A different kind of spiral...
 

OK, this was "inspired" by jaydee16 in this thread here.

He was a bit frustrated, but I think I got his drift.

Comments? Criticisms? (I have my own re: manufacturability!)

Bob

first pic:

In at the bottom, out at the top.

second pic:

Excuse my awful English. I find the center of its block too fine. I have built one similar but without helix.
http://www.arkania.org/~jordimarch/m...05%20copia.jpg

More details can be seen in: http://www.arkania.org/~jordimarch/i...ewlang=spanish

I don't have elements for of carrying out good mensurations.
36ºC cpu
21ºC atmosphere
measured with MBM 20 minutes CPU-Burn ECSk7s5a XP1800@1643

Aitor... that looks sweet...

You are almost at the same temps I'm at (36.3ºc-Core in 24.1ºC air)


what kinda radiator are you using there... and what is you fluid temps to get that....



that coil looks awesome man... but one problem... Manufacturing it...

Re: A different kind of spiral...
 

Yes, that is pretty much what I had in mind. Lost my webhost so my drawing isn't up anymore but yours is superior anyway!

Nice looking block aitor!

Sort of like the Innovatech...

http://www.overclockers.com/articles545/inno14.jpg

That is very good idea Bob, do try to make one and test it. Also flow rate shouln't be restricted too much because of the spirall, now the only thing is what would be the optimal cross section of the chanell and the step of thet helix.

aitor, yes sweet looking waterblock you have there, good job

How about a 'triple' thread, three chanels opening together?...

'splain, please!

Bob

Hmmm?, a double thread would be like slipping two springs into each other from the side, two parralel channels for the water to run up. you could hve three or more but i't be harder the more you had.
Get a pencil and two bits of string, then wrap them around the pencil without them overlaping, think of them as the water channels.
PS, I though of three at first cause thats the height (in threads) of the inlet.

OK, now I see what you are getting at. Well, I can probably model it, but I shudder to think of the manufacturability issues. Let me see what I can come up with...

Bob

Hey! My 100th post!:D

I have a feeling that the pitch of the helix will ultimately be determined by what I can get my friend to machine, rather than any "insignificant details" like optimal cross-sections, etc.!:cry:

Bob

IMO, it would perform dismally:
Since the water would be forced into a circular path (already restrictive), the centripedal forces would push the heavier water out the sides. By "heavier water" I mean cold water, since hot water has a lower density.

I think you'd get better performance shooting water around a central post, even a screw or bolt!

my .02...

I would also expect this unit to perform dismally, because the central post is too massive. A stepped post size might perform better.

edit: just saw temps: 36C isn't bad, but for what CPU, and what OC?

OK, here's one for ya...

Comprised of 3 helixes rotated 120 degrees. Problem with this is the dead spots caused by the single opening slot. Perhaps a plenum-type inlet/outlet....hmmm

Here's the pic:

Perhaps, but it would look pretty cool, eh? :cool:

Kudos for looks! With the right LED combo, it'll kick ya in the eye!

PROBLEM
 

It looks cool, but I suspect you'll get very poor flow where you need it the most--at the base.

At a glance, it looks like water feeds in on several 'ramps' in parallel at the bottom, and goes merrily along to the top, where it exits the ramps.

But it won't work that way. The water on the bottom ramp has to merge with the water entering the second, and they both have to merge with the water coming in on the third...and since water is incompressible, it has to move faster to move a greater volume of water.
I don't see any way of balancing it, either.

For an analogy, think of the parking garages in big cities, and how much of a pain in the butt it is for everyone on the different levels to merge onto the spiral ramp.


Cool looking design, though. :D

<edit>Also, since water is a fluid, it's going to take the easiest path. The path with the least resistance due to direction changes/centrifugal force, etc. is the one where the water enters the highest input, and exits at the lowest output.
The path with the least resistance due to friction between the water and the block, and also between the water and itself also happens to be that same shortest path.
In effect, it's basically only going to touch the middle of the block, and leave stagnant hot water at the base.

Now convection will counteract this, but not by much, I suspect.
</edit>

Regarding the requested values I apologize for not being able to give more data but I don't have probes neither thermometers to be able to carry out any mensuration type. The values of temperature of the cpu and voltages were measured with the mother monitor. The temperature of the water with a fishbowl thermometer.
Athlon xp1800+@1643(143*11.5)
Full= 36ºC (20 minutes CPU-Burn) idle= 30ºC
Temp. H2o = 26ºC
Temp. room =21ºC
Pump Regent ?????l/h ????psi
Radiator = car Fiat147 (71/4"x5"x1")
1 Sunon 120m.m.

utabintarbo my suggestion in the figure (difficult mechanized)
http://personales.ciudad.com.ar/aito...aydee16bis.jpg
aitor

I think a single 'Course' thread, (like a woodscrew rather than a bolt) might have a chance of cooling. but it's ignoring the bottom where all the heat is, saying that Inovatech get along all right don't they?:shrug: ...

I am unclear as to what this modification will accomplish:confused:

Bob