The compressor didnt broke, the condenser did
Printable View
The compressor didnt broke, the condenser did
Ok ive visualized my evap with some paper, a pencil and a red marker ;p
Some overview:
I wanted to make an evap wich had as much copper --> freon contact as possible, right above the cpu so a classic maze design wasnt sufficent, it could give me freon right above the cpu die but the surface where the freon was over the die wasnt sufficient in my opinnion... It would be good if the cpu had a 2x2cm die but its only a dream :\
A hollow block also wasnt sufficent, the freon was in the possible best contact with the copper right above the die, the amounts of freon were sufficient and it didnt flow anywhere pushed by the vapor being produced when it was boiling. The second big problem was te fact it could work in only one position... The last disadvantage in my opinion was the fact that it doesent work too good if the freon is injected as a mist (when using a cap tube its injected as a one stream- a no go wthen using injectors as they produce mist)
Ok, so heres my idea:
Its inspirated with waterblocks, mainly Morphlings MK4 waterblock, its first copper revision.
Using a micro fin design right above the cpu die we maximize the surface area next to the die, the surface area @ copper --> freon is as large as it can get near the die.
Thanks to the much bigger surface it can work with a mist. It can work in 2 positions (when you'll se the primitive design youll se), vertical and horizontal when using a cap tube, combined with a injector instead it can work in any position beacause the mist covers the fins firmly...
The freon could flow through the fins to easy but if the spacing between them is bigh enough and the load per mm^2 is lower it doesent boil so violent and stays between them + the bottom is shaped so it acumulates near the pins...
Ok, enough of this BS, some pics ;]
Hope i didnt make any goofs beacause for the last 72h i had about 9h sleep ;p
What do you guys think about it ?
Ok, its time for some real sleep ;] See ya in ~10- 12h :D:D:D
damn.... :\
berkut
some good ideas my friend.the evap isnt bad idea.you are on the right track.you need mass and enough area to boil off the liquid.
you are doing well.keep it up.:D
pc ice - we make isolation wich is hermetic and doesnt conduct heat, its a chamber for the cpu and evap only wheres no air = no humidity
look here: http://phase-change.com/index.php?ac...s&id=25&page=1
ok, thought about some stuff- subcooling...
To reach those damned -80C temps. ill need some hell'o'mighty subcooling in the II stage... So heres the plan ;]
Im gonna use a 10mm ID flexible hose (hope bowman reply's) and through it will go a cap tube, a really oversized one, its needed only as a "flexible" hose wich will suplyy the injector with freon. But as bowman said- a cap tube inside a only 6' long hose hasnt got enough subcooling power, its surface is insufficient. Cause im gonna use a rather thick cap tube i wont have the option to coil it around some pipe and run it through the pipe as a spiral.
Heres the big plan ;]
Why not cut a 8mm pipe in litle rings, lets say 2mm thick and solder them on a cap tube every 10mm ? They will increase the surface area at least 2-3x (they will work as fins). They will also work as a inner shell so the flex hose cant kink if we bend it to much by accident. The stiffness of the hose shouldnt change a bit exept for the extra one from the cap tube.
And hers the tricky part: how to connect all of this. The beauty of this system is its simpleness... the diagram:
Berkut...this is a truely awesome "walkthrough" done with great skill and detail. :) You have inspired me to get off my Bu** and build something! Thank you!! ;)
The coiled tube on the right side is for stress relieving, when we bent the whole thing stress occurs (the cap tube is streched or crushed and we dont want it cause it weakens the joints) and we can freely tighten the flares w/o any fear the cap rotating cause the coil absorvs these forces also.
The performance enchancers are these 8mm rings made of copper pipes ;]
Or how about this? You could get all but 6" of cap tube inside the suction line the bit between the drier and entering the copper part of the suction line. This was going to be my master plan for my mk2 build but ill share it now :)
I reckon you'd only need 12" - 15" of 1/2" copper pipe to get enough of the cap tube inside the suction line.
What do you think?
Can you not get any1 to supply parker hose to non-trade either?Quote:
Originally posted by berkut
Im gonna use a 10mm ID flexible hose (hope bowman reply's)
]JR[
ok here is what i have.so far.
fitting i have on order.
and hose i have ordered
That fitting sure looks like those on my hose:
aha the part on the right is a flare to flare connector but i didnt want to disconnect it
wow thats a good idea for those fins on the cap tube!! btw did u ever find the leak yet?
nope.. im sending tomorrow that condenser to the repair shop... finally :O
lol damn how much that going to cost? mostly labor hopefully theres nothing really wrong!! and u better tel them u want to know where the leak was!!
I dont know where it is :\
I think that condenser is galvalicly erroded and it has multiple leaks but my dad wants to try to fix it first, i would buy a new one for 110E... wich has 5 or 6x more surface and a different design :O
with all this talk about subcooling i think we need to remember where all of that heat from the cap tube goes. i am pretty sure that it is transfered to the condenser and compressor casing adding a larger heat load on the condenser.
i cant believe 6 ft of cap tube inside the suction line isnt enough. 6 ft of cap tube is usually 2/3 of the entire length, so adding the ENTIRE cap tube in the cooling mix wouldnt yield more than a 33% gain in efficiency if your cap tube was 9 ft long. the coldest part is right out of the evaporator, so if you wanted maximum subcooling, it would need to be cooled where refrigerant is boiling which is the evaporator outlet. somebody need to try and cool their cap tube with some other means like a pelt or chilled water. the heat exchanger would be simple for this.
Thats true but for a 1C bigger temp. on the condenser i can have lets say 10C more subcooling... (nfi what will be the difference) :0Quote:
with all this talk about subcooling i think we need to remember where all of that heat from the cap tube goes. i am pretty sure that it is transfered to the condenser and compressor casing adding a larger heat load on the condenser.
Wont argue it its enough or not, in my system i cant put all the cap tube near the evaps so i have to increase its surface so the vapour also cools itQuote:
i cant believe 6 ft of cap tube inside the suction line isnt enough. 6 ft of cap tube is usually 2/3 of the entire length, so adding the ENTIRE cap tube in the cooling mix wouldnt yield more than a 33% gain in efficiency if your cap tube was 9 ft long.
Its in the II stage, the freon leaving the condenser has ~-25 to -30C ;] No pelt could cool that ;]Quote:
somebody need to try and cool their cap tube with some other means like a pelt or chilled water. the heat exchanger would be simple for this.
I cant build a 2nd I stage loop cause i dont have any good compressor left, all compressors wich i have left are to weak...
well if you havent sent the condensor in yet I found that you can stick the whole thing in water and pump air into it using the compressor. Ive had the psi at over 700 before and let me to you, if there is a leak you will know where it is. I have been told that it can damage the compressor if you keep doing this but so far it hasnt done anything to mine and its only a 1000btu one. Also I see you are using solder so I wouldnt go as high as 700, try 300 or so first.
In the past i tested my evaps @ ~30 bar (~420psi) (using solder and didnt leak :D) but it isnt sufficient cause there are temp. fluctuations... like the leak i had- it was perfectly fine during pressure tests to crack when the temp. lowered...
Now im redesigning my cascade, ill optimize the freon flow (so minimal amounts will be needed in the I stage) - in my current design i didnt like its path, it was to narrow and there were spots where oil could separate, in the new design there's only one place where it goes up (and beacaue of the high flow oil doesent separate) - the tube between the condenser and the compressor...
The condenser will be different- ill use a water to cool it, im thinking about a 10mm pipe as it with a 28mm pipe for the water (the 10mm pipe will be in the 28mm), it will be 6 or 8m long, coaxial...
Ill add a 2nd heat exchanger for subcooling the freon entering the txv.. And few other little adjustments ;]
Ill post some new info in 3-4 days...
ok, some pics of the stuff i bought, no comments this time cause i have to go with my parents...
baseplate + condenser pipes