Another #Rotor style block...

[Puzzdre]

Gonna make something like that, I spent all of my (2 and a half ) dremel discs, and started digging with 2 mm mill bit (yeah, talk about humanized CNC ), and in approx. hour and half did this:

[Puzzdre]

I'm gonna 'eat' a lot of copper in the middle of the block, just wanted to see how it goes. Lot's of WD 40 and up's and down's with the mill bit, and wooohooo!!! It' going nicely!

To be continued!!!

Cheers!

[g.l.amour]

looks more like 'the brain' already. think i'm gonna try and copycat the atlantis block in the #rotor style.

i wonder where it is going, sure looks interesting to see what difference there will be.

[Puzzdre]

Hope there'll be the difference... It should be...:shrug:

Been looking for pic's of Atlantis, not found many...It's surely worth a try! I gotta buy another set of dremel discs, think it'll be enough for center piece. And then (sigh ) again to the machine shop to make me those barb holes tappered...Man, I must find the place with metric threaded barbs!

[#Rotor]

Originally posted by g.l.amour
the total idea of this method is that one needs no mill. so actually there are only the drilled holes with connections made with a dremel.

@puzzdre; i had almost the same feeling when i first fired my block. u invest money, time, sweat, blood so it is normal to expect some sort of small miracle. but, the freaks we all are, gaining 1 °C is already quite an accomplishment. thats because most of the systems we got here are already exploring the boundaries of WC. i think we both got carried away too much at making our blocks high flow. remember that brain design?

what might be a good idea is maybe to start a thread (if rotor doesn't mind) where we open-source minded discuss what can be done through rotor's method. as it seems to have been established by rotor that noone should use his method comercially (and right he is), then everyone using the method can help further refine it.

BillA has already given me some food for though,... anyway, the rotor style blox might not be the prettiest on the inside , but it must be possible to make them as performant as any milled out block. if we try to copy the best known waterblock for what is wanted to achieve; and compile it into a rotor design, then any maze/spiral/atlantis design should be possible.

Nope, I do not mind at all...

as for being pretty, hehe...
I think a lot of extremers will tell you actual performance, is much more rewarding than being pretty.... besides beauty is in the eye of the beholder.

[Puzzdre]

Degustibus est non disputandum, but I find the drilled blocks PRETTY!!! All the reddish glow of freshly drilled or milled block simply brings the satisfaction of the doomed on the surface. Most people (including me) are slaves of the symmetry, and equally distributed network of the holes is just the right thing...

I (think ) I'm finished the digging the insides of the bottom half, time for some sanding...Sorry for the pic being too crappy, but I just couldn't make that scanner do it better...

[g.l.amour]

wow, respect man

that is almost totally as i would have imagined it to be ideally. that is not too far away from the atlantis design. in fact the atlantis is designed for single in - oulet; so i guess your design is an evolution for single in dual out design. way to go.


btw, i picked myself up an epox 8k5a;
factory
cpu vcore upto 2.2V !!!
mem vcore upto 3.2V
.
.
.

(necessary update after soaking my kt3ultra with a mixture of h2o + antifreeze + radiator protector)

g.l. BR34K3R of things

always the same with me, in the beginning when testing something new, i'm always very cautious, and after a while i thk i am the pro, and is start makign n00b errors (starting an amd without a heatsink attached etc.)

[Puzzdre]

Thanx m8!

I'll have to search some more for better look into the Atlantis. It's the one with the 'starfish' thingie in the center below the inlet I suppose (found only pic's with the top on, not one of the disassembled block)...

I'm curious what performance with this block will be...Tomorrow I'll have the top finished, so I guess, temps soon...This making of the block goes much faster than the first one, guess it's due to all the tools at hand...Only that #Rotor's warning about the tapping the LAST hole...never leaved my mind...

Congrat's on the mobo! Think it's a good choice!

Sorry for the 'late' MSI, those dirty bastards don't like the bath... but this says it all (in my case too )

always the same with me, in the beginning when testing something new, i'm always very cautious, and after a while i thk i am the pro, and is start makign n00b errors (starting an amd without a heatsink attached etc.)

LOL, g.l. BR34K3R of things but 133t f1x0r at the end!

Cheers!

[g.l.amour]

i'm very curious about the temps

because i can tell u, vcore of 2.2 will bring the best watercooler at its knees

with fans at 12V
1800+ at +/- 1760MHz
temps seem to be at +/- 46°C!!!

(whereas i used to cool passively at 100%load at 43°C - 1.85V and 1700MHz)

what i also have seemed to notice in the past, and is happening again (could be wrong)

when oc'ing to the max, when i go over 47°C its finished, hangs completely. so i guess that is where the prometia's start to kick in...

[Puzzdre]

Here's the little update...
Got the top today, needs some more sanding. Today I figured that bottom openings of the barbs are small, so tomorrow I'll have to remove some or all of the 'ribs' under the outlets. I wanted to keep them for some benefit of structural integrity, but they'll have to go. The barbs are 10 mm ID inlet, 12 mm ID outlets... I hope tomorrow night I'll have some temps to post...Hopefully good ones...

Sorry for a 'little' big pic

[g.l.amour]

looking solid. am getting really curious now

[Puzzdre]

Here's the pic of the assembled and sealed block... Waiting for the silicone seal to dry, then test in the kitchen sink, and hopefully, in the wc setup...DOH!

[g.l.amour]

waw, thats a little faster than the former block we made,

[Puzzdre]

Yeah, things are a lot easier when you know what to do and how to do it...This one was started with your principle, Cut to size, drill and TAP the holes, if everything goes fine, do rest of the job!

About knowing what to do, this one leaked...Now I re-sealed it, still waiting to dry...

Thx for the , I appreciate it!

[Puzzdre]

After two re-sealings, and five (5 ) in's and out's with the barbs, spent almost all the teflon tape I had, here are the results (for now, 'puter working around three hours with the new block)...

Nothin' spectacular, at the room temp a little high, 25 C on the rad intake, idle 32 C and Cpu burn load for one hour, 38 C.

Guess the setup could use a little bit better pump, copper rad with decent intake area (this mounting of the rad under the midi tower case is not allowing it to breathe fully I think, crappy shroud 1 cm distance between fans and rad) and some decent si tubing (this pvc is kinkinnnnnnnn')... Guess I'm satisfied with the results, gonna try refine the block little more, but that'll happen when I mod the full tower case and setup the watercooling as it should be...

These are temps for amd XP 1800+ @1667 MHz, stock voltage...

[g.l.amour]

that is weird, i have never had issues with sealing barbs... some teflon and gogogo

it is prolly a bit to early to judge completely, (as billA's gospel goes, mount the block 10 x ) but, for now, how do u thk it performs against the previous one?

[Puzzdre]

The problem with the barbs is in 5 mm thick (or better, thin) top, and that 3/8 thread. The thread should be more dense for that thickness of the top...But lot's of teflon solved it (I was ready to epoxy bloody things when it worked at last...you must scare them to behave properly )...

Dunno what to say, I'm 'satisfied' for now, but not completely happy...:shrug: ...actually, I got that feeling that I should rebuild the complete system, than judge, cos' being here at the forums for several months now, I learned much to see all the flaws in my current setup. I guess that this block could do better with the right tubing size, not-kinked-outlet tubing with the proper Y, and a better pump. I didn't have time to try to measure (the fill-the-bucket method ) the flow, but testing the block for leaks in kitchen sink, and comparing it with the one we made in first attempt, this one is much more restrictive. Although, it works better, but not in the temps range I expected it to be, more like 30 C idle...

It's better than first one, and I'll keep it for now. Guess the first one we made would be good with pelt, but cannot try it (no pelt )...

It was worth the try, maybe mk3 will be little better (gonna try that deeper in the middle, shallower towards the outlets thing)...

This one is like this: center of the block is 3 mm base, first circle is 4.5 mm, second 3.5 mm, third 2.5 mm and last (outer circle) is 2 mm base. Toward the outlets I raised the base thickness for 1.5 mm, from 2 mm to 3.5 mm.

Maybe this circular pattern is not the right way to go, or I just don't know how to 'open' it good for flow...

[g.l.amour]

well, better is better. there will be a point where us mere mortals won't be noticing the improvements anymore coz we don't have the right equipment for testing.

block mounting alone gets me 2°C delta in performance. try judging a block in one hour time is almost arrogance; lol

try remounting it a couple of times, thk u will get better temps

[Puzzdre]

LOL, yeah, I think some more remounting of the block will be needed...

Funny thing, I mounted it actually three times, first and second time I got 34 C idle (for about half an hour each), then decided to remount the block with shim I removed before. I guess it helped the block to sit on the core more evenly, cos' now I got that 32 C idle...

BTW, I was in a hurry wanting to see how the block will perform, and forgot to plug my pump in...Luckily, the shutdown option in the bios was fast enough, I actually started to tighten/untighten the bolts holding the block to mobo, and I saw sudden rise of the temp...My cpu is re-born today, bios shut it off on time...

[g.l.amour]

woops close call. on the other hand, would have been a good excuse upgrading some more HW.

so +/- the damn thing is performing the same as your last block? maybe u went over the sweet spot flowresistance/velocity.

what was your baseplate thickness this time?

[Puzzdre]

would have been a good excuse upgrading some more HW.

OOOHHHH, she would freak out....

Hmmm, it's little better of the first one, with the first block idle was 36 C on same room temp...A little improvement...

base is like this
center of the block is 3 mm base, first circle is 4.5 mm, second 3.5 mm, third 2.5 mm and last (outer circle) is 2 mm base. Toward the outlets I raised the base thickness for 1.5 mm, from 2 mm to 3.5 mm.

maybe to go thinner some more?:shrug:

[Puzzdre]

Maybe that channels made with thin dremel discs are too narrow...Considering the wavy shape of the 'walls', maybe I should open the channels more with say 2 mm mill bit, there'll be still much turbulence in the channels, and the block will have less restriction... Think I'll try this one of these days when I take out this block. I don't wanna make another one with questionable results, better massacre some more this one... Snow is still waiting...giving me some more time...

What's happening on your side? That plexy top block?

[g.l.amour]

scary, u dared getting as low as 1.5mm baseplate thickness? waw u got ballz.

well, all in all; i think hooking up another rad is going to give more temp decrease for less hastle. might get my hayden rad out of the attic and find an empty spot in my case to hook it up.

anywayz, after messing around for 3 days with my epox mobo; i seem to have found out my corsair pc2400 from last year needed to be retired (hooked up some pc2700/2.5 samsung). so i'm running reliable at 1.9V/157fsb/1820MHz. this little increase of vcore and mhz has increased my temps by about 5°C!!! looking at it from the bright side; my room temp seems to increase from 17°C to 20°C in no time with just my puter and my ol' 21" heating it. i'm wondering if there is alot more to gain by changing WB design. i'm going to try copycatting the atlantis block though in the far future, but reading your experiences makes me think that there won't be no big surprises...

[Puzzdre]

scary, u dared getting as low as 1.5mm baseplate thickness? waw u got ballz.

Guess it's more like brain shortage...

there won't be no big surprises...

I hate to say it, but you could be very very right...Until proven otherwise...

[Puzzdre]

How many MHz more did you gain? The +5 C temp increase sounds horrific!?