[cnd]

*interesting* - if their 180 watt "fridge inside the PC case" idea keeps an overclocked athlon at -2 degrees, it means we are at least in the ball-park of possability with regular little fridges, and probably definitely within the bounds of reasonable with larger ones?

[christopherpostill]

I have a modified morphy richards mini fridge in my room - i'm one for gadgets i'm afraid.

Took it apart at the weekend, dont know if any of this will help you...

... But you might want to look into removing the fan grill and re-fitting the heatsink with arctic silver thermal compound. My fridge has gotten colder by a couple of degrees i think since i did that... The amount of condensation it produces, however, is pretty amazing - i will probably have to empty the drip tray once a week or so to stop there being a flood every time i open the thing. Nah, its not that bad... but it's noticable.

You might also want to replace the standard fan with a higher power type - mine has 2x 80mm in - though still the standard ones.

I dont know if the below pics will help you in any way with regards to the iner workings on mini-fridges, but its worth a go...!

Regards

Chris

See pics:

[Me__2001]

thats very intresting chris, how long does it take to cool a can of drink ? and at what temperature is it roughly?

now if the PSU is to be mounted on top of the case the wires may be a bit short but until a fridge of the right size is found it will be hard to tell exactly, i have found some interesting things

first i have come across some extension leads for - mouse/keyboard (they are the same right ?), VGA. speaker, USB and firewire ports can be used from the old case if it has them. so the ports can be mounted on the side to allow everything to be plugged in alternatively a hole could be cut for the back plated of the motherboard

second john lewis are doing a zanussi fridge that looks to be the perfect size for £99 the internal measurements should be more than enough room for a PC

http://rswww.com/cgi-bin/bv/browse/M...toid=-97793431

http://rswww.com/cgi-bin/bv/browse/M...toid=-97789351

http://www.johnlewis.com/Shopping/Pr...U&Id=230210737

[christopherpostill]

Define "cool" - heavily chilled?

I can tell you that its quite alot colder than your average kitchen fridge. Most are i think - just because of their size etc...

Got mine from John Lewis a while back - was about £80... The insides aren't filled up with this heating malarky either - fridges are for cooling not heating!!

Can be run off an AC or DC supply, and draws 75 watts.

[Kenny]

Enter the randomness.... Get a watercooling loop for the CPU and GPU, then the noisiest thing left will be the hard disk(s).... Get some long power cables and S-ATA/IDE cables and put it/ them in a mini-fridge?

Or not.

Kenny

[Raje]

You will most likely hear nothing out of the pc fridge. But Hey, going SATA is going to be better all around, and right now it dosn't cost much more. Plus those wires will be much easier to run vs. IDE's. Also they should sell extension for any wire, if not I'm sure we could find a way using adapters or something. As far as that huge amount of condensation, we really should be going low tech as well. As I mentioned earlier, their is this contractor grade moister absorber I saw at my local Home Depot. It can supposedly draw 1 cup (1/4 pint)of moister a day or something. And the entire thing is the size of a small coffee can. It all goes into a container which we could mod with a hose to drain out of the case, and the materieal looks like crushed styrophome, and is easily changed.

Also I don't think we are taking into consideration just how much heat we are reducing from this pc to start. WE are mounting the PSU to the outside or cutting a nitch in the freezer(that a lot of heat right their). And we are running a water cooling kit to our cpu, mem, and video card(this can cool most pcs on its own fanless). We may need to do the north bridge or watever, or put in that 190mm fan.

Can someone test this for us--- can any one remove thier PSU from thier case and run it for a while and see if it has made a temp difference, or at least tell us how much if any difference is noticed??

[TommyTech]

Hi, I signed up just to participate in this discussion! I'm from the States, and am jumping in on this a bit late into the discussion... but ene-ways ... I got the fever about 48 hour ago to figure out how to get greater cooling capacity out of my water cooling rig. I run dual 2.0 GHz Sledgehammers (AMD Opteron 246's). I am in a state of flux right now. I built my rig about 9 month ago around an MSI K8T Master2-FAR (MS-9130) mobo, (one of the few standard ATX dual Opt mobo's). I have been using an all Innovatek water cooling system with a dual reservoir setup with a single 120mm fan/radiator for cooling both CPU's and my northbridge. Because it's just the one small radiator, I get about the same temps I would with air cooling. This got me thinking of upgrading to a huge external (since I have no more room in my 4U server case) radiator with a large (slow, quiet) 14"+ fan. Then I thought... if it's going to be external, it would be relatively easy to use some other method of cooling the water... what about refridgeration?? I began taking apart an old cube-fridge I have, to see how I might take the compressor, radiatot etc. and the freezer cage/element out (which I was thinking I could use to chill water in an an external reservoire by either wrapping it around ... or submerging it *in* the reservoir. I too want to avoid the dreaded condensation effect, and figured my concept of a hybrid water/phase-change system would help avoid this.

I did a feverish internet searche last night into today, bookmarked about 2-dosen links, most of which I havn't even looked at, mainly because your threads here seemed the most well-developed and committed -- (the rest are undoubtedly a lotta logic-impared RAID 0-loving gaming morons!). But I did come across this idea which seemed closest to my own that you might all want to take a look at in case you hadn't already:

This guy built a fridge cooler concept that, while it doesn't house the PC itself, might be built to have room for a 6-pack of beer it seems! http://www.overclockers.com/tips798/

Great reading guys! I'm impressed by your commitment to this idea! I want to help.
----------------------------------------------------------------------------------------------------------------------------------
My current mobo (as reviewed by GamePC):
http://www.gamepc.com/labs/view_cont...ualduel&page=1

My soon-to-be new mobo (also as reviewed by GamePC):
http://www.gamepc.com/labs/view_cont...snf4pro&page=1


Also thought I would mension: While my most recent new concept is to keep the PC it it's *own* case, I also had thought a while back of a concept closer to what chy'all are talking about ... of using a sleek black wine cooler (yes, the nice elegant ones with the lightly tinted glass window) as a sever cabinet or case.

[scotchfx]

This may be a bit off topic but I haven't found too many threads regarding modding the cooling (air - not watercooling just yet) on the MSI K8T Master2-FAR. I've been experiencing a couple issues with my workstation since I installed my HPT1540 PCI RAID card:

2 x AMD Opteron 248 (2.2GHz)
MSI K8T Master2-FAR / Award v1.1 BIOS
2 GB of TwinX1024RE-3200LLPT (4 x 512MB)
2 x 74GB WD Raptor (*boot* RAID_0 on VT8237)
Highpoint RocketRAID 1540 PCI card
2 x 250GB Maxtor Diamondmax9 (RAID_1 on HPT1540)
80GB Western Digital (IDE2 port)
Sapphire Atlantis Radeon 9700 128MB (AGP 8x)
Enermax EG651P-VEFM (550W 24-pin ATX PSU)
Lian-Li PC-69 Aluminum Mid-Tower Case
WinXP SP2 and all critical updates

In short, I've been getting lockups everytime I attempt to transfer a large amount of data (1GB+) via USB. I was thinking this might be some kind of thermal issue as the 2x250GB Maxtor drives somewhat block the main intake fans on my case. To remedy this situation I'm currently installing another pair of of 5.25 intake fans in my last remaining empty drive bay along with a Cyclone PCI exhaust fan/blower to help provide some additional airlflow.

Now for my most tangible question...

I wanted to replace the stock MSI northbridge DDR heatsink with a more customized Vantec ICEBERQ CCB-A1A (copper) heatsink, however after finally removing the stock heatsink from my mobo I find that the Vantec heatsink will not fit!

Can anyone suggest a heatsink that I might be able to use?

In the interim I am going to replace the stock heatsink (add some Artic Silver 5) and add my other mods. Hopefully some extra intake & exhaust will help remedy some the odd issues I've been seeing...

of course, any advice on this would be Much appreciated.

Thanks!!!

[Raje]

Originally Posted by scotchfx

This may be a bit off topic but I haven't found too many threads regarding modding the cooling (air - not watercooling just yet) on the MSI K8T Master2-FAR. I've been experiencing a couple issues with my workstation since I installed my HPT1540 PCI RAID card:

2 x AMD Opteron 248 (2.2GHz)
MSI K8T Master2-FAR / Award v1.1 BIOS
2 GB of TwinX1024RE-3200LLPT (4 x 512MB)
2 x 74GB WD Raptor (*boot* RAID_0 on VT8237)
Highpoint RocketRAID 1540 PCI card
2 x 250GB Maxtor Diamondmax9 (RAID_1 on HPT1540)
80GB Western Digital (IDE2 port)
Sapphire Atlantis Radeon 9700 128MB (AGP 8x)
Enermax EG651P-VEFM (550W 24-pin ATX PSU)
Lian-Li PC-69 Aluminum Mid-Tower Case
WinXP SP2 and all critical updates

In short, I've been getting lockups everytime I attempt to transfer a large amount of data (1GB+) via USB. I was thinking this might be some kind of thermal issue as the 2x250GB Maxtor drives somewhat block the main intake fans on my case. To remedy this situation I'm currently installing another pair of of 5.25 intake fans in my last remaining empty drive bay along with a Cyclone PCI exhaust fan/blower to help provide some additional airlflow.

Now for my most tangible question...

I wanted to replace the stock MSI northbridge DDR heatsink with a more customized Vantec ICEBERQ CCB-A1A (copper) heatsink, however after finally removing the stock heatsink from my mobo I find that the Vantec heatsink will not fit!

Can anyone suggest a heatsink that I might be able to use?

In the interim I am going to replace the stock heatsink (add some Artic Silver 5) and add my other mods. Hopefully some extra intake & exhaust will help remedy some the odd issues I've been seeing...

of course, any advice on this would be Much appreciated.

Thanks!!!

Well before I look into your problem friend I want to advise you of two things, your question is WAY off topic, and this thread has really slowed down lately. We have basically put in all the thought and logic we can so far. Until one of us actaully tries to build this thing with some feed back you may not get many views. You really want to create your own thread and put it in hardware support. Thats why I asked the admin to move it for you. You will get better results.

[SithMaster]

The fridge PC:

Yes it is entirely possible from a mathematical standpoint to build a PC in a fridge. A couple concepts you may not be expecting to deal with...

A couple people have noted that removing the heated air from the unit is a key element to cooling. This is true, but the idea of the fridge is to remove the energy from that air through a simple evaporation process. The whole point of moving air out of a conventional PC is to maintain a low average case temperature. The lower the temperature of the air entering a heat exchanger, the more heat it can carry away. Thus making sure the temperature inside the fridge is uniform (well stirred) facilitates the greatest energy removal. Fans need to be in place to ensure a rapid movement of air across exchangers, and about the fridge/case.

Condensation
ANY TIME YOU HAVE LOWER THAN ROOM TEMPERATURES IN THE VICINITY OF PC COMPONENTS, YOU RUN THE RISK OF CONDENSATION DAMAGE. It doesnt matter how you get the low temp matter there, the air around that point will be cooled and the partial pressure of water in the vapor phase will drop.
This is certainly a key problem to the fridge/case scenario but there are two answers to the problem. The first step is to make sure you don't bring fresh air into the case. Fresh air is warmer, and, like all warm gases, will carry more water vapor with it. When this air drops in temperature, condensation forms. A well sealed fridge/case is the first step to avoiding this. Second, chemical adsorption of the moisture. Chemists use desiccators to dry many samples in standing air. The solution is to simply make sure the air is constantly having water vapor removed. As the air gets drier, even liquid phase water will evaporate (to compensate the loss, a concept called equilibrium). This is often facilitated with a chemical called "drierite." Merely a solid platform of gypsum (calcium sulfate) it readily pulls moisture from the air. This will help remove any excess water from the air and is readily available (gypsum is a key ingredient in drywall).

Power-
Somebody was quite incorrect on the previous quote of removing energy being easier than heating. Heat sources are far more efficient than cooling. In the United States, a refridgeration is normally the largest single consumer of power in any home. Luckily, a computer doesnt necessarily draw all the power needed from the power supply at all times. In other words, a 450 W powersupply doesnt deliver that energy if there isnt a demand for it. If power was constant over your system, heat production would be constant, but systems change based on load. The fridge has a moderately low initial temp, but this will rise as use continues. Long term gaming may result in worse conditions than initial operation in a standard case. The same can go for water cooling. Both scenarios are easy to test, however. Simply place a thermometer inside the cooling resevior or inside the fridge. Your temp will go up at first, but should reach a "steady-state" profile where the heat generated equals that removed. (If the temp didnt go up there would be no drive for heat transfer so its not a bad thing, but the lower the better) This is where you want to measure the temp.

The real deal:
Water cooling vs fridge
WATER COOLING WINS HANDS DOWN. Why? It has nothing to do with temperature, it has to do with the convective heat transfer coefficient. Energy is transferred when a "cold" molecule in your air/water meets up with the "warm" molecules of your heat exchanger. Gas phase is a less dense phase. There are fewer molecules to run into. This slows the heat transfer process. Liquid phase heat transfer coefficients are at least 10x-100x greater than those of gas phase. Many people worry about silver and copper and stainless steel heat transfer coefficients but these are on the order of 2x or 3x of one another. In the end, the temperature limitations are hardly the problem of overall energy transfer rates. It all comes down to the heat transfer coefficients.

I have made systems with liquid nitrogen cooling here on campus. We put a ceramic cup filled with LQ N2 on top of a processor and played around :P It was actually interesting to see. A properly designed interface, with good contact, the right amount of thermal paste, and flat clean surfaces (and heatsinks DUST BLOCKS HEAT TRANSFER) is far more effective than an expensive mod replacement.

If you think your radiator is too small, try adding a recycle stream from post radiator, back to the reservior. Use thinner inner diameter tubing as a fixed pressure drop to prevent large amounts of water from flowing backwards. This will, however, essentially increase the size of your radiator (assuming pumping power is available). The ratio of water to components vs water to radiator is extremely important. Increasing your flowrate to the radiator will help bring down the overall temp of the water. Do you have good convective flow across your radiator tubes? (air blowing) Are the tubes relatively clean and rust free? Is your heat sink contact properly attached to your system?

Hope this clears up some things. Im finishing my master's in Chemical Engineering, I design industrial heat exchangers for a living.