"Nazca" - Waterblock Concept

[diff_lock2]

whys the plexi part so huge and thick?

also it looks VERY good, but, what about this design makes you think it would perform very well?

[JFettig]

That is definately a very nice peice you have there! I'm amased how well it came out, and the 15 hours of determination I'd have given up after 4 lol

Whos testing it?

Jon

[Captain Slug]

Originally Posted by diff_lock2

whys the plexi part so huge and thick?

The top has to be atleast 1/4" thick to have enough threads to hold fittings. But the top also had to have the widening inlet and a 1/4" tall outlet channel. A thicker top will also make the block more durable. Both the top and the middle plate are made out of polycarbonate, which is far less likely to stress fracture over time than acrylic. Polycarbonate can also be solvent-adhered to itself, which makes putting the block together a little bit easier for me because it means I can use one less O-Ring.

Originally Posted by diff_lock2

also it looks VERY good, but, what about this design makes you think it would perform very well?

The design was develop from my intuition, which is based on several months of research into the waterblocks that others have made as well as alot of constructive feedback from this community and others.

The nozzles will be "spraying" both the base and the walls of the channels in the center square inch of the block. The channel walls (or fins) are congruent from the center of the block, which alot of complicated math and evaluation by many others have shown that it should be more effective than pins. And I'm using the dimensional ratios for channel size, channel wall thickness, and base thickness that others have determined to be optimal.
So I'm borrowing alot of expertise from Cathar, BigBen2k, D-Tek, JFettig, and definitely davidzo. All of whom I would like to thank for being so constructive.

Originally Posted by JFettig

Whos testing it?

Not sure yet but I've e-mailed Joe Citarella to see if he might be able to.

[Talcite]

Great job man. I'd love to see the test results. How big is the actual square center?

[JFettig]

For some reason, I couldnt get any of those guys to test my blocks, I know it was partially becuase I was considering selling them, I offered them some machining services and they didnt take it. I went to phaestus(sp?), and joe, maybe another...


Jon

[Captain Slug]

Originally Posted by Talcite

Great job man. I'd love to see the test results. How big is the actual square center?

Here's a picture showing the nozzles and 3 squared dimensions. Center square is 1 inch, then 1.125", then 1.25". Non-IHS processors would fall directly within the nozzle pattern.

Originally Posted by JFettig

For some reason, I couldnt get any of those guys to test my blocks, I know it was partially becuase I was considering selling them, I offered them some machining services and they didnt take it. I went to phaestus(sp?), and joe, maybe another...

Darn. I hope I have better luck. Otherwise I'll just have to ask you to test it.

[JFettig]

lol, sorry, it sounds like you'll be testing your own block, I haven't had a computer to watercool for a couple years, I now have a mATX computer but I havent had the desire to make a mess of it. I still do enjoy watching people make cool stuff.

I'll take a look around to see if I have anything that you can compare it to though if you want to test something along with it, unless you have a retail waterblock somewhere?


Jon

[jaydee]

I still have my test bench pretty much setup. I even milled up a fresh die sim. Just haven't had time to use it. If you don't have any luck getting it tested I can put it up against a Cooltechnica/AquaExtreme White Water SE, MP-05 SE, MP-1 CPU, IceFloe and MP-05 Pro.

Tests will be somewhat limited to my test bench though as with any test bench. I have it setup now with 1/2" x 1/2" die area at 72watts. Will test multiple flow rates (if possible). Only thing I have yet to get setup is the Foxboro dP meter as I still have no way to get a reading from it. Need to pick something off ebay one of these days.

Anyway let me now if all else fails.

[Captain Slug]

That would be very helpful jaydee because the only block I have to compare with is a cruddy old maze 3 that needs cleaning. I'll let you know by this weekend.

[Talcite]

captain slug, the jets don't fall on the IHS edges? Heat transfer from an IHS is primarily along the edges of the IHS, not the center. Is your WB designed for cpus with no IHS?

[Captain Slug]

It's designed for both. Not all of the heat transfer on an IHS processor is around the edges. So you can't concentrate all of the turbulence on just one area.

I over-worked all of the muscles in my right forearm while lapping the base so I haven't been able to do much since Wednesday.

[jaydee]

Originally Posted by Talcite

captain slug, the jets don't fall on the IHS edges? Heat transfer from an IHS is primarily along the edges of the IHS, not the center. Is your WB designed for cpus with no IHS?

I don't follow this at all. All thermal illustrations and IHS tests I have seen show heat transferring pretty much right through the center of the IHS with heat spread slowing way down as it reaches the outer edges. The efficiency of the cooler will dictate how much heat makes it to the edges of the IHS but any good water block should pretty much keep it off the edges.

[Talcite]

hmm really jaydee? I seem to remember the heat being transfered from the edges of the IHS... oh well. I don't remember where I read it at all. It was a thermal camera or something. I really can't remember.

[Captain Slug]

http://www.overclockers.com/articles1284/

[jaydee]

Originally Posted by Captain Slug

http://www.overclockers.com/articles1284/

That pretty much backs up my point of view. Them more efficient the cooler the less heat should make it to the edges (theoretically).

[FooTemps]

Originally Posted by Talcite

hmm really jaydee? I seem to remember the heat being transfered from the edges of the IHS... oh well. I don't remember where I read it at all. It was a thermal camera or something. I really can't remember.

When you apply physics, it makes no sense. If the IHS is thermal epoxied/pasted to the core, the heat should radiate from the core out. Not from the edge towards the core.

[BGP Spook]

http://www.overclockers.com/articles1284/

I don't like his numbers, though I believe his analysis.

[bobo5195]

artical 1284 has some floors but the heat transfer is easy.

Google biot number. Increse the cooler performance your going to change the biot number.

[BGP Spook]

bobo, thank you for the biot number reference but I was refering to page three of artical 1284 and the inconsistency of the die temp vs. center temp relationship.

Look at the "IHS Temp Pattern - Die Simulator" table.

Could be poor calibration or a data entry error. Maybe the sensors are just inaccurate.
It would be nice if he had provided some attempted explanation.
What little(very little) experince I have had working in my college's chemisty lab tells me he owes at least a note addressing the discrepancy.

I did not mean to hijack the thread, I am sorry.

[Captain Slug]

No biggy. This hasn't been discussed in detail anywhere else as far as I've seen. It's strange how much myth there is circulating around concerning the thermal load differences between IHS and non-IHS processors.

The block will be finished on Tuesday.

[Captain Slug]

Well, the block is finished. I fixed the inlet, fused the top and middle plates, and thoroughly cleaned the base using white vinegar and a toothbrush. I'm still trying to figure out how to go about testing it though.

[Captain Slug]

Time to try mounting everything together.

[Colt357tw]

wow, good work to get it done!

[BGP Spook]

It looks like it was well made. Good job!

[Captain Slug]

Jaydee is going to be kind enough to test the block for me against several other modern blocks. Which is good because I have a wickedly time-consuming PCB to get to work on.

http://www.ocforums.com/showthread.php?t=394139