hmmm what bios version should one choose...leave it at 701, change it to 801 or try beta 901.
Printable View
hmmm what bios version should one choose...leave it at 701, change it to 801 or try beta 901.
Am I way off if I say that we should prefer a higher temperature on the heatsinks when we change cooling paste? I mean the paste doesnt have anything in it that makes anything colder, it only conducts heat from the core to the heatsink. And better conduction should mean higher temperature on the heatsink and lower on the core? Heat cant just disapear.Quote:
Originally Posted by davefr
you're not wrong/way off. but this is the ideal situation. less heat on the components that we want cooled and more heat on the components that can take it.Quote:
Originally Posted by Larsson85
I measured at the base of the heatsink right on top of where the chip was.Quote:
Originally Posted by Larsson85
If I increased the efficiency of the chip to heatsink interface area then it's effectively dissipating heat consistently across the entire heatsink vs. dissipating heat in a mosaic pattern across the interface. Wouldn't that result in less temperature at the base since it's dissipating across a wider area?
Took it off and the heatsink gets a bit hotter now, so it's still transfering heat.Quote:
Originally Posted by Zulusman
If you look at the heatsink from the side you can see it touching the chips even without the rubber pad. Looks to me like it's touching all chips, checked with a ruler to see all chips were equally high, and they were...
IIRC thermodynamics, its like this: The temperature of the heatsink is in direct proportion to the amount of energy in Watts that it can dissipate, its a constant (within a range) called themal resistance TR=C/W. Given that boundary conditions remain the same (like air flow and heatsink surface area), for the same thermal load improving (reducing) the thermal resistance will lower the load temperature. Improving the thermal conducting interface (reducing system TR) will increase the heatsink temperature as more energy passes across the interface and out of the load. Heatsink temperatures are a gradient, hottest at the conducting interface.Quote:
Originally Posted by Larsson85
Like what you said.:p:
That's the whole idea behind looking at idle/load temp differences between different heat sinks.Quote:
Originally Posted by Larsson85
Heat Accumulation (on the chip) = Heat generation - Heat dissipation
Heat generation scales slightly, but dissipation grows linearly with the difference in surface and ambient temperature with a slope depending on area and heat transfer coefficient (we increase this coefficient by applying low thermal resistance stuff, like AS5).
When we talk about stable temps, we're showing what conditions at steady state are, which is when heat accumulation = 0. So it's all about how low we can get this steady state temperature to be.
in short..we want to get the heat away from the chips and into thin air ASAP
I admit to be thermodynamically challenged however if you had the "ideal" heat sink wouldn't it absorb the chip's junction heat and dissipate it into the air such that the heat sink itself is remaining at ambient. (ie since it's 100% efficient at transferring heat, then it's not absorbing any residual heat).Quote:
Originally Posted by [G.N.U.]Fragman
If that's true, then as a heat sink becomes more efficient then it's surface temperature is being reduced. Conversely as it becomes less efficient isn't it absorbing vs transferring heat to the ambient.
In my above test. Is a 1.5 degree drop reflecting improvement or degradation of heat sink transfer efficiency? However I think the real answer is that's it's an insignificant change.
the heat sink itself can never be at ambient, since it relies on a temperature difference to transfer heat to passing air. the absolute best heat sink would be nearly the same temperature as the surface it's attached onto to cool, as that would mean that the only thermal resistance comes from the transfer of heat between the heat sink and air.
I don't quite understand the vMHC mod, is "grand" = "ground"?!?! and if yes, is that a common ground, or is that the "ground" potential of that one chip?
Sorry guys if this has been answered before but with the removal of the heat pipes, do you need to replace the thermal pad under the radiator looking thing? My board is in pieces now and I only have that last step to do.
I'm not sure about this. This question gave rise to the thermodynamic discussion.Quote:
Originally Posted by J@mmer
If you decide to take it off, make sure the base is still making full contact with the components.
Yeh looks like it's just thermal padding that covers the MOSFETS. I think I'll leave them and reuse the same ones.
Oh BTW... thanks FCG for this great guide. :)
Just finished your guide.
So I'm about to watch a movie and I notice the liquid in my system is especially pale. I open the box and the pump is making a weird noise, and there is a leak in the bottom. There are reasons for this happening that I'd rather not go into. So then and there I was like "THAT'S IT! BACK TO AIR I GO!" After all, it is just my 805, it's just holding me over until Conroe (by then I'll have phase).
So I remembered the guide, something I wasn't going to do until the next board pull for phase installation, but I figured why the hell not.
Did the wireless removal. No wireless devices.
AS5 on the North and South. I'm used to this procedure, but the tip on the screwdriver off the IDE port REALLY helped.
The back of my board is indeed blue. Sorry about the confusion before.
I revamped my CCFL lighting... the bulbs are now invisible. Only the light you can see. Looks even better than it did before... much more subtle and controlled.
Fixed some minor wiring issues, and removed the chassis intrusion crap. Now the only things you can see are the 24pin humongo and the SATA cable. The rest is hidden... ALL of it! Most behind the board. Case looks soooooo clean...
I was impressed by the stock heatsink mounting. Kinda cool. Relies a lot on plastic, but other than that, very creative and easy.
So anyway... South appears down almost 10*C (EDIT nevermind... nowhere near). Who knows about the North, but I know it's much cooler than it was before. I really wish the board manufacturers would spend the extra $.25 per board and use even an AS5 knockoff instead of white-out.
What are typical idle/load temps on an 805 with the stock cooler? I'm idling at about 50*C... full load? WOW. 86*C. That's only 1.35v in BIOS, too. Does the stock fan suck THAT BAD? I'm sitting at 3.4ghz right now trying to play it safe... I mean, real voltage is 1.3v. 'Stock', according to box, is 1.25-1.4v. So how on earth did 1.4v get it hot enough to shut down?
EDIT ok wow I feel dumb I thought you turned the little things after you pushed down to 'lock' it in place (the push down seemed to easy) but those actually RELEASE the heatsink... so I had... very little pressure. Temps still really high/bad. Didn't even go down 10*C. How hot is too hot for the 805?
Did you check if the IHS is making a proper contact with the HSF? With stock HSF, I often saw like only half the IHS (in the center) showing proof of contact with HSF. I'd also use somewhat generous amount of artcic silver.
My 930 did 3.9GHz with stock cooler / stock Vcore, the load temp was close to 75C. Changed to Thermalright Ultra-120 and bumped the Vcore by 0.025V, the result was 4.1GHz and max load temp 60C.
ok i just ran STOCK voltage (like, AUTO) (which UNDERVOLTS like hell) and got 70*C max. so what i dont understand is, how can the box say 1.25-1.4v, when im running lower than that and getting 70*C? if i use like 1.4v by CPU-Z, the computer WILL shut down. im giong to try a remount today, but come on...
1.325v in BIOS: 80*C during orthos.
about 1.26v in CPU-Z. this is just pathetic.
It says twice to take out the CMOS chip, is this a typo?
"First, pop the CMOS chip out using a tiny flat-head screwdriver."
"Now, boot the operating system with the boot disk we just created. When you're at the DOS prompt.....take the CMOS chip out of the running system"
The guide is actually correct. It might sound somewhat harsh to take out the CMOS chip when its running - but that is the process of hotflashing :)Quote:
Originally Posted by ruler22
Take out the chip from the running (temporary) rig and insert your Asus P5W DH dlx bios chip in the running (temporary) rig and flash away...
-k0nsl
When I read it the first time it sounded like you didnt need another system.....
and if you do need another system (thats what I am getting from the temporary rig reference in your reply) does it have to be another LGA775 board, or any board? Sorry, im just a little confused, I dont have my board yet, but I want to understand this before I get it
I have a SOYO kt-400 mobo, im assuming since it is older and AMD, I can't use it right?
If I can't use it( which im assumign I cant) is there anyway around it, short from having Asus ship me a new BIOS?
Nice guide
Maybe add
Put a fan on the nb and connect to the board connector who is reserved for it.
use the asus wifi mounting to mount your mod's on the left side of the board
like this
http://rol-co.nl/images/asusmod.jpg
2 really nice ideas there
though im going to resist the fan
reminds me of the DFI fan
EDIT and what sort of fanciness do you have on the CPU holder?
Is is possible to hardflash the right bios in my P4P800?
never heard of it before.
You are right. A heatsink that is hot as hell is a good sign because that heat is not in the chip.Quote:
Originally Posted by blink hi
Mounting for the mach II cooler :DQuote:
Originally Posted by AndrewZorn