Block design...need insight and maybe a name?

[MMZ_TimeLord]

Another item to remember when talking about surface area for picking up heat in a waterblock is that the amount of copper you are using in the first place determines how quickly large tempreture fluctuations can be compensated for.

More copper, the quicker the copper soaks up the heat from the heat spreader or core. The whole point of these extreme cooling solutions is not to make a pretty block or a small block, but to make a block that removes the heat from that CPU as fast as possible.

Copper will soak up heat until it's evenly distributed. When it has water flowing through it, it will carry the heat away. Even if that heat sits in the copper for a few moments, it will be OUT OF THE CPU.

The more surface area you have contacting the water the more heat you will ultimately transfer to the water and OUT of the block. The reason fins and pins work so well is that it increases the surface area that contacts the water during it's path through the block.

Presuming your radiator can get rid of the same amount of heat or more that your CPU can generate, then you will have a system that will stay pretty cool.

The block I made happens to be 2" x 3" x 0.5" solid copper and has 0.25" passages bored into it. I have 5 switchbacks in the path which increase my path length to approximately 10.6 inches. Because of that my total surface area inside block is about 8.3 square inches. That's more than the surface area of one of the large sides of the block itself (2" x 3" = 6 square inches).

Hope that helps you...

[Volenti]

Originally posted by MMZ_TimeLord


More copper, the quicker the copper soaks up the heat from the heat spreader or core. The whole point of these extreme cooling solutions is not to make a pretty block or a small block, but to make a block that removes the heat from that CPU as fast as possible.

I used to think like this, I used to think that center inlet blocks we're a fad, I've learned to think different

If you can make a "pretty" block or a small block perform as well (or better) than a big brick of a block then why not make them like that?

[MMZ_TimeLord]

If you can... more power to you... I could not. :shrug:

[Rayman2k2]

1st design, second one is coming soon

[Rayman2k2]

thinking it over....this block is..."out of proportion". The center area with the pyramids is 1/2" long, 1/2" wide. The whole block is 1 1/2" long by 1 1/2" wide, so...yeah...this block is DEFINATELY out of proportion:shrug:

[jaydee]

Originally posted by Rayman2k2
thinking it over....this block is..."out of proportion". The center area with the pyramids is 1/2" long, 1/2" wide. The whole block is 1 1/2" long by 1 1/2" wide, so...yeah...this block is DEFINATELY out of proportion:shrug:

If you are designing this for an Intel P4 then make the pin array wider then the heat/load spreader on the CPU. Figure out the pin array size, then worry about over all demisions of the block. I know jack about Intel CPU's so thats all I can add.....

[Rayman2k2]

aright, im not sure whether the inner area should be the size of a P4, or the outer area should be the size of a P4.

[hara]

The small channels on the base are superfluous IMO.

[bigben2k]

Originally posted by hara
The small channels on the base are superfluous IMO.

They're decorative, like in the Maze3.

It's a good design. The pyramids give the additional surface area needed. I'm not sure that it's optimized for a center inlet, but it'll work well either way. I wouldn't even bother with the walls: just let the flow exit freely, at either 2 or 4 outlets.

[Rayman2k2]

Originally posted by bigben2k
They're decorative, like in the Maze3.

It's a good design. The pyramids give the additional surface area needed. I'm not sure that it's optimized for a center inlet, but it'll work well either way. I wouldn't even bother with the walls: just let the flow exit freely, at either 2 or 4 outlets.

why is it not optimized for center inlet? and, besides flow restriction, will getting rid of the walls benefit performance?

[Rayman2k2]

and also, should the pyramid section be the size of a P4 processor, or should the entire chamber be the size of a P4?

[8-Ball]

Where was it that they discovered that the Heat spreader actually doesn't spread the heat. It's just too thin. The effective heat source is still pretty much the size of the die beneath the heat spreader.

I think it might have been BillA who mentioned it after I mentioned that I was designing a Hammer waterblock with an effective cooling area about 45mm x 45mm.

Simplest thing to do is just pretend it aint there.

Their main advantage is protecting the core.

8-ball

[Rayman2k2]

Originally posted by 8-Ball
Where was it that they discovered that the Heat spreader actually doesn't spread the heat. It's just too thin. The effective heat source is still pretty much the size of the die beneath the heat spreader.

I think it might have been BillA who mentioned it after I mentioned that I was designing a Hammer waterblock with an effective cooling area about 45mm x 45mm.

Simplest thing to do is just pretend it aint there.

Their main advantage is protecting the core.

8-ball

so the water chamber should be the size of the P4?

[8-Ball]

The point I'm making is that the pyramid region need only be slightly larger than the silicon die beneath the heatspreader. Though I don't know how big this is.

Also, don't take my word on this until I can find the article or someone like BillA backs me up.

8-ball

[Rayman2k2]

Originally posted by 8-Ball
The point I'm making is that the pyramid region need only be slightly larger than the silicon die beneath the heatspreader. Though I don't know how big this is.

Also, don't take my word on this until I can find the article or someone like BillA backs me up.

8-ball

oh ok...........wait....dont trust 8-ball? *gasp* (Faints onto floor)

[jaydee]

Originally posted by 8-Ball
Where was it that they discovered that the Heat spreader actually doesn't spread the heat. It's just too thin. The effective heat source is still pretty much the size of the die beneath the heat spreader.

I think it might have been BillA who mentioned it after I mentioned that I was designing a Hammer waterblock with an effective cooling area about 45mm x 45mm.

Simplest thing to do is just pretend it aint there.

Their main advantage is protecting the core.

8-ball

I will buy that. What is the heat spreader made of and how thick is it and how big is the core underneath. I need to get a P4 system one of these days.

[jaydee]

Originally posted by Rayman2k2
oh ok...........wait....dont trust 8-ball? *gasp* (Faints onto floor)

No he is trustworthy. He forgot more about this stuff than I will ever know!.

[bigben2k]

It isn't optimized for flow from a center inlet, sorry, I should have specified. The pattern is not circular in any ways. Nicozeg made "stepped rings" in the first design he presented. Radius is just too obvious

Getting rid of the walls will be less restrictive, and that's an open door for more performance, even if it's just a little bit.

At this point, you might consider your mounting options, because you could have an issue with the structural integrity, as I did with the 1.0mm baseplate. If you incorporate 4 posts where you had walls, and use a copper top (or thick polycarbonate, like me), you'll have that issue licked.

[Rayman2k2]

hmmmm.....it would be impossible to make the center area round, being it that I have SQUARE pyramids.....shucks......anyways...i have a new design that i'll post later....to much god damn homework....

[8-Ball]

Originally posted by jaydee116
No he is trustworthy. He forgot more about this stuff than I will ever know!.

I think perhaps you exagerate. I just have an ability to apply theoretical principles to just about anything to get a reasonable grasp of what's going on. And the ability to visualise things very clearly in my head. What I lack is real world experience, though this WILL change once I've got these damn finals out of the way.

But thank you for your confidence and support.

8-ball

[8-Ball]

I forgot to add, I have a great deal of respect for the experience people like yourself have by building countless blocks and experimenting with new ideas, rather than just babbling about the theory like I do.

I may have a reasonable grasp of the concepts, but that ain't gonna cool my processor down!

8-ball

[Rayman2k2]

I'm planning on having a thick lexan top, working on another design right now, but im having some...uhhh...technical difficulties...

[Rayman2k2]

how effective are center inlets? like, if i were to have one alittle bit off, will i see a major increase in temps? or a minor one?

[jaydee]

Originally posted by Rayman2k2
how effective are center inlets? like, if i were to have one alittle bit off, will i see a major increase in temps? or a minor one?

Depends on the design, but I see about 1-3C better with a center inlet. You can be off a bit, but try to keep it at least half over the center. Closer the better.

[Rayman2k2]

aright, here is another design, the pyramids are in the center. 1 nozzle in, 2 nozzle out.


EDIT: forgot pic