What does an ideal block have?
 

Well...I feel like designing and possibly making a waterblock. I dont know what to look for either. I see "impingment" and "turbulence" and I am afraid. I know how to do turbulence but not impingment, not without some kind of restrictive connection. What makes a good block, and how can I design a block that should perform well? A related question: Where's the best place in the US to get a good piece of copper stock, and whats considered "good" copper stock?

What kind of equipment do you have to make this block?

The best place to pickup copper is here:
http://www.onlinemetals.com/index.cfm?affiliate_id=302
because it subsidizes pHaestus test bench (see link on front page).

Way to cut it down to the basics JD!

I'd suggest that you read some of the better threads in the Water Block design forum right here: it'll all start making sense soon enough (yeah, it'll take you a few days...) ;)

There is 2 ways to design a block:

1) Around the tooling available to you. (What I do)

2) Design it and hope it can be done. (Bad idea IMO)

So I ask again. What kind of tooling is available. Once that is known we can lead him in the right direction. I can tell you right now "impingment" isn't going to work without decent tooling. I good design is only good if it can be made. Otherwise it is just a drawing of theory. Completely useless IMO. Theory is only usefull if it is tested. To many designs on this forum as it is and not enough follow through on them.

For a quick hint though, impingment IS turbulance in an efficent manner. Search the user name Cathar and read his cascade thread.

uh what do I have? not sure. In my garage theres normal handtools, most useful for this is probabily a dremel. Available to me easily is a drill press, which I can kind of use for the big things. Harder to find but possible to get to would be a CAD machine, I know what theyre called but the name slips my mind at the moment.

I'm probabily better off with designing around me. I saw (gotta find the link) an article where they made a waterblock out of the bottom of a soda can and an alum plate, and it performed not bad. I want to be able to do that but make it perform WELL. This is mostly for the fudging-around factor in any case, but if I stumble on some awesome, easy to make design, I would love to have a Kronchev block ;)

Look up #Rotor.

You could design it out of a 1U copper heatsink easily... I did it with a drill press and a dremel.

what I was thinking of back in the day before I even had the money for w/c, was to make a small socket 7 waterblock out of a socket 7 heatsink, basically by soldering on side panels to make it a tight block, and adding an input and output barb. I ran into the wall when I realized I have no idea how to solder big metal plates together :D so can you elborate on what you did a little please?

im searching around and cant find a picture of the #rotor block...google was no help either. can someone hook me up?

Here is a procooling review of a rotor block.

http://www.procooling.com/reviews/ht...view_-_4_2.php

yeh I JUST found that

i more of wanted a good picture of his pin array than to see how he performed...he even says its for TEC's and doesnt have things like an input over the die that I would do. I'm hesitant in putting two outputs but a TDX-like design is so ugly to me I might have to. Finally I need to think about base thickness, not sure what to do there. I might just get 0.5 inch thickness and shave it down thin over the core (possibly on the order of 0.1-0.2, have to research that more). The drill press pin array is looking like what i'm gonig to go with, maybe ill try to get slots in it as well. maybe for my very first one ill just replicate a white water. I really dont know, ive only been thinking about this for a day :)

I didn't put an inlet over the die on one of my blocks which competes with the TDX. ;) It is also based on a #Rotor design.
http://www.customcooledpc.com/myWB/010/001.jpg
http://www.customcooledpc.com/myWB/010/002.jpg
http://phaestus.procooling.com/temp/wbcomparo2.jpg

I assume your block is also less restrictive than the tdx... because it doesnt have an accelerator nozzle?

Actually no (well at least from whatever nozzel pH used in the test above). The reason the block performs so well is because the restrictivness is concentrated over the area the die is. This makes a lot of turbulance and volocity through those channels. I actualy didn't realize what made it work untill AFTER I made it and it was tested by pH. I put zero thought into this design. Took 10mins to think about it and 10mins to draw it up and then I milled it. All done in about 3 hours. :D

so do you think if i moved the input to over the top of the die, then put outputs on the side, itd perform even better? also did you make it, #rotor method?

It would be very difficult to put the inlet in the middle and take advantage of all the channels. The channel area is over 1" wide. A normal Barb is 3/8" to 1/2" ID and that only covers 1/2 the channel area. It most certainly IS possible though and I already have it drawn up and the GNC code ready to mill the peice out. :D Just need the block back and time to do it. I am expecting little performance gain though.

I made it with my CNC mill but it uses the basic concept as the #Rotor blocks. Drill holes and connect them. I used 1/8" holes 1/16" apart and connected them with a 1/16" endmill. You could drill the holes with a drill press and connect them with the dremel. Wouldn't come out as perfect but that doesn't mean the performance would suffer. Sometimes inperfections make the block better.

Do you have any plans to market your design? Cause if you dont, I might :D
Hey, everyone else is doing it, I just wanna fit in :D

If I manufactured and marketed it I would be just as guilty. :D

I dont see why people keep complaining about "stealing" other peoples ideas...

Really, waterblocks can only take so many shapes and be decent performers. You cant really blame honda for building, say, a car with a v-8 engine. It has been done many times before, but could you really accuse them of stealing a successful idea? There are only so many practical ways to build a waterblock and have it perform. It is the differences in execution that should be put under the chopping block for consumers. Even if another person/company utilizes an idea that has been established by another person/company, is that really wrong as long as they impliement it in a unique way?

This is really the best way for the consumers anyways. If a single producer manages to win a patent on a design, it is likely that there would be no competition for that particular unit and therefore leaves the consumer with the short straw. If the producer fails to implement the idea in a way that appeals to the public, then it is a total loss and the possibility of owning one is basically dead for the average consumer (a very bad thing for me at least :)).

I see this kind of thing happening with Cathar's cascade. So far, I think that he is the only one who is producing that particular style block. Because he cannot mass produce these blocks (yet :D ), and because the cost is too high for the average joe to buy one (not to mention there arent any left), we are left with a demand without a supply. I'm sure if DTek were to begin making these things, people would be all up-in-arms about stealing Cathar's idea, but I guarentee those same people would be buying them like crazy.

A similar predicament happened at the beginning of microsoft. The original creator of DOS had a great idea, but no way to market his software. People accused Gates of being a thief, but really what he did was a great thing. Now we all have DOS and everyone is happy. Catch my drift kinda...sorta...maybe I'm just thinking through my keyboard again.

i dont remember hearing a huge stick about the vtec knockoffs so youre right


as for the blocks, if i stumble on some really really good design ill gladly post scans of it, pictures, etc so people can copy it. I just want it to be called the Kronchev Block, much like its called #Rotor Pins :D

i may or may not have access to a CNC mill, im trying to convince someone who has access to one to etch a block for me but it seems people who have access to tools are jerks about letting other people use them :rolleyes:

i dont question your measured dT, but I just wonder how well itd do on a small die processor like an AMD. also is it wide enough that it hits the pads?

That test was DONE on an AMD XP1700. Read pH's test bench logs. The Block itself is 2" wide and 3" long. Yes it most certainly covers the pads. 2 more channels and it will be a TEC block (design in the works). I designeed it around AMD CPU's as that is all I use.

People are jerks about using mills because they are very expensive and take a lot of time to learn how to use correctly. Not to mention thay can take a finger off in a second and no one wants that liability. :D Tooling is very expensive and easily dammage aswell. The only person I would consider using my mill is my buddy that is a design engineer. He showed me how to use mine properly. :)

I have been pondering buying a larger mill and make blocks spacifically for people like you who come up with a design yet have no way to produce it and be a USA manufacture and distributor for people like Cathar with the Cascade. I am finding the market to be to low though. Just not enough people interested in WC'ing to make it worth while financing a $20,000 mill.

Manufactures BUY a license from the patent owner to use the technology. That is why we see various versions of everything and no one is getting sued to often. Intel and AMD both buy rightes to use each others tech in the CPU's. GM buys tech from Toyota to use in their's ect... If water cooling became mainstream then we would have to buy tech licenses aswell to copy the design. If the white water was patented Cathar could easily sue and win many cases. Problem is Cathar would actually LOOSE money as there is no way he made enough money from that block to even cover the patent fee's let alone the court costs.

Water Cooling is a very small market, a lot smaller than some want to belive.

jaydee: Wouldn't that block work better if you reduce the fin area to 14mm * 7mm, right in the center, so there is even more turbulence and all of that turbulence right over the die?

i didnt want to use it, i asked him if he would do it for me and he said no. hell i wouldve paid him to do it if he wanted. ill keep bothering him and eventually he will :D

sounds good about your block. i might mimick that design. is there a page anywhere where you/someone explains a little more about it?

[technicality]AMD and Intel have an agreement to share technology[/technicality]

;)

Yeah, one agrees to pay a fortune for the privilege to use the others tech. :D They are not giving their tech away free.

Oh I see. That is a little odd. I paid $1,800 for my little mill and don't mind using it for others. Gives some justification for buying it.
Click the link to my site in my sig and then the Lumpy Channel link on the site.

ill give you 30 dollah to make me one ;)

I would do it if I had the time. My mill is setup at my grandparents shop 130miles (one-way) from here. I live in an apartment in the city. Costs me $30 just in gas to get up there and back. :( I also travel for a living. Headed out of town for a couple weeks on Tues. Sigh....

Are you in the USA?


Here is a pic of the drawing for the Lumpy Channel 2. Minus the top peice. The only difference is there is 2 less holes drilled on each channel. I think that will make it a little less restrictive and shouldn't hurn performance as those holes are past the point were the heat from the CPU will spread.