Link Mode to Slow (start at Slow, and move to Faster).

I'll give it a shot and see what happens. But in the past I have been unable to enter post if I set the ram timings manually. The only way to post was to set it by spd.

My memory is 4 gig of OCZ XTC Platinum Rev.2 pc2-6400. The rated timings are 4-4-4-15 at 400mhz.
 
Lower the Vdimm to 2.1v, and change only the four primary timings (tCL-4, tRCD-4, tRP-4, and tRAS-15) and leave the rest of the timings on Auto for the time being. Since you're running 4GB, you might want to start w/ 5-5-5-15, and then tighten them once your system becomes stable.
 
Well I changed the settings you suggested one at a time. Right now I'm at vcore 1.425v, fsb termination 1.5v, mch 1.65v, ich 1.20v, peg buffer short, link lat. slow, and peg link mode slow. Looking good so far.

When I made sure that I could boot into windows at those settings I focused on the memory. This mobo refuses to post unless spd is enabled. I will try to up the vdimm to 2.1 and 2.2 to see if that helps.

Thanks for the help so far.
 
With 4GB installed and disabling SPD, while setting the timings to the default of 4-4-4-15, the timings are probably too tight regardless of supplying the RAM w. 2.2v

Start w/ 5-5-5-15 at 2.1 Vdimm, and leave all of the sub-timings on Auto for the time being. With only 2GB installed, you'd have a better chance of running the default timings at a lower voltage. But since all four slots are populated, this puts more strain on the NB... which in turn requires more MCH voltage (just not more than 1.6-1.65v).
 
With 4GB installed and disabling SPD, while setting the timings to the default of 4-4-4-15, the timings are probably too tight regardless of supplying the RAM w. 2.2v

Start w/ 5-5-5-15 at 2.1 Vdimm, and leave all of the sub-timings on Auto for the time being. With only 2GB installed, you'd have a better chance of running the default timings at a lower voltage. But since all four slots are populated, this puts more strain on the NB... which in turn requires more MCH voltage (just not more than 1.6-1.65v).

No luck, I tried the vdimm at 2.1, 2.2, 2.3 and the max of 2.4v and tried the loosest settings of 6-6-6-18 and it would still hang at post requiring a full power off to restart.

So would 2 2gig dimms be better? I'll pull out two sticks and see how that fairs.

Orthos just stopped with a fatal error at 5 minutes into the blend test.


What BIOS are you running?

The latest...2406.
 
Well, I pulled two sticks of ram and booted into windows with these settings.

MCH 1.65v, ICH 1.2v, fsb termination 1.5 (what does this do actually), vdimm 2.0v, timings set by spd.

Fine so far. Started orthos and I got a fatal error 3 minutes in.

I'll tinker around some more.
 
I used to have issues with my P5wDH with booting with manual memory timings unless I ran a different divider...don't run it 1 to 1. (If you can)
 
What is the temperature variation from the northbridge chip to the base of the heatsink? Any ideas or guesses?

I went out and bought a laser temp probe from the local computer store and if I aim between the fins to the center of the base I get a max reading of 42.3°C. Orthos errored out about 15 seconds after I took the reading.

It's gotta be the northbridge or memory or a combination of both that is causing the instability. I just ran the small fft cpu test for four hours and it was still going strong when I stopped it. But as soon as I run a blend test ot the large fft test it errors out. Sometimes it's within a matter of minutes and other times it will take an hour or so.

Off to tinker some more. I think I've reached the limits of what this mobo can do.:confused:
 
What is the temperature variation from the northbridge chip to the base of the heatsink? Any ideas or guesses?

I went out and bought a laser temp probe from the local computer store and if I aim between the fins to the center of the base I get a max reading of 42.3°C. Orthos errored out about 15 seconds after I took the reading.

It's gotta be the northbridge or memory or a combination of both that is causing the instability. I just ran the small fft cpu test for four hours and it was still going strong when I stopped it. But as soon as I run a blend test ot the large fft test it errors out. Sometimes it's within a matter of minutes and other times it will take an hour or so.

Off to tinker some more. I think I've reached the limits of what this mobo can do.:confused:

Actually, I think you are right...I traded out my P5WDH for a P35 chipset and gained 100mhz FSB. I remember that now.:eek:
 
Yeah, I've been doing some reading and from what I gather the p5w doesn't like it when fsb gets up to far past 400mhz.

Right now, I've got the mch at 1.85v with a 80mm fan blowing directly on it. FSB is at 460mhz and orthos has been running now for 28 minutes. According to my laser probe the northbridge heatsink is reading 39°C, but I have no idea how to convert this to actual northbridge temps. According to the intel spec sheet the max temp for the i975 is 103°C.

The last stable oc was at 3.08 ghz with a 440fsb. This was with vdimm at 1.95, vcore at 1.3875, and mch at 1.65, ich at stock. This is my brother-in-laws computer, so I think that this is where I'm going to set it. I know it is 8 hours orthos stable at these settings.
 
thats a good oc for a 975 chipset im only running 320 24/7 its stable at 340 but i have to bump the cvore up and i dont like to 24/7 so i just run it 340 for gaming
Rich
 
Here's an interesting thread on the relation between the fsb and the mch on the p5w dh deluxe.

http://www.xtremesystems.org/forums/showthread.php?t=105940&highlight=MCH

This pretty much correlates with what I have been experiencing. Right now I have the fsb at 460 and the vmch at 1.85 which is the max available in bios. Orthos has been running now for almost 3 hours.

My core temps are at 59°. According to my thermo-laser the base of the northbridge heatsink is 37°, that's with an 80mm fan blowing across the stock heatsink. I would really like to know what the exact temperature of the chip is. Does anybody know of a way to find a closer approximation of the temp of the chip?
 
Here's an interesting thread on the relation between the fsb and the mch on the p5w dh deluxe.

http://www.xtremesystems.org/forums/showthread.php?t=105940&highlight=MCH

This pretty much correlates with what I have been experiencing. Right now I have the fsb at 460 and the vmch at 1.85 which is the max available in bios. Orthos has been running now for almost 3 hours.

My core temps are at 59°. According to my thermo-laser the base of the northbridge heatsink is 37°, that's with an 80mm fan blowing across the stock heatsink. I would really like to know what the exact temperature of the chip is. Does anybody know of a way to find a closer approximation of the temp of the chip?

That is a very nice thread, thanks for sharing.

You don't really need to know the temperature, but if you want you can always get a thermal probe and stick it to the underside of the heatsink just next to the area that the core of the NB touches
 
Back