Great progress man. One question for the last two weeks I've been running my computer with chilled water 11-16°C, and after seeing how great it worked under those temps. even memory went higher because of water cooled nb. and no problems with condensation, I think I've changed my mind a little, and now believe that water chiller that would keep my water 5-10°C would be realy great and more than enough for me, condensation problems would be also very easy to control at those temps. Now I want to know what do you think about heat exchanger chiller, basicaly what I mean is that water will get chilled only by running through chiller with inline pump and no res. so amount of water in this sistem wouldn't be more then 1L . Is this doable, I want to have everything as compact as possible so i don't want res. If I would have thermostat in the water to keep my water at 5°C do you think that will cause compressor to turn on/off to often? How would you keep water to ~ constant temp so that I could have long lasting compressor ?
I believe 24/7 running the compressor is better then on/off cycling right.

well i am using a chiller in my project i am working on now to chill the video and chipset.so i have though about that to.well my compressor is going to have to run 24/7 because of the direct evap block running so my water will be extra cold.so how do i slow down the condensation to say the chipset that doesnt need -20c temps?well i am going to use a small vavle inline going to the chipset.and keep closing it down untill the water will barely enter the chipset block,or i should say just enough to keep it say at room temps.now this to me is the easest way to control the temps and get no condensation.since i am using a small water pump or large for that matter wont make any differance,the impellers inside this mag drive pumps will just not pump there rated flow.sorta like holdong you finger over the discharge line with the pump running.doesnt hurt the pump at all.

Are you talking about essentially having a dual-loop setup? One going thru the CPU block, etc. and then thru a chilled-water heat exchanger rather than a rad/HC? I was kind of thinking about doing the same thing with a bong...

Bob

Single loop cpu-nb-gpu ,then instead of cooling water with radiator and fans I would made custom heat exchanger/evaporator all in one block and in-line pump. But the goal is to keep water at aprox 5-10°C.

well sorry i havent posted much latley.but i am making some progress on my chiller tank.so this is where i am so far.
p.s i know someone is going to comment on the gate vavles i have installed infront on the chiller.well they are for return coolant from the chipset and video.,......but using vavles in the return i should be able to regulate the temp of the chipset and gpu .buy slowing the water down i think i can stop the condensation problem i would have on the chipset.and on the video.i should be able to fully control the temps on those 2 items.

http://members.fortunecity.com/ajr01...treeme1/re.jpg

well sorry been a while.but i have been testing my chiller for leaks and such.tank finished and i have filled it and been running for a couple of days to be sure i dont have a problem with any of my parts inside the chiller.used a clear lexan top you might can see in the pic.all i am waiting on is time to start machining a new evaperator block.then i fill this puppy up with refridgerant.....also I purchaced a second simpson temp controler for the chiller temps.proberly start the pc by measuring the chiller temp.set the temp say at -20c on the chiller tank output.so the evaperator by them will be at least at -50c or so.well testing in a couple of weeks and then i will settle on a startup temp.
http://members.fortunecity.com/ajr01...221resized.jpg

WELL BEEN A FEW DAYS THOUGHT I WOULD GIVE EVERY ONE A UPDATE.WAITING ON MY SLOW BROTHER WHO HAS ACCESS TO A AUTO CAD CNC MACHINE.HE SAID HE COULD DO A BETTER JOB WITH THE COMPUTER MACHINING MY BLOCK.I TOLD HIM, OK PUT UP OR SHUT UP.
SO HE IS GOING TO MACHINE THE BLOCK IF I CAN GET HIS LAZY ASS TO DO ANYTHING LOL.
BUT I DID RUN ACROSS A BOTTLE OF R502 CHEAP TODAY.TOLD THE MAN TO BRING IT ON.SO I WILL HAVE A LITTLE SOMETHING EXTRA NOW IN THE COOLING UNIT.HE HE HE REAL COLD.....ONLY PROBLEM THE COMPRESSOR MIGHT NOT BE ABLE TO HANDLE PURE R502 SO I MIGHT HAVE TO BLEND IT WITH SOME R22.I WILL JUST HAVE TO WAIT AND SEE.

Hot Damn! You're the coldest guy around!

I'm impressed. VERY impressed.:eek:

Man, you know your stuff!

thanks guys, i just like to keep myself busy


well i finished machining block this weekend.i still have some final tube connections and stuff but here it is ..new maze design.this design should improve my cooling. I designed it so the refridgerant will stay on top of the cpu longer i hope.and everyone wanted my cnc designs so i am posting the cad print outs of my newest block.I still changed it from cad drawing for the back section.but i think this will give every one a idea of how it works and goes together.
http://members.fortunecity.com/ajr01.../Scan10003.JPG
http://members.fortunecity.com/ajr01...reeme1/cnc.jpg
http://members.fortunecity.com/ajr01...finalblock.jpg

Please explain something for me.

Why is it that every phase-change cooling system that I have seen only has one hose going to the point (Cpu, GPU) where it is going to cool it down. Why isnt there a outlet, where the heated coolant return(the gas) ?

I and really dont understand your block...I have surely missed something :(

that is because most have the capillary line (inlet or pressure) running inside of the return hose that you see.as mine does.so in effect you have a hose inside of a hose.this alows the return gas (which is still very cold)to help cool the capillary tube down to help in the cooling effecientcy of the system.this is call subcooling.
and my block is very complicated in how it works.but very simple in design.the block has 6 parts.
1 ....the base to which trasfers the heat from the cpu.that plate has a chamber machined into it and a precup brased to the face.the precup has 4 holes drilled in it, i hope to cause the refridgerant that is spraying from the capillary tube to collect and spray in 4 directions.i am hoping this design will help in heat transfer from the cpu quicker.also the chamber will hold refridgerant over the cpu die longer.the block is designed to mount vertical so if you look my maze design it trys to hold the refridgerant longer to help in heat transfer.
2 ....capillary tube that passed though the entire block and finally sprays into the precup chamber.
3......the first section has a maze design with a chamber and the refriderant passes into the back side to another maze where it reaches
4......the divider plate that has a hole to alow the refridgerant to pass into the next section.
5.....last maze section where i have a maze on front and then passing to the back side for one last maze.
6.....finally block cap with #8 ac fitting brassed to it.


if you notice the block has been beed blasted to rough up the suface to give me more suface area to aid in heat transfer.

2 pics one a good picture with a break down of the componants.

second you will see how it looks together and look you will see the capillary tube running inside the fitting for the return hose.
http://members.fortunecity.com/ajr01...uevapblock.jpg
http://members.fortunecity.com/ajr01...l/cpublock.jpg

Thanks for the reply.....NOW I understand, the rather advanced design of your block.

It really looks like a gem now for me :)

But why arent we using hose in hose for regular watercooling then ? I know BB2K is in his "radius" thing, but what about all the big companies like DangerDen etc.

Also..it is very difficult to understand that such a little inline with coolant(Your callipary tube), can cool so much. There must be an enourmous presure built in that tube!

And again...really cool block there :cool:

One more...

What do you use for hose around the callipary tube ?

two intirerly differant cooling principles.i use refriderant that is under high press the it cools by evaperation.nothing like water.tube inside a tube would not help a water cooler any but hurt him.

well i use a refridgeration hose disigned for the ac market.mine is made my parker inc.i am a parker hose dealer where i work so making the hose is easy for me.

Admitedly, although it looks simple, it actually has a number of inconveniences, at least from my perspective.

I'm trying to incorporate a drop-in nozzle, but interchanging them will be a fair amount of dissasembling.

One major disadvantage is that sold comercially, Radius would have to be shipped fully assembled, and would be judged as extremely heavy, and bulky.

I just ran across this thread today and I am BLOWN away. This is some very cool custom work.
I'm embarrassed at my setup now. I've been using a 350gph pond pump in a 5 cubic foot freezer with a mixture of glycol/water through a Z4 waterblock. Condensation is a pain, but it does the job.

I'm amazed at where you have taken this project.

Good luck bro and keep on going.:D

Rattlehead thanks for the compliment.
little update.

and i spent 3 hours sunday morning soldering the block up.but now block is together.monday i let acid set in the block to clean it some after the solder and heat got to it....now i have finish machined it and currently being pressurized for leaks.280 psi for a day or two i hope will do it.so with a little luck this weekend i can get the chiller cranked up and see if it works.
block being tested.http://members.fortunecity.com/ajr01...ze.of.Clip.jpg

well a little weekend update.i got the cooling unit up and running.am running now and testing for leaks.i hope to get my load bench finished so i can start loading and testing the unit under stress.i used some r12 just to give it a fill and test with.my chiller tank seams to be doing good.i checked it and seams to be flowing real well at -27c with no signs of thickening up.evap temp was running well under -40c .not too bad just breaking it in.
http://members.fortunecity.com/ajr011/complete.jpg

Dang, that line isn't just cold, it's cold all the way down the pipe! That's going to be one cold CPU!

What do you have planned for anti condensation?

WELL YOU ARE RIGHT IT IS COLD. I DESIGNED THE CPU EVAPERATOR TO HANDLE 200WATT LOAD AT -23C SO THAT SHOULD HANLDE ANY THING I CAN GIVE IT.BUT I AM GOING TO POST BETTER PICS LATER IN THE WEEK.AND YOU WILL SEE THAT THE FROST LINES GOES STRAIGHT INTO THE CHILLER WITH A SECOND CAPILLARY TUBE SPRAYING ANOTHER SHOT OF REFRIDGERANT INTO THE TANK TO CHILL THE COOLANT.THE RETURN LINES TO THE COMPRESSOR AFTER LEAVING THE HEAT EXCHANGER OVER THE COMPRESSOR IS RUNNING ABOUT 10 TO 15F OVER ROOM TEMP SO I HAVE NO LIQUID MAKING IT BACK AT ALL, EVEN AT NO LOAD LIKE NOW IN TESTING.

Simply amazing work. Great job, man! Wish I had the knowledge, tools, technical skills, ballz, and sheer audacity to attempt something like that :)

Bowman, I plan on attempting a similar project over my christmas holidays. I already have a few questions and I was wondering if I could pick your brain about them. Can I e-mail you sometime?

well a little weekend undate.been running unit on and off all week.well today (sunday) i set out to try and load up the unit with a 156watt pelt on the evap and 80watt pelt on the chiller.well middle of the week i changed the capillary tubbing and made it longer to help bring down the temps while i was using r502.seams i set the unit up for r22 capillary and r502 flowed too easy,more than i expected.making the evaperator flood with excesive refridgerant.this makes the temps rise slightly.best it would get is about -44f on the evap.but because i am running dual capillary tubes one feeding the evaperator and one feeding the chiller..it is very hard to get the best of both the chiller and evaperator block temps.get one good and hurts the other,,so i have comprimised to come up with a nice in the middle setup.i lenghened the evaperaor block capillary tube to 120" od .036id tubbing. and set the chiller to 90" of .031"id tubbing.now this hurt by load temps with a 156watt pelt.evap block cannt keep up with the heat load.rises to about 32f(0.c).but then again no cpu will put out that kind of heat.but the chill did very well with a 80watt pelt on load.keep temps around 24f (-4c)in the coolant after a couple of hours of running.

overall the cooler is working pretty good.got a no load temp of -52f (-46.6) on the evaperator cpu block and a -11F (-23.8c) in the chiller tank.chiller is taking 2 hours to reach -12f(-24c) and has a reservoir of 2 gallons.
here is a little pic
http://members.fortunecity.com/ajr011/52f0f.jpg

Bowman: earlier in the thread you said that you are going to reduce flow to keep temps on the NB and GPU in line, but won't the reduced flow have a tendency to allow your coolant to get awfully viscous, especially with an 80W load temp of -4C?