My first waterblock design

[Paxe]

What do you thing of this waterblock?
Don't know if its dimensions are right. It's made to cool down an amd athlon xp 2100+ on an Asus a7v333.
As you can see it's similar to Fixxit's Spir@l waterblock.

[TerraMex]

This is more practical ... : (for those that dont actually have acad)

[erbiagio]

http://www.cooltech.it/ctengine.php?...ge=wb_003.html

it's the same design

[MadHacker]

It's a little bit diffrent...
the center walls is a lot thinner in this design

[liquidcooler]

I did go the cheap way and it worked fine for me, here you can check how I made it: http://www.eng.losconectados.com

[#Rotor]

liquidcooler....
Now pardon me for my stupidness.... but how did you get the holes in every 2nd fin of that sink....

you did a damn sweet job man.... this is good stuff....

[liquidcooler]

Hahaha, that's easy: I did the other way round: first I drilled one hole till the end and then covered with epoxy (mixed with aluminum powder) every second hole.

I like very much your publishing of you system temperature and load (though I couldn't see them). Compared to what I've been reading around my system is clearly an underperformer, but I don't believe there can be such big differences.

The difference in temperature between the water (measured at the waterblock) and the processor (measured by the Asus Probe) is very stable at 10 ºC. I can lower the water temperature giving more power to the radiator fan, but what really worries me is these 10 ºC, I don't really believe the waterblock perfomance can be so bad compared to others.

[#Rotor]

aaaa but that is cheating....


you mean you can't see a picture, when clicking the link in my sig?

it should look like this..


your temps sounds reasonable, remember that your block design has a lot of smooth long flat surfaces. That is very advantageous for laminar flow. Which in turn, is very bad for block performance. The Zig-ZAG of the flow path helps a lot to overcome this drawback though, from there your very respectable temps.

[liquidcooler]

I didn't mean to cheat, sorry for some reason in the English version there was not the pic, anyway in the pic you don't see very well the holes covered.

I did some testing using the waterblock with transparent walls and the flow was clearly turbulent (though when testing it did'nt share the pump with the PSU and GPU). The flow doesn't need to be laminar just because the walls are flat, it depends on the Reynolds number which in turn depends on the fluid density, pipe diameter and the fluid velocity. I didn't bother to calculate it for this case because I "saw" the flow was turbulent.

[crxy_sasquatch]

How will you keep the water flowing through the spiral without it leaking over its sides

sorry, I"m a noOb.

[artsite]

Sorry , but for what is this




Tanks

[bigben2k]

It's a motherboard backplate.

You mount it behind (or under) your mother board, and it's supposed to help prevent the board from bending, as the mounting pressure is applied.

[artsite]

What material can it be made ?

Acrilic?................or cooper...............

[erbiagio]

the backplate is made of alluminium with 4 plastic washer

[Pritorian]

Originally posted by TerraMex
This is more practical ... : (for those that dont actually have acad)

Wón´t that be really hard too mill when you have so thin walls.

i thought that the mill varied some i width when it took turns.

-P-

[FL3JM]

You people should check this block out:
http://forum.sweclockers.com/showthr...hreadid=163491
Although the page is in Swedish, the pics is the important thing.
The really bad-ass thing about the block is that it has a silver bottom! You clearly see the silver part of the block on pic 2.

[Pritorian]

It´s only got a silver bottom, not the innards of the block, so i would think that the TIM joint between the copper and silver would just make the silver a total wast, IMO.

-P-

[FL3JM]

It´s only got a silver bottom, not the innards of the block, so i would think that the TIM joint between the copper and silver would just make the silver a total wast, IMO.

I talked with the guy who makes the blocks (and infact sells them in small quantities), he didn't want to say how but he told me that the block is completely solid.
I have ordered a block.

[liquidcooler]

Best of all (by far): http://www.3dnet.hr/dr-ice11-en-01.html

[Pritorian]

Originally posted by FL3JM
I talked with the guy who makes the blocks (and infact sells them in small quantities), he didn't want to say how but he told me that the block is completely solid.
I have ordered a block.

Hmm, find out how he does that,

[tex707]

Originally posted by bigben2k
It's a motherboard backplate.

You mount it behind (or under) your mother board, and it's supposed to help prevent the board from bending, as the mounting pressure is applied.

I think it could be better to use a piece of elastic matherial in the center of the backplate, like on the attached screenshot.

[Nuson]

wouldn't heat travel better through those walls if they were thicker?

[MadDogMe]

I think it could be better to use a piece of elastic matherial in the center of the backplate, like on the attached screenshot.

Make the rubber pad bigger, remember the CPU is mounted in a socket contraption so making the pad the same size as the CPU is pointless cause it's the socket that 'transmits' the pressure to the mobo>plate not the CPU, idealy it'd be a [ ] square shape the same 'footprint' as the socket(the pins are clearly visible on the mobo's back)...

I'm surprised more people don't use a backplate mobo bending can realy **** up the equilibrium of you mounting pressure leading to an uneven mount and poor temps. Graphics cards especialy bow like crazy when the holes are used to mount a block, it can't be good for the traces, 'specialy the inner ones!...

[tex707]

Originally posted by MadDogMe
Make the rubber pad bigger, remember the CPU is mounted in a socket contraption so making the pad the same size as the CPU is pointless cause it's the socket that 'transmits' the pressure to the mobo>plate not the CPU, idealy it'd be a [ ] square shape the same 'footprint' as the socket(the pins are clearly visible on the mobo's back)...

I'm surprised more people don't use a backplate mobo bending can realy **** up the equilibrium of you mounting pressure leading to an uneven mount and poor temps. Graphics cards especialy bow like crazy when the holes are used to mount a block, it can't be good for the traces, 'specialy the inner ones!...

Thanx for the tip...however, this is just an example for presentation purposes only...I do use a larger rubber pad (made of an old mouse pad) and this is a good way to keep a MB in a normal conditions, witout any bends.

[MadDogMe]

Cool ...

It's such a good idea I don't know why nobody's used it much before now . I have a few perfect bits of Alu laying around I'm going to use. I'll 'double' the VPU~card one up as a 'rear' heatsink. I'd prefer to use some polycarb for the mobo one but don't have a 'score' where I live for it...