You need to decide which drawing program you are going to use. In school I’ve learned to use AutoCAD really well so I always use acad. This is as precise as it gets; lets you print in a 1:1 scale. But with a little bit more effort my friend Puzzdre succeeded in making the same kind of blocks perfectly, using nothing more than Corel draw (congrats to you on that one Puzzdre). You can get your hands on any kind of software, but if you need to learn how to use it for months on end before something usable comes out of it, you will shy away. You can use the most modest drawing program, draw it up approximately, print it, and measure it with a ruler and keep working at it until it fits your work piece perfectly.
I would advise you not to make it too complicated for a first timer. In a future article i will explain how to make a moderate performer into a stellar performer anyhow.
These 3 drawings tell the story of my evolution as the skills developed. The first design gets you a block that performs somewhere between a maze2 and a maze3. It is a very low restriction block, and that is also its relative weakness. If you are cooling a very small die area (TbredB) it is slightly ineffective. It would show its strength in cooling a large pelt area though. It is very easy to drill and Dremel.
The second design is a lot more complicated to Dremel and performs approximately on par with a maze3. Notice how the restriction has gone up with a lot less channels the water can use to reach the outlets. I will use this design to show how to tweak it in an upcoming article. It wouldn't be that good in combination with a pelt, but is a decent compromise if you would want the flexibility to cool a small AMD type core and a large P4 core. The thing that makes it weak (and for that matter every commercial block except Cathar's White Water) is that the water at the entry point gets deflected too fast and the velocity of the entering water is too low for it to reach the base plate (hot point). This gives us an excuse crank it up even more in the design after this one, where I will explain the term 'jet impringement' (and I think it was that visionary Cathar that invented/implemented it for the first time).
My final design to date gives the stellar performance i was aiming for from the start. it shaved off another 3°C cpu temp from my previous top performer which was nothing short of amazing. Jet impringement: I can't explain it better than a showerhead. If you take the showerhead off the water tube the water flow won't be nearly as potent as with a highly restricted showerhead. Now imagine a water block on a scale that would make it as large as a human being. If you take the showerhead off you need to hold the water source a lot closer to your body to 'get wet' than with the restricting showerhead installed. For our water block this means that we add an extra layer or two to convert our inlet barb flow into multiple small jets that will have no problem spraying that cool water all over your die area. The blue layer is the plexi plate where the shower holes are drilled into. It shows an area of 12x12mm where the jets will maximise cooling. Does this make sense?
Printing the design -->>