Need adcice on building a phase change evaporator

redleader · 06-09-2004, 08:34 PM

This is an extension of my last thread where it became clear that using a swiftech block as an evap was not an option. In that thread it was pretty obvious that i didn't really understand the pressures at work in an evap. So heres what I need to build a block capable of:

*Must be able to withstand at an absolute minimum 200 PSI of inside pressure. Ideally more in case i ever want to switch to a higher pressure refrigerant and for safety in all cases.

*Must be able to return a small amount of liquid oil to the compressor.

*Must have a large enough internal surface area for ~150w of heat to be transfered to the refrigerant during phase change.

My current idea:

Modify an old Maze 2 classic block:

First step would be to unsolder it. Then I'd need to clean it out and braze it together and replace the barbs with fittings for a gas line and a cap tube. Finally I'd bolt on a socket A to 754 conversion bracket. As an added bonus the bracket would allow me to bolt the block together in the four corners, further increasing its strength.

Again though this brings me back to pressure. I remember Dangerden pressure tested the old maze1 blocks which were very similar at 1000 PSI. Since I'll be brazeing and bolting I think i'll hold at least that much pressure, but i have no concept of what high pressure like this really does to a joint. Does this seem reasonable?

Also, about brazing. Is there some special braze I should be getting, or should i just go with the regular stuff used in AC systems? Will ordinary copper braze work when i braze in brass fittings? Finally does braze use flux? I've only worked with plumbers solder (though i have a MAPP torch).

Edit: Noticed i misspelled the topic sentence

crane · 06-10-2004, 01:26 AM

I've been working on a retro-fitted water cooler recently...took the plunge and finally evacuated the little .5 HP compressor and chopped off the evaporator coil to try some experimentation with different configurations. First thing i did was solder on a service port at the process tube on the compressor... i figured I would try my luck at pulling a vacuum and recharging before I went too far with evaporator mods. I forgot my 3CFM 2 stage vac pump at work,so i was limited to using my little hand vacuum pump(don't laugh) to pull a vacuum. Moisture wasn't a concern as it was only open a few minutes.
I pumped like hell and got it down to -25 and shot the gas to it and it started happily and cooled just fine...seems to me the little hermetic compressors are pretty tolerant of under/overcharge as i was able to fine tune the cooling by adjusting the R134a amount. It holds a whopping 1.75 oz.
I then decided that since the evac/charge scenario was simple enough... i would cut out the original evap and try some alternate configurations. My first evap replacement was a coil of 3/8 copper with the cap tube soldered in one end and the suction tube soldered on the other end of the coil. I then did the hand vac deal and recharged again... Once again it fired right up and made a big chunk of ice on the coils. I used some old plumbers solder I had lying around forever.. can't even read the type on the roll its so old. It seemed to hold pressure fine and i wasn't a bit concerned about something popping loose under pressure.
I wasn't too happy with the size of the evap coil I had made, so once again...chop chop and this time I made a coil out of 10 foot of 1/4" copper tube and soldered it in place,using the same crap solder i used before. Only difference was i bought a little container of tinning flux from Ho' Depot and that stuff is the cats ass for soldering copper. I will look at the bottle when i get home this weekend and post the brand and type.
To sum it up.. i wouldn't concern myself with brazing anything on the phase change system...just use a good flux and plumbing solder and have at it.
My results from playing around with my chiller were really quite good... it makes zero degree celsius water and and I had my chip down to 8 degrees celsius. But without any condensation proofing and insulation, I didn't run it but a little while as it was sweating terrible.
Next step,this weekend coming up is to remember to take home my HVAC vacuum pump... and also i'm going to replace my evap with a modded water block , insulate everything and make it a direct die phase change system.
One thing you may want to think about is the block orientation... of course the center would be where you put your cap tube, but where does that leave the outlet or suction tube? At the top? If its at the bottom,gravity is going to make the refrigerant leave the block quicker than if it was at the top of the block... But I guess with a maze type block it is going to go where it has to go... out the suction tube. Running the suction at the top would allow the block to be flooded with cold refrigerant,which is a good thing. I will be using a modified Z4 (cringe) water block I bought years ago... replacing the aluminum top with a copper top including fittings to solder on the cap tube and suction lines. I used a simple brass cap with a hole drilled in it for a place to put the cap tube.
Edit to add: Are you sure that the refrigerant oil travels throught the system on a hermetic type compressor? I don't think it does...Not like an automotive AC system where the oil is misicible and mixes with the refrigerant. I may very well be wrong, but I didn't see any evidence of oil inside of the piping on my system.
I beleive it stays in the crankcase for the most part.

redleader · 06-10-2004, 06:05 PM

Thanks for the input. Its good to know i'm probably being caustious about pressure, even if R134a is a much lower pressure refrigerant then the R290 I plan on using. I think i'll try it and see.

Also, the oil does travel through the whole system. Fire up your compressor while its detached from the system and it will shoot oil everywhere

crane · 06-10-2004, 07:59 PM

Quote:

Originally Posted by redleader

Thanks for the input. Its good to know i'm probably being caustious about pressure, even if R134a is a much lower pressure refrigerant then the R290 I plan on using. I think i'll try it and see.

Also, the oil does travel through the whole system. Fire up your compressor while its detached from the system and it will shoot oil everywhere

Thanks for clearing up the oil travel mystery for me....makes sense,I suppose.
Fridge compressors are a new animal to me...i just tinker with auto AC , but always like to learn more.
R290 gives lower temps with the trade off of higher pressures ,correct?
We should brainstorm a bit on evap designs as I'm ready to make one....

Edit to add: I just bought 3 ea. copper -phosphorus brazing rods from Ho' Depot.
Tensile strength 40,000 psi , working temp 1310 f to 1475 f. Says on the pack "joins copper to copper without flux and copper to copper alloys (brass & bronze) This alloy is widely used in refrigeration,electrical and plumbing jobs to join copper tubing and pipe,electrical wires and flexible tubing. Excellent for soldering copper tubing connections even with water in the lines ".
I'll let you know how they work.

redleader · 06-11-2004, 05:10 PM

Sounds good. I'm extremely busy at the moment with school, but that'll end in a few weeks and i'll have more time to devote to experimenting.

I have been thinking about design. One big problem is that the whole thing runs on its side. So with a maze design a lot of the base surface is not always in contact with the fluid because the liquid will tend to fall to the bottom side of the channel and the gas will float up to the top. This isn't a critical problem, but i suspect that an ideal arrangement for an evaporator is going to look a lot different then your typical waterblock because of the need to keep gas out of the way. It would be pretty interesting to see if CPU temps drop when you run an evap made like a Maze on its side.

Regarding pressure: R290 evaporates at a lower temperature then 134a, so at a given pressure drop between the high and low side it will be quite a bit cooler. However i think that in the real world the difference would be less then you'd expect just because it'll be somewhat harder to compress and thus you'll get a lesser difference across your two sides. I'm using it just because its virtually free and doesn't require any certs. Plus its compatable with R-22 and R134a systems. Downside is that when it warms up the pressure increases a lot more (though still less then R22).

Good luck. Definately post anything you try.

06-09-2004, 08:34 PM	#1
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Need advice on building a phase change evaporator This is an extension of my last thread where it became clear that using a swiftech block as an evap was not an option. In that thread it was pretty obvious that i didn't really understand the pressures at work in an evap. So heres what I need to build a block capable of: Must be able to withstand at an absolute minimum 200 PSI* of inside pressure. Ideally more in case i ever want to switch to a higher pressure refrigerant and for safety in all cases. Must be able to return a small amount of liquid oil to the compressor. Must have a large enough internal surface area for ~150w of heat to be transfered to the refrigerant during phase change. My current idea: Modify an old Maze 2 classic block: First step would be to unsolder it. Then I'd need to clean it out and braze it together and replace the barbs with fittings for a gas line and a cap tube. Finally I'd bolt on a socket A to 754 conversion bracket. As an added bonus the bracket would allow me to bolt the block together in the four corners, further increasing its strength. Again though this brings me back to pressure. I remember Dangerden pressure tested the old maze1 blocks which were very similar at 1000 PSI. Since I'll be brazeing and bolting I think i'll hold at least that much pressure, but i have no concept of what high pressure like this really does to a joint. Does this seem reasonable? Also, about brazing. Is there some special braze I should be getting, or should i just go with the regular stuff used in AC systems? Will ordinary copper braze work when i braze in brass fittings? Finally does braze use flux? I've only worked with plumbers solder (though i have a MAPP torch). Edit: Noticed i misspelled the topic sentence Last edited by redleader; 06-11-2004 at 05:52 PM.

06-10-2004, 01:26 AM	#2
crane Cooling Neophyte Join Date: May 2002 Location: so cal Posts: 82	I've been working on a retro-fitted water cooler recently...took the plunge and finally evacuated the little .5 HP compressor and chopped off the evaporator coil to try some experimentation with different configurations. First thing i did was solder on a service port at the process tube on the compressor... i figured I would try my luck at pulling a vacuum and recharging before I went too far with evaporator mods. I forgot my 3CFM 2 stage vac pump at work,so i was limited to using my little hand vacuum pump(don't laugh) to pull a vacuum. Moisture wasn't a concern as it was only open a few minutes. I pumped like hell and got it down to -25 and shot the gas to it and it started happily and cooled just fine...seems to me the little hermetic compressors are pretty tolerant of under/overcharge as i was able to fine tune the cooling by adjusting the R134a amount. It holds a whopping 1.75 oz. I then decided that since the evac/charge scenario was simple enough... i would cut out the original evap and try some alternate configurations. My first evap replacement was a coil of 3/8 copper with the cap tube soldered in one end and the suction tube soldered on the other end of the coil. I then did the hand vac deal and recharged again... Once again it fired right up and made a big chunk of ice on the coils. I used some old plumbers solder I had lying around forever.. can't even read the type on the roll its so old. It seemed to hold pressure fine and i wasn't a bit concerned about something popping loose under pressure. I wasn't too happy with the size of the evap coil I had made, so once again...chop chop and this time I made a coil out of 10 foot of 1/4" copper tube and soldered it in place,using the same crap solder i used before. Only difference was i bought a little container of tinning flux from Ho' Depot and that stuff is the cats ass for soldering copper. I will look at the bottle when i get home this weekend and post the brand and type. To sum it up.. i wouldn't concern myself with brazing anything on the phase change system...just use a good flux and plumbing solder and have at it. My results from playing around with my chiller were really quite good... it makes zero degree celsius water and and I had my chip down to 8 degrees celsius. But without any condensation proofing and insulation, I didn't run it but a little while as it was sweating terrible. Next step,this weekend coming up is to remember to take home my HVAC vacuum pump... and also i'm going to replace my evap with a modded water block , insulate everything and make it a direct die phase change system. One thing you may want to think about is the block orientation... of course the center would be where you put your cap tube, but where does that leave the outlet or suction tube? At the top? If its at the bottom,gravity is going to make the refrigerant leave the block quicker than if it was at the top of the block... But I guess with a maze type block it is going to go where it has to go... out the suction tube. Running the suction at the top would allow the block to be flooded with cold refrigerant,which is a good thing. I will be using a modified Z4 (cringe) water block I bought years ago... replacing the aluminum top with a copper top including fittings to solder on the cap tube and suction lines. I used a simple brass cap with a hole drilled in it for a place to put the cap tube. Edit to add: Are you sure that the refrigerant oil travels throught the system on a hermetic type compressor? I don't think it does...Not like an automotive AC system where the oil is misicible and mixes with the refrigerant. I may very well be wrong, but I didn't see any evidence of oil inside of the piping on my system. I beleive it stays in the crankcase for the most part. __________________ dirt rules Last edited by crane; 06-10-2004 at 01:32 AM.

06-10-2004, 06:05 PM	#3
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Thanks for the input. Its good to know i'm probably being caustious about pressure, even if R134a is a much lower pressure refrigerant then the R290 I plan on using. I think i'll try it and see. Also, the oil does travel through the whole system. Fire up your compressor while its detached from the system and it will shoot oil everywhere

06-11-2004, 05:10 PM	#5
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Sounds good. I'm extremely busy at the moment with school, but that'll end in a few weeks and i'll have more time to devote to experimenting. I have been thinking about design. One big problem is that the whole thing runs on its side. So with a maze design a lot of the base surface is not always in contact with the fluid because the liquid will tend to fall to the bottom side of the channel and the gas will float up to the top. This isn't a critical problem, but i suspect that an ideal arrangement for an evaporator is going to look a lot different then your typical waterblock because of the need to keep gas out of the way. It would be pretty interesting to see if CPU temps drop when you run an evap made like a Maze on its side. Regarding pressure: R290 evaporates at a lower temperature then 134a, so at a given pressure drop between the high and low side it will be quite a bit cooler. However i think that in the real world the difference would be less then you'd expect just because it'll be somewhat harder to compress and thus you'll get a lesser difference across your two sides. I'm using it just because its virtually free and doesn't require any certs. Plus its compatable with R-22 and R134a systems. Downside is that when it warms up the pressure increases a lot more (though still less then R22). Good luck. Definately post anything you try.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)