Check this homemade WB out

[JoeBar]

Keep the work on a new block nate. I would skip the whole fan/wb idea totally though. Focus on making a plain wb...

[nateman_doo]

Why would I want to quit?

[UrbanSmooth]

Why would I want to quit?

ROFL! Exactly!

NEEEVVAAAARRRRRRR!!!!!!!!!!!!!!!!!!!!!!

[nateman_doo]

Version 1.1 has been completed. Unfortunately, it broke my mill. Now the gears are all messed up.

[nateman_doo]

Replacement Mill is on the way. What are opinions on baseplate thickness?

[Manicdan]

1/4" sounds like a safe middle point

[Church]

Simple. As thin as you can get from cpu's IHS to water without it overly loosing structural rigidity to not get too deformed when it's mounted on cpu + choosen height of microchannels (higher ones probably will result in lower flow restriction, as overall cross-section for water will be bigger). Afterall best thermal results will be with one atom thick layer between water and coolable surface

[BeepBeep2]

Heatkiller is 2mm.
...and the microchannels are 1.5mm...meaning parts of the baseplate are only .5mm thick :O

[Church]

HF's microchannels are ~ 1.9mm though.

[nateman_doo]

2 mm sounds like a good number to me. Its do-able.

[BeepBeep2]

HF's microchannels are ~ 1.9mm though.

http://www.skinneelabs.com/heatkiller-i7.html

It's less complicated than it sounds really--the base is 2mm thick and over the width of a typical IHS, it has ~52 microchannels that are 1.5mm deep into the base. What does that mean for you? You have water flowing within .5mm of the very bottom of the base and have a lot of surface area really, really close to the heatsource, your CPU.

[Manicdan]

i wouldnt get that close at first, how bad would temps be with an extra 3mm in there? cant be more than 1-2C difference

[nateman_doo]

You wouldn't got 2 mm to the base?

[Alexandr0s]

If a block's that thin it will bend a lot under the strain of the mount, reducing contact surface. So too thin isn't always that great as well. Though you could try doing a very thin centerpiece, which isn't big enough to cover the entire CPU. That way, the thicker part of the base also still sits on the CPU, which should make it bend less.

[Waterlogged]

If a block's that thin it will bend a lot under the strain of the mount, reducing contact surface. So too thin isn't always that great as well. Though you could try doing a very thin centerpiece, which isn't big enough to cover the entire CPU. That way, the thicker part of the base also still sits on the CPU, which should make it bend less.

Can also be made to "bend" the other way. Use an o-ring to seal the inlet area pathway and create a bow in the base.

[nateman_doo]

There is no bending of this base, trust me

The new version is the same exterior (mostly) but the internals are completely different. The motherboard will flex WAY before the block does. So will the space time continuum.

[Church]

nateman: no bending - so probably solid base part very thick? It won't affect performance of block too much?

[Manicdan]

the mounting is built to take 70lbs of force like its nothing. if you dont put center pressure on it, then it can flex, even just by a few thousandths of an inch, which then gives horrible contact where it needs the best contact (in the center)

[nateman_doo]

a few thousandths of an inch is what the thermal paste is for.

[BeepBeep2]

a few thousandths of an inch is what the thermal paste is for.

Not exactly.
Thermal paste is for little gaps in the surface that you can not see.
Not a surface which is not flat.

Most waterblocks on the market are bowed, so they end up being flat when mounted. There is a certain amount of pressure needed to mount them properly.

[nateman_doo]

On that same token, most waterblocks are also a thinner base plates. I don't see much flex in my block due to its massive size, It has over 1/2" of solid copper. The likelihood of four 6-32 screws warping it before the board is just inconceivable to me.

[Waterlogged]

On that same token, most waterblocks are also a thinner base plates. I don't see much flex in my block due to its massive size, It has over 1/2" of solid copper. The likelihood of four 6-32 screws warping it before the board is just inconceivable to me.

Nate, that is air cooler thinking and it will lead you down the wrong path to a lot of wasted material. Thinner is better for water. . .please, I beg you. . .trust us on this.

[nateman_doo]

I'm sorry, just was saying that the HK LT block is only a thin piece of copper. My baseplate is 2.3mm thick where it makes contact with the CPU. While my baseplate is thin, it also has a surrounding copper body which it will not flex from the "strength" of a PCB.

[Manicdan]

I'm sorry, just was saying that the HK LT block is only a thin piece of copper. My baseplate is 2.3mm thick where it makes contact with the CPU. While my baseplate is thin, it also has a surrounding copper body which it will not flex from the "strength" of a PCB.

heres what you can do and its very simple

mount your block normally and get the cpu running hot, when your temps stabilize start cranking down the screws a little, like a half turn each, wait 1-2 minutes and see if the temps go up or down or not at all. (the actual amount you tighten the is based on if you use springs or just thumb screws against the block, if u dont have any springs then a full turn will break something cause they are already super tight, so go very minor then)

if your temps go down then you have room to tighten it, if your temps are going up, then you have a base thats flexing. repeat the same test but loosen the block, which should obviously raise temps unless it was already flexing.

theres so much math involved with energy transfer, surface area, and other crap, so knowing whats best takes a rocket scientist, or we just do that simple test and see the before and afters. thats why i like adjustable mounts, its hard to have a bad mount cause you can fix it without having to take everything apart

[nateman_doo]

UPDATE: Baseplate is >1mm

.95mm is what I measured.