Aaaarrghhhhh...:eek:

Just like BillA said, Zeros are the most expensive numbers...;)

Hey Puzzdre, you can use any size barbs on the sensor,use a reducer, does'nt matter if the sensor is restrictive cause the other channel is open for flow, as long as you get enough going through the sensor to show movement alls OK :) ...

Yup, that will do, now I have to find agood Y piece with somewhat 10 - 12 mm ID. That won't be so easy, cos I cannot buy one here, dunno if automotive parts stores have that kind of stuff. Maybe I'll make some somehow.:)

And seeing Volenti's solution for one way valve, I'm going to make one too...copycat...:D

Just a little update...:D
I'm gonna try both setups, serial and paralell pumps then decide which one to go...

Chipset block on the way...pics (more like scans cos I don't have digicam...:p )

first, grid on the chunk of copper
(somebody pls. enlighten me, how to post multiple pics in one post:confused: )

punched start holes into grid, removed the sticker, drilled and...

and ofcourse, forgot to attach the pic...stupid...:mad:

both halves, looking better in real than put in scanner and scanned...:shrug:

can't wait to see the finished block. so far it seems relatively doable with a drill on a standard; so long as , like u did, let someone else dimension the copper piece as needed. chipset and gpu blox don't seem to need ultra performance, just a good amount of flow.

btw, multiple pics can be done if u host them pix on some personal webspace and use the 'img' tag to adress the http'z of the pix.

Thx m8!!!

I thought that for multipic's I'll have to put them somewhere else!

As for the block, just picked my copper from machine shop, all perfectly sized and flattened. That will be the base for cpu block.

This one on the pic's I cutted myself with handsaw (had some 40 mm wide copper bar), sanded a little, than drilled today.

I think that tommorow I'll have enough time to drill one cpu block, and post the result.

Cheers!

oh yes. looking good... :) I see you had a little bit of a problem with some of te holes, drifting..... did you punch each hole, before you drilled, or did you just drill right through the printout...

Hi #Rotor,

Yes, I punched all the starting points, but drifting is caused partially due to my cheap drill press, paritally due to my 'puncher', I couldn't find any better (at home, without going to the store), so I used one bullet (12.7 mm slug) I had from past times, thought it would serve a good purpose at last ;) , so the centering was mostly good, except in few holes. The worst holes are my first drilled, I put too much pressure on drill so the drill bit drifted towards neighbour hole. The others I drilled with less pressure on the end, turned out good.

This will be the chipset block, so I made it first, to make less mistakes on cpu block.

I thought that I'll be able to drill cpu block today, but it's snowing and pretty chill on my balcony, so dunno would it be done today or tomorrow.

Thx again for all the advices!:)

so basically, all you are using is a cheap drill press?

Yup, with some good quality drill bits...simple, fun, nice results, fun, and ofcourse, fun. :p :p

allow me to say, this manufacturing method is by far, the cheapest, as far as high tech equipment goes. it is also the most versatile as far as design options. It is the soul reason Why I can literally, in a matter of 10 minutes, communicating with someone on the other side of the world, design and finalize a 100% customized block,


to put it in another way.... If I was AMD, I would have been able to build you your CPU, to your spec, via Email and I would have been able to do so for less than the norm as far as mass produced product available.

and it's all thanks so this manufacturing method. It took me 4 years to get it fine tuned, And I'm still improving every time I make one.... the beauty of this.... doing it with CNC would greatly enhance accuracy and speed.... but it is absolutely not a requirement. a Steady hand and eye, and you will do just awesome... as them pictures prove, That for a very first effort... is way much better, than my first effort... My first one, ended up getting the hole cavity cleaned out... :) none of the holes was where they should have been :D

Wow!:eek:
Four years! Another thanks for sharing the info on manuf. the block! When somebody puts that much time in something, and literary gives it to the people, what can I say, BIG THANX!

Thx for encouraging words.:)

Tomorrow I'm going to buy a nice plexy shield for my face, and remove that crappy blurry piece of plastic shielding drill chuck, one cannot see anything through that.

One question though, I have couple of 3 mm mill bits here, would it be any improvement if I try to drill the holes with those flat end mill bits? It would be much trickier to centre the bit over the hole without xy table, but is it worth to try?

actually, the cone shaped ends of each hole, adds to the total surface area, as well as increasing the turbulence. endmills, are very "grabby" if you do not have proper retention on the workpiece.

Have you made a block where the holes are not directly one on top of the other before Rotor?, imagine stacking bottles on top of each other, on their sides. they each fit into the space between two others...

Indeed I have, you mean staggered, so that the centers of the holes in the top plate, are located over the pins in the bottom plate..... or do you mean like this....

http://3rotor.homelinux.com/images/c...P04480_sml.JPG

problem with doing like this, the pins are so thin they can barely carry the heat all the way to there tips....
http://3rotor.homelinux.com/images/c...P04482_sml.JPG

Im just curious, did you ever tried the circular hole pattern like I posted here? How did it work?

I'm still eager to try it, if anything, it would still look gorgeous on my comp desk here at home (I'm still impressed with the inside looks of that pieces I drilled two days ago...:cool: )

nice thread ya got going here, i thought i'd throw a little something into the mix. while you are talking about dual pumps, i came across some interesting heater cores:
http://members.sigecom.net/unloaded/2-271.jpg
http://members.sigecom.net/unloaded/2-726.jpg
I have been curious how these would do, specially the one with dual outlets. I picture res->5/8ID->pump->5/8ID->radIN->rad1OUT 1/2ID->cpu->1/2ID->res & rad2out 3/8ID->gpu->nb->3/8ID->res
this would provide higher flow thru cpu where its needed, and less thru the chips. also only the nb would be getting second hand water. i'm not sure where you'd put a second pump in this setup. why are you wanting 2 pumps anyway? are those the only ones available there and you don't feel one would be enough?


peace.
unloaded

now that core would be ideal for running SMP with.... have the radiator act as the Y split in the loop... then have each split form it's own loop all the way back into the reservoir. IE, have two return fitting in the reservoir as well....


I have not tried the circular pattern yet.... I think you are going to find that cutting away so much copper in between each hole, is going to cost you a little fortune in Dremel Disks. that is why, having the holes in a 90º SQR pattern, gives the best result, the formed pin is concave square with very sharp corners, maximizing the surface area related to its volume... and yet it still has enough "body" to be able to carry substantial amounts of heat out into the pin...

I meant the bottom one, but the other sounds interesting as well!, except only the bottom plate would have heat going up the pins, the top pins would be pointless would'nt they?, if they have no pin to mate to and carry on the conduction from?. or do they act as blanks to keep the water around the bottom pins?...

As I understand the design, given that the top is identical to the bottom, the "pins" would probably help in the structural integrity area, if not the heat spreading realm. #rotor, please correct me if I am wrong.

Bob

Thx for contribution Unloaded!:) Think I'm gonna run separate loops on one rad, just drill and solder the fittings on the rad. Yeah, the pump I own is pretty much only one EASILY available in line pump here, and the price is decent. I can buy two for price of one Eheim 1046 (also Eheim is not available locally, I'll have to wait for few weeks for delivery).

#Rotor: I never thought about wearing of Dremel disks, you're right, this might become an issue...but I already punched the start holes in one half of the block, so I guess I'll have to try to make it...

This is what I've done today:D

Puzzdre..... DAMN.... you are getting good at this! See, I told ya so.... not too hard at all. :)


Structural integrity as well as heat transfer, you guys are both correct yes. the two plates get bolted together really hard, with 4 SS bolts so as far as the heat is concerned, it's pretty much one continues pin, all the way to the top. which, if one side is drilled 4mm, would mean the pin to be 8mm long..... now that is a mighty long way for heat to travel, when you consider the environment that pin finds itself in... just think of all the miniature little vortices and turbulence, and it changes with the smallest rise or fall of fluid velocity.

Haiku...:p

/puzzdre is coming to the stage, taking the mic/

Thank you, thank you, thank you!!!

I'd like to use this opportunity to thank all the good people who supported this project from the beginning, especially my executive producer of all the funds - monetary and neurally dependant - my wife, and also, The Man Pointing Me In Right Directions - #Rotor. Thank to also all the good people sharing their knowledge on this forum.

/puzzdre is leaving the stage distinctively staring into the hosteses' wonderbra, trips on the second step and falls into the deep of copycat shame. The crowd goes wild./

;) :) :cool:

oh no, its getting late again. puzzdre is going berserk again. keep em coming dude !!!

:p LOL:p

I've been busy little bee today:cool:
Off to buy some sanding paper.
Here's some pieces straight from drilling...Yo ho ho, and a bottle of rum!!!:D :D :D

that is beautifull puzzdre!!!!


i found some bench drill for +/- 50€ , looks solid enough. but the table is fixed. so for every other hole, you'd have to unclamp the copper. and position it for the next hole and then reclamp. is that how u do it puzzdre, or do u have a mounting table that can be moved in X an Y axis? looks like a bitch to drill some 50 holes that way without a flexible table.

i am on the brink of going down the same road.

This is the drill I have, all I ever drilled for wc I drilled on this. The drill was 54 Euro, and the vice was 13.5 Euro. The vice is bolted for the drill table with two bolts. What I do is, clamp the copper tight into the vice, than unscrew a little bolts holding the vice to the drill table, so I can move entire vice forwards/backwards with some force. When drilling I just hold the vice steady while positioning the drill bit over the starting hole, than drill and position the bit over the next starting hole. This way I manage to drill one entire column of holes in the square pattern, then turn off the drill, unclamp the copper and position it for the next column. Similar to that I did when drilling the circular pattern.

It would be surely much better having the xy table and everything clamped tight (it would be more precise and less dangerous:D ) but this is the way I do it without (for now, hehe) xy table.

I hope I explained it good enough, if you didn't understand I'll try to explain it better.

Glad to hear that!!!

:) I'll keep posting to give you some final 'push';)


pic of the drill: