My AC Chiller Build, Suggestions Needed

[Budwise]

Im working on this chiller build and am just left sitting here scratching my head as to how to go about building or getting this evap into a res. It seems to me that with these pipes there is not alot of slack for bending them to get the evap into a cooler so i think i am pretty much left with very little movement options and trying to figure out how to build a res up around it. Any tips or ideas?

The plan for now is to use a temp switch so that at a certain temp the chiller will turn on. Until i move in 6 months to a bigger place i plan to just run sub ambient temps yet stay above the dew point so needless to say im gonna need a temp probe to keep it above dew point.

Here it is with the fan housing removed. The fan is frickin LOUD, so i figured id use something else once i get the res figured out.







As you can see, it doesnt appear to have much slack to move the evap very far... I dont know how much help getting a pipe bender would do.





Im curious as to that little section of coiled up pipe, what is that? Im in need of some suggestions as i am kinda stalling out on this one...

[ricecrispi]

what ive seen is guys will place the evap in a really large plastic storage containers or ice chest cooler.

What you have to do is bend the tubes to the evap gets positioned off the base. Unwind the cappillaries towards the evap. Now bend the larger tubes to where you want the evap. You can place the evap into a container underneath or next to the base.

little coil pressurizes the liquid. Determines temps for idle and load. Not sure how it works but depends on the length of the tube.

Extreme overclockers or extreme systems is the best place to go. I think there is a sticky on one exactly like yours on extreme systems
http://www.xtremesystems.org/forums/...t=22968&page=1

[Ice Czar]

split level condo
there is easily enough line to drop it below the level its currently on with the added benefits that it will need no throughwall fittings and the "upstairs" will have an unobstructed breezeway

you need a copper tube bender is all
get a reasonably good one
like this


pull that peice of putty and unwrap the thinline

best take it outside and hose it down first however

[Budwise]

UPDATE:

I found the old soft sided cooler that i was looking for. Although it is a tight fit, it does fit.



The problem is the cooler is not exactly leak proof.



After about 10 minutes it went from tap water to freezing cold icy water, yummy...



Now i gotta figure out how to get that cooler water tight. The lining in it looks pretty rugged and like it would be water tight, but sadly its not. Any ideas? All i can think of is to get some sort of heavy duty rubber bladder to fit to the inside of it. Anyone know where to find something like this or have any better ideas?

[Budwise]

i decided to scrap the soft sided cooler idea as it seemed like it was gonna be a pita.

Here's what a trip to the store has brought.



One thing i am worried about is dirt particles getting into the loop from the cooler. As you can see in the pic there is some little specs floating around. I have rinsed off and cleaned the evap and cooler already and still it seems that stuff ends up floating in it and i dont want that stuff in my loop.



The better cooler definately helped drop the water temp faster.



What if i were to simply run my rad from my current watercooling and submerged it with the evap? Would that be much more inefficient?

I still need to figure out how to regulate the water temp as to avoid condensation. Can someone help me out and link me to a thermostat i could use or something? cuz im not sure what to look for...

[Ice Czar]

1. if the "stuff" floating didn't come from the air (open top) its debris that was previously caught in the evaporator (one of the reasons I mentioned to hose it down). A filter will of course cutdown the flow. Ideally you keep the pump matched to the resistance in the loop at the flowrate that gives youy the best performance. So the less resistance the better. A more powerful pump adds more energy to a loop, a very small amount as direct heat, the majority as kinetic energy, but both impact the lowest temperature your able to maintain.

2. Just placing the rad into the cooler will be a less than ideal heat transfer solution, ideally (given what your working with) the return from the computer would be on one side and the pump on the other forcing the flow to move through the evaporator
(The ideal solution is a heat exchanger with the rerigerant on one side and the water on the other, convoluted coil like a wort chiller or a plate heat exchanger being another approach, that of course means recharging the system ect)

3. Regulating the water temperature to avoid condensation is far easier said than done, to begin with room temperature is but one variable, the others are barometric pressure, relative humidity and the temperature of the block (which changes with the ambient temprature cooling the condenser)
The dew point moves around on you. Trying to get sensors that will compute that then compare the temperature of the block and turn the AC on and off is the exceptionally hardway to try to solve the problem.
Insulate the blocks to prevent condensation.

but just to illustrate the point

Psychrometric Chart

The psychrometric chart gives the relative humidity for a given wet-bulb and dry-bulb temperature. To use, find the vertical line for the dry bulb temperature at the bottom of the chart, and the slanted line for the wet-bulb temperature at the curved top-left edge. Find the intersection point of these two lines, and read the percent relative humidity from the curved lines nearest that point.

if you want to try to do it manually
Id recommend a cheap weather station with Relative humidity, temperature and barometric pressure
and a demo of http://www.megawatsoft.com/humidair.asp?cat=5
if the block drops to the dewpoint temperature (for the given dry bulb temperature, relative humidity and barometric pressure), you have condensation

(hmmm... the demo has just one preset barometric pressure )

example (at 8bar)
room temp (dry bulb) 70F
relative humidity (RH%) 40%
dew point 44.5F (block hits that you get condensation)

room temp (dry bulb) 70F
relative humidity (RH%) 70%
dew point 59.7F

room temp (dry bulb) 70F
relative humidity (RH%) 98%
dew point 69.4F


again

room temp (dry bulb) 90F
relative humidity (RH%) 40%
dew point 62.3F

room temp (dry bulb) 90F
relative humidity (RH%) 70%
dew point 78.8F

room temp (dry bulb) 90F
relative humidity (RH%) 98%
dew point 89.3F

of course 8bar is kind of useless (its a demo after all) that works out to some 236 inches of mercury
where I am its currently 29.77 in (with a dew point of 14F, 14% RH, and 62F dry bulb)
so the "air" pressure in my examples would be like 8 times "standard" or about 267 feet underwater
the chart above is a "standard" atmosphere (29.92 inches of mercury (Hg) or 1.01 bar or 14.6 psi absolute)
finding a weather service that provides your dewpoint might be a better approach to do "manual" control
that of course assumes your exterior temp matches your interior

if its "dry" where you are you'll likely be OK with quite a decent subambient potential,
if on the other had its "muggy" youre screwed if you drop just a little below ambient

[Ice Czar]

found a calc that actually works
http://www.michell.co.uk/downloads/index.html#humid

[Budwise]

my plan was to just keep an eye on the weather sites and watch the dewpoint. I guess that wont be reliable enough?

I decided on submerging the rad rather than run a line directly due to the fact that the evap fins are aluminum and my blocks are copper and corrosion is not my idea of fun. When all is said and done, if this is gonna end up too much of a pain in the butt ill just toss the chiller idea and force myself to be content with watercooling.

[Ice Czar]

would be simple enough to employ a sacrificial anode
zinc would likely be the best choice, with a cooler at least its relatively easy to "flush" the system and swap out coolant.
Of course cobbled together chip chillers are a pain in the butt, your cutting corners.

1. youve already invested considerable effort and money you might as well "try it"
2. for the first runs, provided the dewpoint is low enough, no insulation required, galvanic corrosion generally takes some time, if you employ an inhibitor (water wetter) you fine for testing
3. if its looking good insulate and employ a sacrificial annode, or determine the effectiveness of your inhibitor and how long till you need to swap out coolant
4. upgrade with the help of a local AC Tech to a convoluted counterflow wort chiller ($110)



5. once your new friend the AC Tech is fascinated with your insane plans, have him help you build a cascade chip chiller
with another AC unit and refrigerant to refrigerant heat exchanger cooling the first (which you again swap out refrigerant on to a lower one),
then toss it into a case of sorts with wheels

[ricecrispi]

Some nice work Budwise. You done pretty good so far.

I wouldn't continue unless you want sub zero temps which is possible. Using this as a plain chiller without condensation isn't that hard but not worth it. I faced the same problem. With the amount of work, size of the unit, $$ dinero, and time, you get little in return in terms of perfoamnce except for the joy of completing the project which is priceless.

[ricecrispi]

If you want to continue the project

1) To clean out the radiator I first get some brake cleaner and spray a whole can thru the evap and then condensor. Helps remove some gunk and some of rust corrosion off. I would remove all the electrical stuff and high pressure wash the rads/heat exchangers at carwash or with garden hose. If you dont want to remove the electronic I would wrap it in a few plastic bags and wrap it with duct tape several times. My dad sprays the AC unit at his place dozens of times and its works.

2)Remove the stock fan and build a shroud for some 120mm fans at low speed will reduce noise and EDIT (increase temps)
3) instead of running inline put a rad in the container and submerge a pump. The rad makes the setup less efficient and the inline pumps helps produce some heat directly into the loop. might reduce condensation.
4) Having a seperate loop in the container you don't have to worry about corrosion as much. Still drop a zerez or water wetter and anti-freeze in the res.