1. The Rotor design is intended to be made with a drill press and a dremel. It can be done with a hand held drill or a drill in a stand, but you will need a steady hand to do it. If you use a good drill, and plenty of coolant / lube you should be able to make it without breaking bits. The dremel with a cutting wheel is used to break the walls between the holes, creating the pillars. A milling machine will make the process easier, but is not required.

It is worth noting that if one uses a 1/8" mill bit to just cut the channels the way I did, you end up with square pillars. If you use the Rotor method of larger holes with thin channels connecting them, then you get the concave sided pillars which per Rotor are supposed to offer a theoretical cooling advantage.

Also, you should NOT attempt to use a drill press as a mill substitute, it is not designed to handle the stresses, and you run severe risk of damaging your tools and workpeice, and possibly yourself.

2. I made my blocks using a similar pillar design, however I rotated it 45* so that the flow entered the crossing channels from both directions. This makes the water flows collide at every intersection, causing major turbulence.

3. Soldering tubes is NOT going to give as good a heat flow as having pins that are part of the baseplate no matter how good the solder is! Furthermore, only those tubes that are directly in contact with the baseplate will see any real heat, coolant going through the higher layers won't see any heat at all to speak of. Basically it's a bad design.

Gooserider

Thanks for some good tips ;) .

I am buying a 1250 unused Eheim pump, how much is it worth? 40-60? USD?

Home Made waterblock
 

I am almost finished with my 2nd home made liquid cooling system, I have spent (so far) a total of $17.53USD (I only lack a PSU and fan controller). I have a water block made of an old 486 air cooled heat sink. I scavenged some 1/16th aluminum and epoxyed it around the outside of the old sink, then I scored a great waterblock top made of machined aluminium, pics are at http://arapaho.nsuok.edu/~landrynk (main site is down, thats a backup).

I am happy to offer more assistance but tech support wants to CCM this machin, I guess I'll have to take the rest of the day off :)
N8

The hookup I had broke the CNC machine, I don't know what happened...

Anyway back to Ebaying... :p

I'll get the Eheim 1250 at least, and I still have to find a radiator... My case is a bit small on space on the top, but I was thinking the HWLabs Black Ice Pro II Radiator, which is 25mm thick and has 2 120mm fans on it (not included), but with some fans on it it should only be 50mm thick yet perform better than a single 120mm Black Ice Extreme, So do you think this would work good?

As for the water blocks I was looking at PolarFlo since I might be able to find some cheap deals...

Whenever I get anything used I usually figure about 50-60% of new, possibly 75% if it's a really nice item or is high demand. As I recall an Eheim 1250 is around $70-80 new, so the low end of that range is probably about right. Don't forget to include shipping costs, if any, when calculating the price.

The best value is usually automotive heating cores both for price and cooling performance. Check Airspirit's Heater Core data base to figure out what size you need. However, the pix I've seen of the BI rads look like they use an automotive style design, so they wouldn't be to bad. IMHO the critical things when setting up a rad is that you should NOT just strap the fans to the rad face, instead use a shroud to space the fans AT LEAST ONE INCH off the face of the rad (2" is better) to both improve cooling and reduce noise. Per BillA, fans work better when they 'suck' through the rad instead of trying to 'blow' through it. Last, you should have at least 1-2" of clear space on the other side of the fan as well.

Gooserider

I'm not stupid, thats with SH...

Hmm Although it would reduce some of the sound it does take too much room, sucking through and pushing through should perform the same if you put the fans on right...

I also think of making my own WC setup and i really like the design of Jaydee's block.
I've got a CNC @ my disposal so i can make a nice block and im planning 2 make a few of those #rotor blocks. Only my question is how thick/thin should the bottom be from the water ? My copper plate is 10mm/1cm and how much should there be left @ the bottom?

I got the PolarFlo pro set, it amazes me how good it works, you just have to be sure you use specialized water that does not have any minerals or salts in it, to prevent corrosion since i have 3 types of metals throughout my system.

Anyway that was a dry test off my system, ps im using a eheim 1250 pump, anyway I still need a radiator. I'll probably get a dual radiator and hang it on the top of my case with about 1 to 3 CM between the 120mm fans and the radiator.

Now for the tubing everyone is saying get TYGON, or get this and that, but they have reinforced tubing at HomeDepot it has white nylon threading through the rubber/silicone tubing, i could not get a knick in it with my bare hands, the only thing is getting it on the water blocks... Anyway would this work?

Alot of people get burned on S&H costs though, so I thought it was important to mention. Glad to hear that you're keeping them in mind.
[/quote] Hmm Although it would reduce some of the sound it does take too much room, sucking through and pushing through should perform the same if you put the fans on right...[/quote]
It isn't an intuitive thing, but air will flow faster through an obstruction (like a rad) in order to fill a vacuum than it will to get out of a high pressure zone. In addition, airflow is driven by pressure differential, an axial fan will create a greater pressure differential trying to pull a vacuum (sucking) than trying to generate pressure (blowing). Thus the same fan will move more CFM of air volume when it 'sucks' than when it 'blows'. This is both per BillA's testing, and also some of the fan mfg website's engineering advice. I don't make up the laws of physics, I just tell you what they say!

Ditto the guide on spacing, 1" (2.5cm) is about the minimum that it takes for air flowing through a rad to recover from the turbulence of going through the rad, it also allows air that flows through any area not directly under the fan blades (such as the hub and any area bigger than the fan size) to move to the blades cleanly. If you have the fan closer, then the fan itself will act to impede the airflow.

Your cooling performance is all about how many CFM you move through the rad, so anything that increases CFM is good, anything that reduces it is bad. Note that the higher the CFM you get at 12V, the greater the potential noise reduction you might be able to get by lowering fan voltage until the CFM produces just enough cooling to be acceptable.

----------------------------------------

Opinions vary, depending on block design, but typically about 1/8" (3-4mm) is needed to keep the baseplate from distorting under mounting pressure. If you have a design that offers structural support to the center of the block, or does a honeycomb like structure, then you can get away with slightly thinner. The other issue is that you need to have enough material for heat dissipation, a high turbulence block can be thinner, but a low turbulence block (i.e. a simple maze design) needs to be slightly thicker.

The best approach if one is borrowing a design from someone else is to ask what they reccomend for that design. If designing your own, read a bunch of the design threads to get a feel for how to determine the best thickness for yourself.

Gooserider

Tygon is reccomended for a reason... That reinforced stuff is very stiff, and unless you position it exactly right, you will put stress on your block mounts and risk tilting a block or cracking the die. The non-reinforced vinyl tube usually found at the hardware store is not as rugged or durable as Tygon and is more likely to kink. You should use Tygon (in the R-3603 formulation) for at least the tubes that connect to any blocks. You can use the reinforced stuff on those connections that don't connect to blocks (i.e. res -> pump, pump -> rad) if you like since they aren't as fussy about side loads.

Gooserider