so whats best then?

if you dont get these blocks made in a month or 2 email me
i should have my mill by then...

and why are the blocks soo large? the ones i make are 50mmx75mm aprox, or the ones i have made in the past...

Wouldn't we all like to know!

I've got room for 100mm long blocks, so why not make them 100mm long?

I'm hoping I can convince someone who has posted in this thread already to make them soon for me

I know how you feel. I would still like a radeon cooler from morphling but he's working on new tools so it will be a while.

Brad..... YGM :)

http://3rotor.dns2go.com/images/r&d/...00-cucu-01.jpg

http://3rotor.dns2go.com/images/r&d/...0x100-cucu.jpg


something like this? maybe...

Webmedic, I'm looking, I'm looking, patience my friends :)
Brad why on earth would you need such large block if you won't use two TECs, that's complete waste of space, it won't cool better than some good designed 60*60 block, and also this design doesn't look to efficient to me, you've got nothing in the center where all the action happen, you're going to have some death spots behind those T fins, plus block being so huge the water will move very,very slow inside, because there is just one water path with very large cross section and even that gets split two ways so velocity is again halved going left and right, no,no, this would be very bad the more I look at it, sorry. Design block the way you keep velocity at max and keep surface area large.
It all abot proportions if you got inlet cross section 120mm2, you want to have (if you're thinking two outlets) inside paths 60mm2 going in two directions, maybe a little larger if your pump can't handle a lot of preasure drops, or even smaller to improve velocity->heat transfer if you have pump that can do that.

Rotor - my mail checks every minute all day long, I probably had read the email before you typed that :D I was just too busy at work to reply all day. But I'm at home now, finally, and you've gone offline. Maybe your TigerMP is on fire right now? :P Just emailing me the two pics would be fine whenever you're online next.


Morphling, the gap between the fins would be as thin as possible, also the open part of the block would be fairly shallow, only 5mm or so. If the block was 3/4" deep then sure the performance would suck. Also I intend the base thickness to be 10mm, to spread the heat out as wide as possible.

Ok, that I little better, but just how thin are you thinking, because 2mm will be is as think as anyone would be willing to mill in copper. Then I just read you want 3slug mounting on socket !!! Do you realize that just bottom alone (100*60*10) weights ~0.5kg, the that dense fins and 5mm copper top and you're close to 1kg, no socket is going to hold that secure, rethink.

ok, lets go back to the holes again then ;)

this block would use a rotor style grid. The line near the top is the start of the rise in the socket 462 design. The outline is the area of the grid.

any comments?

hehe no it's not me that's offline, it's dns2go not able to do what they say they are best at.... my server has been non stop for some time now.... and it's frustrating me to have to deal with this kinda service, but ya, it's free, so I can't really complain now, can I..... :)


I am online now, or shall we say, DNS2GO thinks I am...

my pin-grid-array can go right up to 0.5mm in base thickness, easily, and since the pins go all the way up, the block will still be structurally sound, I.E.. you wont be ably do dent the middle of the block by pressing hard on it... of coarse the average thickness is always a tad thinker, due to the cone shapes of the hole bottoms....

would you want it so you can run the block with Tec's later on... or is TEC absolutely not in the picture... reason I'm asking, the VDFC pattern looks very different for TEC, than for just straight Core cooling...

no pelts at all - I'd need to run 220w's on them, which isn't practical.

which do you think would be better, your design, or my version?

Rotors design looks more practicle. If you really want something big ask him about his quatro with dual ins and out. It's a monster block.

Better yet here are some pics:

http://3rotor.dns2go.com/images/Turb...ro/indexx.html

those first two pictures was just to get the imagination on a role.... as you can see it's plain and no holes anywhere, a blank in other words..... :)

here is what it has evolved.... mutated is a better word, I think :D... into.

http://3rotor.dns2go.com/images/r&d/...00-cucu-03.jpg

and in 3D it looks like this.....

http://3rotor.dns2go.com/images/r&d/...00-cucu-02.jpg

and
http://3rotor.dns2go.com/images/r&d/...u-02_under.jpg

I've seen the quattro, that's what inspired this block.

Rotor, looking really nice, can you email me your account details so I can DD the money?

hmmmmmm..... haaaaaaaaaa.....

the smell of freshly milled copper..... :)

eerrmmm or is that coffee I smell.... :rolleyes:


picture 1

picture 2


I dono about that socket arm issue, I looked at my boards, and they all clear the top of the CPU by a substantial margin...

do you have some pictures of your board maybe?

very very nice.

there are pics of my board all over this place: http://homepages.ihug.co.nz/~otama/MPX2/

http://homepages.ihug.co.nz/~otama/MPX2/MVC-819F.JPG


looks like yours are fine too... seems like it even clears the socket without any CPU installed, or am I miss-understanding you as to the clearance you are talking about...

I'm talking about the area's in red

oh ok... yes, you mean the step in the baseplate..... yip, all nicely milled away....

Rotor, got a q about yoru blocks

what do you have to make them? it looks like you cnc a bunch of holes that interconnect a little bit,

how thick are the plates you use?? it looks like its only 1/4 for tthe bottom and the top... and what kind of flow you get with those? it looks really restrictive...

Yes I "HYC" the holes with a somewhat cranky and not much used super-computer. :D


the plates are indeed relatively thin, and there is a good reason for that. in order to keep the flow velosity right up there, even with the most puny of pumps emaginable, the block volume needs to be held at bay, though the ratio between the effective surface-area and the volume, is way up there.

To design the block to make the pumps life easy... that is not what the block is there for, sucking heat out of the CPU is what the block needs to wory about. Let the pump worry about getting the liquid through the block.

"If it can't do that, don't go open up the block.... get a stronger pump...." is what I always said


flow rate.... well this is with 1000mm of lift.... 3 feet

which equates to about 169 GPH effective flow.

what i dont get is how you can keep such high flow when it has such a small area to escape from the fitting! do you make the design on both sides of the peies? whats the performance like compared to regular blocks? like the swifty?

I have the pins on both sides yes, the fact that there are more than just one path for the liquid to take, reducing the flow-resistance of the block considerably, is what accounts for the apparent "not so bad flow" .... :) also, the fact that the liquid spend a very short amount of time in the block, makes for awesome replacement rates, I.E.. the liquid does not have to follow a long and winding road to be able to get out again, it takes the first available path of least resistance and then it's on its way back to the radiator, making space for new cool liquid to enter....

I'm kidding.... I'm using magic cheat codes to get the flow up... :D:D

lol your crazy;) but the thing i dont understand is how the flow could be kept up... when it has a tiny spot below the barb to escape from.....

it's not that small, the chamber is abour 5 to 6mm high, right where the water comes in, but it' because there is more than one way for the liquid to go, that makes the difference. In a single channel design, you have one ticket-office, and 10 000 fans.... no way everybody is gona get in for the consert... :D

and with a 660gph pump per block I'm very concerned with flow rate ;)

oh Brad......and while you are here.... check it out


http://3rotor.dns2go.com/images/r&d/...05582.JPG.html
http://3rotor.dns2go.com/images/r&d/...05583.JPG.html


I'm about half way, the most tedious part is done... :)

looking good. you've done almost all the damned work before the money arrives ;) still, I suppose you'd sell them fairly quickly to other people.

I guess it wouldn't be that hard no, though these blocks are a tad bit on the big side.... won't fit on most boards... doesn't even remotely come close to fitting on any of mine....... I had a bit of a dilemma, as it was almost surely a once off, I didn't want to spend the materials on making a template, nun the less it was 4 plates, that had to be accurately re-produced, so templating it was kind of a requirement, I ended up using a piece of plexiglass, worked like a charm, though the softness of the plexi, will render the template rather inaccurate after about ten uses or so...