AIRCHILLER + Single-stage (worklog)

[piotres]

Hi

Mine next project - rather won't be boring like all 2 stagers - it'll be cooling chamber (to put inside mobo + all other stuff, except hdd and psu ) plus SS , all on one plate ...

Specs :

Compressor: two Danfoss SC18CL - one for airchiller, one for SS
Evap for SS: Steeper 40mm tall/42 mm diam., enclosure for all newest CPUs, 1.0 meter braided DN10 suction line (with
stainless steel mesh)
Evap for airchiller : Evaporator from split A/C unit
XV: captube 0.9 mm for SS, Alco TI-HW for airchiller
Condensor: two 1.9 kW + 254 mm fans
Refrigerants : R507 in both
Case : custom made, welded frame, custom perfored side panels, all powdercoated on black color, wheels.
Chamber :HARD moded Ladycook small fridge
Others: Control panel on top of chamber (switches for both SS and airchiller, switch for PC), Danfoss controller for airchiller, Frigomec 0.88l receiver.









Bone stock frigde ...brand new ... .






















Some reduction on parts used in frigde .







Mine new welding machine- MIG/MAG Esab C250 3ph .





Mine father helped me with welding a bit and later with grinding .



That was only 4 fun ...just a try on trash piece of metal .





















Fridge got small accident and had must been painted ...so paint is fresh high gloss black acryl ).





















I decided to weld suction line holder for more safety .

























Clear coat this time also .



















First time I've must use ladder when I build cooling system .











First tests, temperature inside chamber without any fans mounted.





Here with two 120mm - nice gain .









I was pumping hot air inside chamber from hairdrier working at 1st speed and taking over 700W energy - during 30min test temperature moved from -31*C (I've wait long, for lowest temperature) till -10 C - thermal probe was mounted in piece of cooper put inside chamber. It's not final .





Some heavy piece of brass inside - that's tip from future owner of this unit to use big pieces of cooper (but I don't have it, only brass) as a "cold magazine" .





Probe inside small cooper piece (don't have bigger ).



That's what we should say "ARCTIC FAN" .



That's loadtest of SS loop - make -40C@260W but I don't think that's most important part of this unit .

Regards
Peter

[biglipzit]

Wow very impressive as usual piotres! I wonder if the board in the airchiller will be affected by condensation at all. I love -40 at that load on a SS. You are the chiller god!

[quintus]

Hi piotres,
Just an ideea: the enclosure of the hole setup shoud be purged with CO2 or N2
What about the hdd and dvdrw and psu, what are the plans?
regards,
q

[biglipzit]

Its a refrigerator... I doubt that is has the sealing air tight capabilities to do so and when benching, to do anything you will have to vent the gas out. Won't be worth it. Just have to run it for a while before putting equipment in to ensure ample humidification. Probably worth setting up in a dehumidified room.

[sdumper]

Love the work and attention to detail ... but am wondering how your going to condensation proof the mobo, psu, videocard ect...also wondering how it will perform with a hot video card inside and psu inside. I think my 4870x2 is much hotter than a blow drier...

Cant wait to see it tested with live equipment its going to be very cool ... yes the PUN was intended

Again wonderful work!!

[t800]

Since im getting this unit, i give some answers

Its a refrigerator... I doubt that is has the sealing air tight capabilities to do so and when benching, to do anything you will have to vent the gas out. Won't be worth it. Just have to run it for a while before putting equipment in to ensure ample humidification. Probably worth setting up in a dehumidified room.

Love the work and attention to detail ... but am wondering how your going to condensation proof the mobo, psu, videocard ect...also wondering how it will perform with a hot video card inside and psu inside. I think my 4870x2 is much hotter than a blow drier...

Cant wait to see it tested with live equipment its going to be very cool ... yes the PUN was intended

Again wonderful work!!

- Actually good fridges are airtight and reason why they get frost is because ppl will open and close fridge door. And if some reason it would not be airtight, i have some ideas in my pocket to make it air proof.

- Unit is for daily use

- Of condensation, i will use many silica gel moisture bags to remove moisture (eg one bag is capable remove moisture from 1..1.5 cubic meter of air). Basically i put parts inside (in dehumidified/very low moisture room) , tens of gel bags and then do several test runs (only compressors are turned on, not computer) and let bags to catch the moisture. No moisture, no condensation

-4870X2 is far away from hairdryer, it produces about 300w heat. But even -45c air temprature, gpu's wont be below 0c under load, due air temp condictivity is bad...but freezer gives possibility to run 4870X2 fan very high rpm due lowers the incoming noise from fan, what normally is way too much

-PSU,HDD,DVD will stay outside of system.

[EvoCarlos]

it looks great i must say but

as for the frost
our home fridge has a fan so one opened and closed the air is sucked out by the fan as if to create a vacuum in the case maybe you could do something like this?

would be easy to set up just a fan with a rubber flap on the back so when the fans off the rubber flap seals it shut

[Sink]

Vacum are bad if you want the air to cool something down really.

[piotres]

Please nottice that on few last picutres there is a lot of frost inside because I've been pumping hot air by hairdrier inside the chamber for 30 minutes .

[biglipzit]

Please nottice that on few last picutres there is a lot of frost inside because I've been pumping hot air by hairdrier inside the chamber for 30 minutes .

Oh yeah that would do it. I am sure it will work great, even for everyday use. Once it is setup and running there won't be much need to open the fridge. Should be pretty neat. Can't wait to see what performance will be like when it is all up and running with the PC inside.

Keep us posted!

[quintus]

The fridge is air tight enought.
I think will work fine just having a CO2 botle of say 2 liters, open it up and blast the enclosure, purge all air outside, when you set up the system or when u open up the door for dealing with the hardware...and u will have 0 humidity silicagel cant remove all water from air

[t800]

The fridge is air tight enought.
I think will work fine just having a CO2 botle of say 2 liters, open it up and blast the enclosure, purge all air outside, when you set up the system or when u open up the door for dealing with the hardware...and u will have 0 humidity silicagel cant remove all water from air

I think with silica gel packs i can remove most of water from air (i hope so), that rest is just so minor that it can't do any harm (we are talking 1~ gram).

Of course if this method is not sufficient, then replacing air with co2 or nitrogen is only solution.

[piotres]

Sorry pics should be working again tomorrow - I've reached limit for transfer for basic web account (3.4GB sucked from http in 12 days by You (mounthly limit is 3gb) - thanks a lot people :P - really I'm happy that so much people look at mine topics) so I must have upgrade for plus account which should be active tomorrow.

[one_servant]

Rock and Roll Peter. Keep up the great work. I look forward to seeing where this goes.

[piotres]

More photos, unit is nearly finished :

















I've added 4 pcs of 80mm fans behind the evaporator - to mount them I had must cut hole in back of fridge which later I've covered with styrofoam and new cut piece of steel plate. I couldn't mount fans before evaporator because in that way motherboard with long GPUs can't be put inside chamber (too little space) .



Testing Rig.



Noloaded I've reached nearly -49.5C air temperature inside chamber (I haven't shooted that on photo, only -48.2C) .



From that temperature I started load testing - pumping hot air throught 30 mintes produced by hairdrier (taking over 700W from 230V) by hole for cables on back of fridge .





Big brass block (that round one) temperature after 30 minutes loadtest (remember - over 700W) .



And same but measured by small cooper block .





SS unloaded.



SS loaded with 260W after 1 hour test .

Now unit is standing for next 2-3 weeks for testing if it's leak free for sure (of course leak detector sprays haven't showed any leaks but I must be sure), after that - packing & shipping .

Regards
Peter

[piotres]

Hello

Unit is already 100% completed :



Finaly, small cooper block after temperature (so also air temperature inside chamber) pumping hot air by hairdrier (30 min test) - start of test was from -43*C.



Big brass block after same test .



Then I start some real loadtests







































And that's all, unit is 93kg, total pack weight is 157 kg without palette . As You can see packing is heavy, but it must be heavy to be safe .

Regards
Peter

[Duniek]

ladycook overclocking lol
great job Peter

[tiborrr]

1-5-7kg! :O

Respect Peter!

[sdumper]

Nice house too...oh and yes the unit looks like its mass produced by a factory very very professional in appearance much respect!!

[SexyMF]

Unit looks good but I suspect it will be a motherboard killer due to condensation.

[sofos1990]

Congrats piotres this is a innovative construction

[Rise]

that is awesome, i love seeing *inside the box thinking

[eligray]

Very great looking and innovative system there Piotres

And to Rise: Condensation will not be an issue I'd imagine. Most would form on the evap anyways. And the Silica will get it out of the air pretty well.

[systemviper]

Wow you do great work, love the ladder to access the project.

I just love innovation, and that's what you keep pumping out,

Top NOtch!

[t800]

Okay, After months of tuning & testing finally so small results available of this cooler. Currently bit different look of cooler




Added sound insulation and rerouted aircirculation inside of compressors, eg air is doing S loops before getting out and walls are insulated with bitium + softcells. Currenly doing 55-60db from 2m away (using iphone as meter :P)

Also controlbox added quickswitch for resetting cmos..eg holding switch does reset. The fridge door is sealed with silicon and havind side bolt+screw lock system that door is not opened by accidently.

Before closing door:




Air inside is circulating from behind blowed with fans end of fridge (door), and there is goes behind mobo (eg where all wires are) back to evap.

Small results with set to keep air 0+ to -15c ...takes 15mins to cool down and 10mins to cool back to -15c.

Result after running 30mins of 3Dmark03 NatureTest

idle


And finally, the myth of condensation, does it exsist? Simply no..i have opened door during run, no water visible and during run haven't have any symptons of condensation failures so far.

In future pushing more cold (-40c) and see what happens