One more solid container.

[ilkkahy]

Its not yet as i planned and it isnt finished either but here are some pictures anyway:






I will test it hopefully next week with 160w heatload tester (that thing wont give any ridiculous -75c temps with dryive i can promise that). From the thermal capacity that responds 2.5kg copper container so its kind of brutal. There will propably be improvement in internal desing. Holddown will be from upper side of container.

[Dani]

Sick container, very nice and effective cross in the middle of that "thing"
How thick are that wall and bottom?
..and for whole container?

Is bottom like:

|  |
|  |
\_/

or

|  |
|  |
---

[Vapor]

And how would it mount?

[Dani]

Holddown will be from upper side of container.

Nice avatar, you should use it too

[ilkkahy]

Is bottom like:

|  |
|  |
\_/

or

|  |
|  |
---

Its like the first one so there is little "cone" in bottom to improve the heat conductivity. Walls are 7,5mm and base is 10mm. Container outer diameter is 75mm.

[gloatlizard]

wow, your light alluminium monstah will handle all the load you want and keep some crazy cores at very low temps for that highly effective design and so high surface area
great job friend

[ilkkahy]

Well the internal surface area in bottom is about 110cm^2 but because the cross is solid too it will improve heat conductivity pretty much from cpu to pipe compared different bottom so i think this will perform very well. I will put cross section drawings here too when i get to the computer where they are.

[k|ngp|n]

I like the base

[k|ngp|n]

Any updates on this beauty ilkkahy?

[ilkkahy]

I straightened the container with lathe to 74,5mm because it was about 77mm to 80mm cone before. The operation reduced mass from 1100g to 850g. My copper containers have been about 1100g-1750g heavy but copper thermal capacity is 0,387kJ/(Kg*k) where aluminium is 0.900kJ/(Kg*K) so from that attribute it still responds 2kg copper container (naturally it stil doesnt conduct heat as well). Here are few pics and the original plan of the container:





Oh and by the way 110cm^2 was incorrect. I dont know what i have been thinking. Real value of bottom surface area is about 70cm^2 (flat bottom 30cm^2 for example).

I try to get some dryice this week to test out few containers including the old heatpipe container i made ages ago which have never been tested. Here is a pic from the 160w heatload tester:

[gloatlizard]

its fkn good lookin' with the insulation,damn!

[ilkkahy]

Okay the research with four containers has now been made and the results were kind of embarassing for my clever little "innovations". The aluminium container turned out to be total piece of junk. The heatpipe container didnt work with heatpipe either (just as normal container). Im pretty certain i have just let too much pressure out when checking it. Reason for aluminium container not working is propably due the aluminium alloy since it is made by casting. Its obious that the model would be effective if right material were used (copper or right aluminium alloy type). Even though things didnt work out as i planned im actually still in good mood that i have finally made this test that has been in my mind for ages

Here are the pictures and results:

From left to right: heatpipe container, "half-solid", "iron cross", rev 1.1

Rev 1.1 gave -23c

Aluminium container gave totally crap +9c

Half solid container came out to be the best with -27c

I dont have picture from the heatpipe container at use but it gave -20c and unfortunately i didnt have iron cross for the test (i sold it away week ago). I think it might had been the best of those.

Oh and the heatload tester worked like a dream. It is badically just 200w 330ohm power resistor in 230v line which puts out 160w heat. The resistor is on the back side of small copper piece and the thermal probe is inside it. Naturally it gives hotter temps than same heatload giving pentium4 system because there is copper plate between heatload and cooling element and the heat output is being "pushed" from 7cm^2 to 3cm^2.

[ilkkahy]

I made this "rod container" yesterday:






And today i tested it with the heatload tester:



So the basic idea with this one is that the rods go trough the base connecting directly to cpu. However in this testbench the "die" connects only to centre rod so this container may even benefit from heatspreader a bit because then all five rods would get contact to cpu.

Anyway it gave -29c. I was kinda hoping to brake -30c barrier with this one but its still okay i guess The rods are 12mm in diameter and 100mm high. If you look this picture you can see the construction:

http://koti.mbnet.fi/ilkkahy/ln2&co2...tainer%203.jpg

[gloatlizard]

nice contaiiner, nice temps and nice machining work



























ok now send to the address that i'll send you with a pm 200€

[ilkkahy]

ok now send to the address that i'll send you with a pm 200€

offtopic: This is some Italian propaganda. Real price for this prototype was like 1/4 of that + shipping.

[gloatlizard]

wtf italian propaganda??

[k|ngp|n]

I love the one with the rods..

[Thrilla]

The alu cross looks pretty efficient.
btw what's the brand of that thermometer?

[ilkkahy]

The cross in solid container would work nice if only correct material were used.

The thermometer cost something like 5e so it isnt exactly Fluke. Anyhow 0.1c accuracy is more than enough and it seems to give somewhat correct temps so far. The downside is very narrow temp scale (only -50-+70c) and slow updating of the temps. The big varying with the outside temp is because all the tests have been made with window open (i dont want to sufficate in co2 gas).

If i update for proper thermometer some day (for ln2 testing for example) i still have the basic -21c brazed container for calibrating check. If there is difference the tests made with old thermometer are still valuable and can be compared to each other since the testing setup has been the same.

I dont think even solid copper container will get that thing below -50c with dryice but i would be really exited to see it. Anyway i will try it some day too.

[G H Z]

Nice stuff Ilkkahy. I have never seen anyone test containers on a test bench before so kudo's for that. Would be interesting to see where a 3" Mousepot, Kingpins unit or any solid copper container stacks up in comparison.

Can you borrow one to try it?

[Thrilla]

hmm maybe I could get around to testing containers on a bench, I got a 150watt and 50watt in the makes.

btw ilkkahy what did you use for the machining?

[Kaffebord]

I like it. You are one of the few (if not the only one) that really put documentation with your work. And I also like that you describe the sience of behind the making of your cooling gear.
I am sure of that your creative thinking and innovative design will put you at the top someday.

[ilkkahy]

Thanks mate. Here is bit better discription about the very simple tester:

And the resistor is in this pdf-file:
http://www.elfa.se/pdf/60/06083901.pdf

I also have 680ohm resistor which would put out 78w of heat in 230v line but i think 160w is more suitable for dryice containers. With 78w resistor tester gave about 60c temps with intel stock p4 cooler and with 160w temps would rise above 70c very fast.

Anyway i may get load of containers for testing from Dani in the summer so there will be more action. One solid container is a must.

[ilkkahy]

btw ilkkahy what did you use for the machining?

I have drill, lathe and the basic tools everyone got in my use (saw, hammer.. you get the point.) Mill and some sort of machine saw would be nice due the nastiness of copper but i dont have em in my use.

[Dani]

Anyway i may get load of containers for testing from Dani in the summer so there will be more action. One solid container is a must.

Yea, I will send it to you, but first I most isolate my container with polyurethane, it is'nt easy
Polyurethane isolation