The "Cascade" - mini-cup directed jet-impingement block design (56K warn)
 

Well this is my final conceptual merging of a number of block concepts drawn from a number of sources into an affordably machined design.

Basically drawing upon the White Water concepts, Volenti's directed jet pipes, BigBen2K's cups (rather than dimples), and jaydee116's mass-jet array block, and bringing them all together into a cohesive and (somewhat) affordably machinable design.

Anyway, this is what came back from the work-shop today. It's hard to describe just how fine and intricate the array of jet tubes is when looking at it "in person". The eyes/brain really does have trouble focusing on just what's going on there. Fortunately the wonders of close-up digital camera flash photography gives a clearer image.

Anyway, the block is presently being tested. At present it's testing at about 0.5C worse than the White Water, but this is a fresh application of Arctic Silver 3. The White Water's temps were taken after 2 days of the thermal paste "settling in", and typically 1C or so is gained after an initial overnight "burn-in". This makes testing somewhat slow and painful, but I guess that this helps to give a clearer idea of final performance. I'll find out tomorrow if the design really does exceed the White Water for cooling performance. I have a number of bases with differing hole/cup geometries to try, so it's going to be a few weeks of testing to isolate the best base setup.

Anyway, here's the piccies.

http://www.employees.org/~slf/cascade/cascade1.jpg

http://www.employees.org/~slf/cascade/cascade2.jpg

http://www.employees.org/~slf/cascade/cascade3.jpg

http://www.employees.org/~slf/cascade/cascade4.jpg

http://www.employees.org/~slf/cascade/cascade5.jpg

http://www.employees.org/~slf/cascade/cascade6.jpg

http://www.employees.org/~slf/cascade/cascade7.jpg

http://www.employees.org/~slf/cascade/cascade8.jpg

http://www.employees.org/~slf/cascade/cascade9.jpg

http://www.employees.org/~slf/cascade/cascade10.jpg

http://www.employees.org/~slf/cascade/cascade11.jpg

http://www.employees.org/~slf/cascade/cascade12.jpg

http://www.employees.org/~slf/cascade/cascade13.jpg

Damn, thats really sweet...

How are those jet pipes attached?

It looks to me like they're machined.

Though I could be wrong.

That is one damn sweet block.

8-ball

The whole middle plate, including the jet pipes, are machined out of a single solid piece of polycarbonate.

Whoa... :drool:

did it cost about the same as the WW for them to machine?

Now anybody wanna try and make a copy of that with only a dremmel?

really nice work Cathar!

:drool:

At present it works out to about the same machine time as the White Water. Despite polycarbonate being dramatically easier to machine than C110 copper, there's a LOT more cutting passes that are being done. The machinists were being very conservative though for this prototype though, and there exists a number of design tweaks and a more aggressive cutting approaches to speed it up.

;)

bp thickness? cup depth? any specs? or can you not tell us:D


btw, VERY nice block, that is kind of what I am going after in my block, but my will be...different?

Ice Water?, for a name?...

Do the jets protrude into the holes or do they 'hover' above?. If they hover above: do you think the jets reaching the full depth and striking the BP with 'enough'(optimum?) force?. I'm guessing the BP is 5mm thick?, that'd make the holes 4mm+?...
Do you think it's be possible to make jets that protrude into the 'cups' and still allow enough room for water to egress?...

Nice block anyway!, even if it does'nt beat WW it's a top performer. If enlarged it'd make a good TEC block for that Iwaki MD30RZT of yours! :D ...

I want to create something similar to that...

Cna you give us a little more info, Cathar?

Nice, that's exactly what I had in mind, here.

Looks like it'll be a good performer too!

Yep, look at the 3rd post in that thread - the one straight after your initial posts. I had been pondering it for a while, but your post suddenly gelled it all into my brain at that instant. The block pictured above is exactly what I had in my mind, and would've happened 4 weeks sooner if not for the machinists being overly busy.

Like I said in this opening post, I owe some inspiration to yourself, Volenti and jaydee116, and to washing dishes (sticking cups under the faucet to rinse them out), which initially got me thinking.

Not just yet. I wouldn't like to think that I had all the dimensions worked out just yet myself. I'm still playing, and as I said, I have some different bases to "plug in" to see how they go.

Simple copper machining and injection molded plastic - near White Water performance.

This one looks like an even bigger winner than WW. (particularly commercially)

Why sandblasted?

Make the 'water exit chamber' taller, and go to a single exit barb?

LOL! That's got to be the most original source of inspiration!;)

Is'nt there an optimal height~lenght ratio for a jet?, to long/thin(or short/wide) and it reduces velocity?...

This is a headache block for tweaking by theory!, (at least for me!!) :D ...

Just wondering, how thick is your baseplate right now?

I was thinking of the space outside the CPU die area. Beyond the immediate vicinity of the die, the chamber in the top could get taller and still allow the nozzles to be short.

DIY'ers might be able to make use of a Stimpson eyelet to make nozzles. I'm not sure they like to sell in quantities less than 100K though.

Could you finish the bottom of the 'cups' off with a ballend millbit?, or do you think it'd not make enough difference compared to a [b\/[/b] from a drill bit?...

Is there an optimal shape?, have to get all the old spoons out from the cutlery draw! :D ...

It varies, depending on what one I choose. 3/16" or 1/4".

Will be trying all 3 "shapes".

Drill-bit end (pointed)
Ball-bit end (rounded)
Flat-end.

I suspect that the balled end may be the best one.

Not entirely sure if injection molding could get that level of detail. It's something worth looking into.

Bead-blasted, just because the machine was there.

Am considering a single exit barb. Since it's a prototype and I wanted to cut costs, to save on the programming time we just copied over as much of the White Water program as possible.

I just wanna know how the h&ll they got the jets in the array in the shape of little stop signs,when the holes themselves look rounded??

Cathar, another very fine design !

I love your attention to all the details, like how each little tube's top is countersunk with it's own little funnel shaped entry.

How does the pressure drop of the Cascade compare with your WW design?

Any word you can offer on how your Gfx block is progressing in development?

Again very fine workmanship & design! !:drool:

Yeah, the countersunking was done by hand, hence the uneveness. I've since gone over it my home drill press and they're all very even now.

Pressure drop is about the same as the White Water.

Depending on how this works out, the GFX block will get a redesign.

Cheers - off to bed now.

I'm fairly confident that it could be done, with the right plastic.

The biggest problem I see, is that when injecting the plastic, the plastic is going to have to 'flow past' the tip of the nozzle, in order to ensure that that part of the mold gets filled. This will tend to leave flashing at the nozzle tip which might be a pain to trim off.

Anyway, I think it's worth looking into. Probably $3000+ USD, to get a cheap aluminum mold made. Then maybe $3 to $5 per piece at qty. 1000. (Take these numbers with a big grain of salt. They're based on things I've overheard at work, not on work I've actually done.)

Make and sell 1000 before the cloners swoop in?

Cloners might have their work cut out with this one :)

Without the straws, its relativly simple to make, probably even more so that the WW. I mean, I've got one with simple holes in the spray plate, so has Volenti, and Jaydee I think as well as a few others.

Those straws make it a LOT harder though, I'd never like to say impossible without a cnc machine, but tough, very tough.

Excellent design, very nice indeed.