New Block REV. 1.0 (pics)
 

Alright here we go. First off I am going to aplogize for the crappy pics. I had to use my cheap webcam as I still cannot get the good one to work.

Here is the top. 1/4" acrylic. It took 50 seconds on the lazer to make. I couldn't put the mounting holes in at work today as I didn't have time to re-lookup the deminsions so I will do that later if I decide to make this a working prototype. REV. 2.0 will use the base for the mounting instead of this top peice.
http://www.customcooledpc.com/chill1/002.jpg

Here is the middle section with the nozzle cut into it. Will take some testing to figure out how many holes to use. These holes are 1/16" and the material is 1/8". This peice took 39 seconds on the lazer to make.
http://www.customcooledpc.com/chill1/003.jpg

Here is the bottom peice. I needed to make this so I has a place for water to flow. :D It is 1/8" thick and took 20secs to cut on the lazer. This peice will be replaced by the base in REV 2.0. As I will mill out the ovel.
http://www.customcooledpc.com/chill1/004.jpg

And here is the base. Right now it has 1/16" pits drilled into it right over where the core will be on an XP AMD CPU. I have the nozzel aimed right at it and should help in the heat transfer. REV 2.0 will still have the same identical pattern but in copper. This will take a lot of testing to know if more pits will be needed/useful.
http://www.customcooledpc.com/chill1/005.jpg

How it is put together.
http://www.customcooledpc.com/chill1/006.jpg
http://www.customcooledpc.com/chill1/007.jpg
http://www.customcooledpc.com/chill1/008.jpg
http://www.customcooledpc.com/chill1/009.jpg
http://www.customcooledpc.com/chill1/010.jpg

Looks good. I think a few too many holes though. Too close together. This causes the streams to interfere with each other and messes up the impingement activity.

Not sure on your separations. You want the jets to be about 4-6mm high from the surface they're hitting, and you want them to be no closer than 4mm to each other from edge to edge (not middle to middle) to avoid cross-interference. For an AthlonXP sized die, that pretty much limits you to a 2x2 grid of circular jets of 3mm in diameter, or a 3x3 grid of circular jets of 2mm in diameter for common pump use (<2m head height). For higher pressure pumps (>5m head) drop the jet diameters to 75% of that stated. Base-plate should be about 2mm thick (in copper).

Note, these are just my personal thoughts on the matter, others may differ.

Thanks for the input, very much appriciated! I am going to use this in REV. 2.0. I am not fluid in metric so I will have to do some conversions. I think I am headed in the right direction though. I do not belive you need big winding channels with lots of copper around them to make a good block. I am pretty confident I can turn this into a very simple to make (at least for me) and decent performing block. It may take until REV. 40.0 to do but I am good with that.

I am going to put this one together and hook it up as is and see what it can do. REV 2.0 should be a large step forward as it will be in Copper and several modifications to this one. I am going to build one per every Revision being they are so easy to make. It will get a little harder when I mill the oval out but I think I am going to make that oval much smaller than it is now. There is to much area for the water to flow. Sounds backwards I know, but I think it is right. I want the water to enter and leave as fast as possible so less area for the water to move around in the better I think. We will see. :)

Feels good to be building blocks again! Didn't realize how much I liked doing it. :)

And for anyone wanting a conversion source this place rocks. http://www.onlineconversion.com/

I'm very fluid in both imperial and metric. Result of being brought up by a father schooled in imperial just as metric was introduced.

Here goes:


Not sure on your separations. You want the jets to be about 5/32" - 1/4" high from the surface they're hitting, and you want them to be no closer than 5/32" to each other from edge to edge (not middle to middle) to avoid cross-interference. For an AthlonXP sized die, that pretty much limits you to a 2x2 grid of circular jets of 1/8" in diameter, or a 3x3 grid of circular jets of 5/64" in diameter for common pump use (<7' head height). For higher pressure pumps (>16' head) drop the jet diameters to 75% of that stated. Base-plate should be about 5/64" thick (in copper).

;)

Love the lazer cutting:drool:, very swish.

pump some water through it without the base on, that way you can check out the jets, see if they interfear with each other or not.

i like it alot.

from a practical point of view, this might make a cheap block to manufacture in high nrs, since the Cu - Al cost would be relatively low because most of the block is plexy. practical: u could make different middle pieces for , lets say cathars line of thinking, different head pumps.

too bad getting an objective means of comparing blocks is nearly impossible. i would really have liked to know the performance of the block.

congrats.

Doesn't work like that. The intereference comes from the back pressure of the jets striking the surface. If they are too close what happens is that the water from the inner jets really has to "fight" to get to the plate. This results in a dramatic drop in jet efficiency. If they are too close, then it becomes little better than a single inlet approach.

The jets need room to impinge onto the plate without the flow bouncing off the plate smashing hard into adjacent jet streams.

You might take a look at amd cpu core not being in the center of four mounting holes, I think that is almost necessary with your jet impingement action. The difference is not that small, but I don't realy have precise number yet, I'll measure it tomorrow (I need it too)

That would be great morphling! I have a spell of food posining so I may not be back for a day or to. :(

At least I am getting all my preying to the porcelin gods done. :( :rolleyes: Not doing to good....

lazer! lucky bastard...


:D

the offset is btween .05 and .06inches, I think its .0539inches or so..
looks real good, can you lazercut aluminum? make it out of aluminum instead of acrylic?

Nope, it is only a 80watt lazer. Can't do any type of metal with it. I am going to hook up the nozzel I have done tonight and see what it looks like. May end up going with some other type. basically I just need to water to be accelerated into the pits of the base. This type of nozzel I don't think will do that that very well inless I have rediculous amounts of flow. That would kinda defeat the pourpose of a budget block.

Lots of little holes that equal the same cross section area as a single bigger hole will have much higher head loss for a given flow. This is the other consideration along side that which Cathar has already mentioned.

You may also try cutting those holes at an angle off vertical such that each jet is already starting to turn in the desired direction towards its exit. If you manage to get this right, you can reduce total head loss while also reducing interference from adjacent jets.

The only real downside I see is that you don't presently have any stiffness other than the thickness of the baseplate. This will limit the minimum thickness you can use. When you get around to optimizing further, you may wish to consider some "struts" to gain the added stiffness available from the top.

My rough guess for PD based on dubious extrapolation of BIlla data* .Equated the PD in a 0.25" Nozzle to that in a 14" length of 0.25" ID pipe then used this calculator http://www.aps.anl.gov/asd/me/

http://www.jr001b4751.pwp.blueyonder.co.uk/JetPD.jpg

* http://www.thermal-management-testin...vs.%20flow.gif

I hooked the nozzel up and flow wasn't much an issue but the watter looked the same as if no nozzel was used at all. :D It just all combined like they said it would above. :D I think I am going to use more of a jet type nozzel that will increase the volocity over the area. I have a lot of thinking to do about this,

As for the base, REV 2.0 will be 1/4" thick with the oval milled out 1/8" deep. I will then still have the same spacers above making the distance from the bottom of the nozzel 1/4" to the highest spot on the baseplate. Which at the moment is the best I can do with the materials I have. Everything will be bolted to the base using the 6 holes. I will tap the holes in the base to acheive this. I think that should add a lot of strength. The oval will also be much narrower than it is now, probably only 1/2" wide and maybe a slight rounding around the core area. I may cut the modified oval out today in acrylic if I have time.

heres a real handy converter program that i use. Its small and its just an exe, so no installation it convert EVERYthing
You guys might find it handy:)

http://www.joshmadison.com/software/convert/convert.zip

REV 1.0 has came to an end! It will be a non functional prototype. REV 2.0 Already under construction and should have a working prototype this time. It is going to be considerably different than this one. Thats all I can say for now. Look for REV 2.0 Thread Sunday evening. :)

jaydee116, do you have the socket offset in there? .053something offset to one end
and lemme guess? the crappy way of trying to make a jet? do the 1 hole jet

jaydee, what are using (or planning to use) to seal between the acrylic layers?

8-ball

Acrylic cement like IPS Weld-On #16. REV 2.0 may not have the layers though. Probably just an acrylic top and copper base, but I havn't ruled out a 1/8" spacer with or without the nozzel. If I do not use the 1/8" spacer then I will just use a 1/2" thick base insted of the planned 1/4" one.
REV 2 is going to be nothing like anything said/shown in this thread except it will be single inlet/dual outlets, with a drilled base. Thats about it. As for the jet I am undecided if it will be applied yet or not. If it is it will be of a one hole design. Possibly oval shaped. That will be in REV 2.1. :D

The base will look nearly identical the Cathars white water exept no micro channels, in place will be the drilled base. The offset may be included in the mounting. I have to measure it out later. It isn't very much. I need to order some copper stock. I have the 1/4" acrylic top already done but still need to do some thinking on the base sizes ect... Didn't have anytime this weekend as it is my daughters birthday so I spent the majority of it doing stuff with her. Always first on the priority list. :D

Heheheheheh...:evilaugh:

I made the base design today at work while I hade a few mins. I lied, it isn't going to be drilled. it has a pin array with PYRAMID shaped pins. :evilaugh:

The base is 1/8" aluminum (because ofa lack of copper), and the pins hight is .030". The cuts where into to base so they go down .030" aswell. The cutter used had a V shaped tip so that is how the pyramid shaped pins came about. The pins are very small. At the bottom they are 1/16" wide and at the top 1/32".

Bst part is it only took 13mins to do. I will get a pic up one way or another tonight in a new thread called "REV 2.0. Micro PYRAMID Pins". I have to go cut out the spacer plate then get back to real work. :D.

REV. 2.0: http://forums.procooling.com/vbb/sho...&threadid=5559