New water block idea...
 

I can't freeking sleep because I keep getting to many ideas on what I want to do with my site and water blocks ect....

Anyway here is a thought for you all.

I know there is a lot of talk about not mixing Aluminumn and Copper ect. ect.... BUT there are a few blocks out that do do this and don't seem to be that many problems. Also if you use distilled water and watter wetter it shouldn't be a problem.

Anyway here is my idea. You all seen my current blocks here: www.custom-cooling.com. What if I where to take the Aluminum block and instead of using an aluminum top I used a Copper top. And then i fflipped the block over to where the copper was in contact with the CPU. That way I would have the heat absorbsion of Copper and the quick cooling of Aluminum. That way I can have the best of both worls and still keep the price down.

And instead of solder or epoxy to hold the copper plate on I would narrow up the channels a bit and use a rubber O ring and screw the plate on to seal it up. That way you can take it apart and clean it in the future if growth ever does become a problem.

It is late (12:30am) and I may not be thinking to strait but what do you think. Is it worth trying???

That might work, but would the copper top and the alum. block make good contact? Also, if the water to the block does not get supplied due to a bad pump or leak, the O-ring would melt due to the rising temperature of the CPU and spill out all over. (The XPs should be no prob, they got that built on Temp. probe that shuts the system down at a specific temp.)

thats just my $.02....

The prob is that you don't have the channel walls supply heat to the water so you massively lose surface area. Also the "quick cooling of Al" is a myth, convection (which is what we're talking), is entirely dependent on surface area and fluid dynamics. The only advantage copper has in water blocks is that it spreads the heat from the cpu to more of the water due to it's higher conductivity.

Just to be fair there aare allot of otherdesigns that use screws, orings, and what not to fasten the blocks together. Jaydee #rotor has been makeing them for quite a few years and he uses silicone and screws to fasten his blocks. He even says that he has a hard time getting those appart to clean them. Someone at another forum (don't have the link) stabbed himself trying to get something like this appart.

I have talked to the guys that own dangerden and they said that they have seen the cpu's under pump failure heat up enough to melt the solder off thier blocks. If this can happen with solder I don't think it much matters if you use solder or not. Besides I have seen a picture of what can happen to your block with tap water.
http://www.voidyourwarranty.net/revi...de/index3.php3

Being able to clean this type of stuff out whould be good.

If that O ring gets that hot the computer is on fire. They use O rings in car engines that withstand 300F. If your comp gets that hot the CPU will die and the comp will shut down. Then the water will just dribble a little because it isn't running or non of that would have happened in the first place.


Heat transfer will still be ok. Metal to metal screwed together at 100PSI or so will still get the job done. Most of the cooling happens on the bottom of blocks anyway.

Fast cooling a myth??? Have you ever held a aluminum block in your hand and turned the water on? It gets cold instantly, as to where Copper takes longer.

My head is killing me, I ididn't sleep at all last night.:(

Thanks for your input. I will work on it just to see if it will work.

I can first hand vouch for the solder melting in Dangerden blocks. In my case it was a run away TEC (uncooled Peltier due to pump failure), but regardless the solder melted, killing my video card along with the already dead CPU and MoBo. Oddly enough my silicone hoses survived the 300F heat intact!! Literally you could not tell thy had been exposed to the heat. Personally I think soldering your blocks would be better, however if you were to use a silicone O-ring that can withstand those temps in case of a pump failure or human error (forgetting to start the pump) you would be OK. I have seen the results of someone forgetting to start a pump with a T-Bird and a Spir@l, the acrilic warped under the heat and caused a catastrophic leak. The O-ring was intact, but the acrilic failed. If you had a metalic top you could probably avoid this as the Al probably wouldn't warp like a Spir@l would. Just my $.02. Let us know when you've got these blocks ready, I am very eager to give one a try.

the only concern I have with that is heat transfer from one part to the other. I'd also prefer to have solder, not epoxy/screws/o-rings.

If the solder melts, lets face it, your computer is ****ed regardless

I think nothing at all would actually be better than solder to transfer heat. Unless it was 100% silver or something. Also the solder would be right there at the heat source. This I wouldn't feel to solid about. Nor would I want epoxy right there. I will probably scrap the idea for now anyway.

I got bad news today, they may want me to come back to work by Febuary. That screws up my whole situation!!!. If that is the case I will be back on the road for weeks at a time as the job requires ne to travel. I was hoping for a few more months off to get things setup to where it wouldn't matter if I was here or not.

If that happens then my website will turn into a information site on how to do water cooling. It may anyway. I will still do blocks in my spare time though.

yeah, I wouldn't like solder right above the cpu, but as I said before, if the solder melts, you still have major problems regardless.

maybe you should look into silver solder, or other solder compounds which have a higher melting value, just to make the people happy

Normal 60 / 40 electrical solder is melting point of 183C, if your cpu temp is that high you're ****ed anyway.

Yeah the solder I have been trying to use melts at 409F!!! The computer will be the least of your problems at that temp.

...which is what I have been saying all along. if the solder melts on your wb, you've got major problems anyway

Thats Bad Physics. Al does NOT release heat faster then Cu!

Again say it with me: in a closed system, specific heat capacity does not effect the overall thermal resistance of the system.

However Al is much easier machined, is lighter, and cheaper. So while Al is not a bad idea, it will not help with cooling performance compared to Cu assuming all other factors are equal.

And I personally would not buy a block that has both Al and Cu in direct contact with the water. If you go that route, I would have to see that at least some annodizing or some such precation was taken.

Fear not. Idea tossed!!!

hey Jay, I love bouncing ideas off of people who enjoy trying new things.

What if we cam up with an aluminum base for a block, but have a 1.5 inch circle cut out where the contact would be made, then have a screw in copper incert. If the threads were done right they wouldnt leak, and have really good aluminum to copper contact that way. Then inside the block, have screw in type of incert poles that could be customized to different configurations. Say just for example, an open pocketed block, with small holes (not drilled all the way thrue the bottom) with threads in them, then take copper or aluminum dowles that were threaded to screw in the holes. U could use as many, or as few as you would want, to get the best concact area and flow rate.

Just something I thought up a while ago.

I like the idea of copper and aluminum mix blocks, its easyer on the cutters and the cost. I mean why have all this expensive copper in a block, when it isnt doing jack. Have the hottest parts copper, and as you work your way outward away from the heatsource, aluminum would work just as good.

Also, I sent in a copper Spir@l and an aluminum Spir@l, I am very excited to see if the copper is in fact any better in this design.

what about having a copper base, then screwing/soldering in a fairly complex Al channel?

Al is much easier to work with, so it'd be better for the complex channel, and the copper would allow you to have good transfer properties

That sounds pretty good. I like the inside dowl idea. it would take some time to make but if it worked......

People are all worried about the copper and aluminum causing problem. If they just use water wetter and distilled water all will be good.

Hell, I have been running a copper tubed rad and an aluminum block for months with city tap water with no problems at all. The water gets a little discolored after about a month but I just change it then anyway. I took one block apart and there was nothing in it.

The problem is getting people informed on what to do and how to do it.

Is complex good though? That is the question.

presuming you can transfer the heat to the block, you are interested in surface area and flow rate. So if you can come up with a parallel tube design or something with a lot of surface area, but still with little flow resistance, you should have a good block

My design engineer was talking about this very thing the other day. I will run it by him next time he comes over. Luckily we are good friends. We both graduated in the same class. To bad I didn't go to engineering school with him. Oh well still making more money.:D

Whait till you see my new block tomarrow I thnk you will be pleased as it is high flow and yet should have good surface area.

I should be up there about 12 noon.

apparently my EXTREMLY simple block does the job cooling my duron perfectly .... it kind of a bigborde singelchannel block design.... keeps me below 30c fulload w/ duron800@1000, 1.92v...

so whats my point here ? complex is sometimes not much better than really simple, tried and true design.

dan,
That single pass block I am sure works really really good with the lower heat processors, But what we have to do, is think in terms of here and now, and later in the future. Sure the XP processors got a little cooler, but what is next? I am sure they will crank them up more and make more heat too. I would be interested in seeing how your block would work with a much higher heatload.

Unfortunatly I don't think so. From what I have been reading when they shrink the dies with the .13 process it not only will use less power but run much cooler. The New P4 2.2 is down to like 50watts as compared the the 2.0 that was around 70watts. Thats just from the die shrink. From what I understand the AMD chips will follow suit.

This is a good thing though as we will be able to get the CPU even colder and should be able to OC even higher?!:evilaugh: