FSX & SLI

[Nick N]

Mmmm there's only one rectifier bridge in every power source(~+-~)or (+~~-) IF only If the problem shows again try to connect your new fan or all of them in another power souce ,an AT perhaps an oldie

Yes, you can run 2 PSU's in tandum without any special wiring as LONG as you -do not- under any circumstances connect the fans running off the 2nd PSU to the motherboard in any way. Also, you CAN NOT run any drives or other devices from the 2nd PSU without a special wiring harness that runs the 2 supplys in tandum and BOTH units MUST be equally ATX version spec'd

You CAN

[Nick N]

[quote]Ah yes.... my infamous cold boot starts

[alrot]

I know Nick, the only thing is Malfuntioning seems to be my english :-/ , I think it the season or the approach of the moon , I know dude! ,I was trying to tell our mate Black ZR 1 that if the problem persist /insist /remainds ,..to try to add JUTS FOR THE FAN AND WITH SEPARATE PLASTIC BUILDS only to add another Power supply, you know maybe that mosnter of fan  could be the issue,it work with 12 V but maybe 10 amps..lol  (Personaly I think is an exageration unless the CPU its made of Tungsteno.. )But I would place one just to look the interior of my PC like a JET plane

Hey Nick is it normal the cpu Fan at 4800 rpm? ,My both PCs the old P3 and the P4 (according to speedfan)are working with that speed? and Why the Fan of the P4 is so noisy? ,looks like you know those european cars (Reanault)its like its going to burn, instead my oldie I can't even hear it ..

[Nick N]

[quote]I know Nick, the only thing is Malfuntioning seems to be my english :-/ , I think it the season or the approach of the moon , I know dude! ,I was trying to tell our mate Black ZR 1 that if the problem persist /insist /remainds ,..to try to add JUTS FOR THE FAN AND WITH SEPARATE PLASTIC BUILDS only to add another Power supply, you know maybe that mosnter of fan

[justpassingthrough]

[quote]Ah yes.... my infamous cold boot starts

[Black ZR-1]

That unit will drive a sherman tank of a tower and then some.

It also weighs like one!  ;D

[justpassingthrough]

That unit will drive a sherman tank of a tower and then some.

It also weighs like one!

[Black ZR-1]

Thats nothin,., try this one: http://www.pcpower.com/products/viewpro ... how=T1KWSR

With that thing you can power the flux capacitor. Who needs lightning?

[Black ZR-1]

I left the sim on while I ran out.




I came back and these are my current temps:

[Nick N]

I left the sim on while I ran out.




I came back and these are my current temps:

Thats looking sweet

I think we nailed your thermal problems. Just make sure when you get a chance to blow the perf metal out of those fan boars. I usually mount the fan to the outside of the tower where I want the hole and use the inside of the fan edge to mark the perf with a fine line sharpie. When the fan is then un-mounted it leaves a perfect mark that when cut and filed so the black line is gone, no perf edges will hang into the airstream generated by the fans.

I dont want to hold you up but I really would like to be sure this all remains stable and we give the compound time to cure. If you feel you put the screws to it and everything seems solid, the next numbers to boot will be:


NEW FSB 1300
NEW MEMORY 650

Leave memory at 2.0v unless it is unstable, then instead of raising to 2.1v I may want to run the latest memtest86+ on those sticks.

If that is the case, set it back to:

NEW FSB 1280
NEW MEMORY 640

and let me know what happened. Between now and then I will consider if the memtest runs should be done. I may want to try raising vcore to 1.475v

If all remains stable with the next boost in speed I would like you to stay on those settings for at least 48 to 72 hours. At that point and with the system being turned on and off for curing we will go for another round.

I will be quite honest, that chipset is only rated for 1355-1385 max with extreme (chiller type) cooling and DDR800 memory. If we can get you to 1333/333 on the swifty and DDR667 I will really be impressed. Don't jump the gun on that and try it because it may require Vcore and other changes and I want to make sure that compound is properly cured before we go slapping it silly.

[Black ZR-1]

Thats looking sweet

I think we nailed your thermal problems. Just make sure when you get a chance to blow the perf metal out of those fan boars. I usually mount the fan to the outside of the tower where I want the hole and use the inside of the fan edge to mark the perf with a fine line sharpie. When the fan is then un-mounted it leaves a perfect mark that when cut and filed so the black line is gone, no perf edges will hang into the airstream generated by the fans.

I dont want to hold you up but I really would like to be sure this all remains stable and we give the compound time to cure. If you feel you put the screws to it and everything seems solid, the next numbers to boot will be:


NEW FSB 1300
NEW MEMORY 650

Leave memory at 2.0v unless it is unstable, then instead of raising to 2.1v I may want to run the latest memtest86+ on those sticks.

If that is the case, set it back to:

NEW FSB 1280
NEW MEMORY 640

and let me know what happened. Between now and then I will consider if the memtest runs should be done. I may want to try raising vcore to 1.475v

If all remains stable with the next boost in speed I would like you to stay on those settings for at least 48 to 72 hours. At that point and with the system being turned on and off for curing we will go for another round.

I will be quite honest, that chipset is only rated for 1355-1385 max with extreme (chiller type) cooling and DDR800 memory. If we can get you to 1333/333 on the swifty and DDR667 I will really be impressed. Don't jump the gun on that and try it because it may require Vcore and other changes and I want to make sure that compound is properly cured before we go slapping it silly.

Check.

I had to raise mem to 2.1. I coudldn't boot up with 2.0.

[Nick N]

Thats looking sweet

I think we nailed your thermal problems. Just make sure when you get a chance to blow the perf metal out of those fan boars. I usually mount the fan to the outside of the tower where I want the hole and use the inside of the fan edge to mark the perf with a fine line sharpie. When the fan is then un-mounted it leaves a perfect mark that when cut and filed so the black line is gone, no perf edges will hang into the airstream generated by the fans.

I dont want to hold you up but I really would like to be sure this all remains stable and we give the compound time to cure. If you feel you put the screws to it and everything seems solid, the next numbers to boot will be:


NEW FSB 1300
NEW MEMORY 650

Leave memory at 2.0v unless it is unstable, then instead of raising to 2.1v I may want to run the latest memtest86+ on those sticks.

If that is the case, set it back to:

NEW FSB 1280
NEW MEMORY 640

and let me know what happened. Between now and then I will consider if the memtest runs should be done. I may want to try raising vcore to 1.475v

If all remains stable with the next boost in speed I would like you to stay on those settings for at least 48 to 72 hours. At that point and with the system being turned on and off for curing we will go for another round.

I will be quite honest, that chipset is only rated for 1355-1385 max with extreme (chiller type) cooling and DDR800 memory. If we can get you to 1333/333 on the swifty and DDR667 I will really be impressed. Don't jump the gun on that and try it because it may require Vcore and other changes and I want to make sure that compound is properly cured before we go slapping it silly.

Check.

I had to raise mem to 2.1. I coudldn't boot up with 2.0.

[Nick N]

I am going to do a little research on that board and see if there may be a way to manipulate the BIOS into tricking the northbridge into thinking its running slower than it is. They call it 'bootstrap' on the 965-975 chipset motherboards where the northbridge is locked at a DDR2 800 ratio but still allows frequency changes, maintaining the same mathematical ratio all along. Being your BIOS is very limited it may simply not be possible, but still if it is possible to find a very carefully calculated ratio that might have that affect we may be able to take it up a notch again after we run through the 1:1 setup.

I want to get the system stable and set up @ 1:1 with the sim and VC tweaked before we do anything like that so if there may be a positive or negative change with attempting a bootstrap ratio it can be seen correctly.

[Black ZR-1]

[quote]It would appear we are finally seeing the 'true' wall on the memory. Before you could get no where near 1280/640 @ 2.0v and now it ran stable through that up to 1300/650

Right now you're CPU is running well above its rated total bus speed of 1066 (4x266.5MHz) and instead @ 1300 (4x325MHz) a gain of 234MHz which is a 22% increase in discussion rate between the memory and CPU.

The CPU is now at 1950, 64MHz higher than stock or a 3% increase. Between the 2 you are running at just about a 25% true system boost.

Unfortunately there is no multiplier availability which would allow us to jack the CPU independent of the other values. With that and the limit imposed by the chipset you are a bit stuck but we will tweak it out the best we can with what you have to work with.

The HT (LDT) of the system is 3x (LTD) the mem speed @ 975Mhz (3x325) We want to keep the HT @ or below 1000 (for right now) and you are at 975 so its good. Later may raise this again and see if we can boost the system from the back end numbers.


It would not appear your BIOS gives you access to the follwoing values in memory:

Trrd
Trwt
Twtr
Twrwr

and it would be interesting to see what those are.

I think based on what I am seeing it may be a good idea if you have any instability to raise the memory timing to

5-7-7-19-31

[Nick N]

[quote][quote]It would appear we are finally seeing the 'true' wall on the memory. Before you could get no where near 1280/640 @ 2.0v and now it ran stable through that up to 1300/650

Right now you're CPU is running well above its rated total bus speed of 1066 (4x266.5MHz) and instead @ 1300 (4x325MHz) a gain of 234MHz which is a 22% increase in discussion rate between the memory and CPU.

The CPU is now at 1950, 64MHz higher than stock or a 3% increase. Between the 2 you are running at just about a 25% true system boost.

Unfortunately there is no multiplier availability which would allow us to jack the CPU independent of the other values. With that and the limit imposed by the chipset you are a bit stuck but we will tweak it out the best we can with what you have to work with.

The HT (LDT) of the system is 3x (LTD) the mem speed @ 975Mhz (3x325) We want to keep the HT @ or below 1000 (for right now) and you are at 975 so its good. Later may raise this again and see if we can boost the system from the back end numbers.


It would not appear your BIOS gives you access to the follwoing values in memory:

Trrd
Trwt
Twtr
Twrwr

and it would be interesting to see what those are.

I think based on what I am seeing it may be a good idea if you have any instability to raise the memory timing to

5-7-7-19-31