anyone ambitious? almost silent subzero case?

Talcite · 05-29-2006, 06:45 PM

I was thinking in the car today, and i was wondering if it was possible to create a TEC adapted or a LN2 system case. The idea is just to combat condensation, and perhaps a bit of noise as well.

Why not build a case that encloses the motherboard, HDDs and the PSU? the rad and optical drives can be outside. Basically, the enclosed part can be airtight, and then put a hygroscopic material inside. Absorb all of the humidity, kinda like silica gel or something. The idea is if you remove enough moisture, there's no need to silicon your board, or dielectric grease the socket.

The PSU would have to be watercooled, and the CPU, northbridge, HDDs, and vid cards as well. Mofset's too probably. This is a crazy idea, but it would mean that we could have almost silent cases, plus very fast systems. The input panel would have to be sealed with a rubber gasket or something. Rad hoses can go directly into the enclosure wall. Put a metal tube welded into the wall with barbs, just so the liquid can get through. Rad would have to be outside the enclosure, because it needs heat transfer. Optical drives too, because they need to open up. All the other cables like front panel, ide cables, power cables, could be sealed with silicon. PCI cards, rubber gaskets. Make one side of the enclosure openable, so the silica gel can be replaced periodically.

am i crazy? or is it something someone wants to try? I'm still thinking about my waterblock, so i can't do this one.

bobo5195 · 05-29-2006, 07:12 PM

need to get heat out of the case as air is a good insulator and there are still things like mosfets which will heat the air.

Make a close fitting case around it and fill it with FC77 by all means, that will work 2.

Sealing will be a big problem

Long Haired Git · 05-29-2006, 09:34 PM

Put a radiator inside the case as well with air flow through it, to absorb heat and transfer to the rad outside the case.

Talcite · 05-30-2006, 12:23 PM

yeah. Sealing actually won't be that much of a problem. You don't need a complete seal. Only enough so the silica gel will be able to keep humidity low. However, a good seal will mean silica gel needs to be replaced less often. Hmm... rad inside and rad outside the case is a very interesting concept.

What is FC77?

bobo5195 · 05-31-2006, 01:21 PM

everything leaks. expensive liquid and leaking does'nt go well

flourinert (FC77 being the best for our purposes)
http://solutions.3m.com/wps/portal/!...i_IcFQEAOIxDW4!

Its water with flourine instead of hydrogen. Highly inert and looks and feels like water. could probably drink it and take a bath in it with some high voltage cabling.

Brians256 · 05-31-2006, 03:19 PM

Quote:

Originally Posted by bobo5195

everything leaks.

Lol! That is God's own truth! Yes. All we can do is spend time and money to try and slow it down to "acceptable" values.

I've seen this idea of cooling the air inside a sealed case and then transferring the heat to some outside radiator. One way might be better than a complete seal. What you could do is treat it like a refrigerator

Refrigerators have a drip pan, and this might work too. Have a very VERY cold surface inside the case (like a TEC) that freezes water vapor. Then, have a four hour timer to defrost it and dump the liquid water into a tube that dumps into a drip pan outside the case. This dries out the air without noise. Either put a really quiet fan onto the drip pan that evaporates the liquid or just empty it often enough that it doesn't grow stuff.

Off Topic: Isn't flourinert related to the "liquid" that high pressure divers use for breathing? I think I saw a rat submerged in some flourinated (sp?) liquid that had dissolved oxygen. Was impressive.

bobo5195 · 05-31-2006, 06:33 PM

TEC chiller setup is way to advanced get some anhydrous copper sulfate and be done with it (or of course silica gel). Drip plate won't really work as cards will get in the way and act as a very expensive drip plate.

Flourinert is as chemically unreactive as just about anything. The stuff your talking about is unstable enough that the blood stream can suck the oxygen out of it yet heavy enough a room temp. Plus free flourine in the lungs im guessing is not a good idea.

Radiator + fan won't work so well. You chuck air out the back of a case you lose heat as your losing 20 CFM air or whatever. Thats a case full of hot air very quick.

Best option in my option is use SLS to make the case and then put some FC77 or better (FC 40 or so) arpound to get rid of heat. In short all options lead to green backs.

Brians256 · 05-31-2006, 08:14 PM

Quote:

Originally Posted by bobo5195

TEC chiller setup is way to advanced get some anhydrous copper sulfate and be done with it (or of course silica gel). Drip plate won't really work as cards will get in the way and act as a very expensive drip plate.

Flourinert is as chemically unreactive as just about anything. The stuff your talking about is unstable enough that the blood stream can suck the oxygen out of it yet heavy enough a room temp. Plus free flourine in the lungs im guessing is not a good idea.

Radiator + fan won't work so well. You chuck air out the back of a case you lose heat as your losing 20 CFM air or whatever. Thats a case full of hot air very quick.

Best option in my option is use SLS to make the case and then put some FC77 or better (FC 40 or so) arpound to get rid of heat. In short all options lead to green backs.

With proper positioning, you really could put a TEC on a finned heatsink, but reverse the current when you want to melt the frost. You could put the heatsink down low so that the drips wouldn't go onto any card. But.... yes it's not easy or cheap. It just happens to be lower maintenance.

Rube goldberg for the win!

Talcite · 06-01-2006, 07:09 PM

anyone actually want to draw these out? 0.o how much is fc77 or 40? 3m needs you to call a sales rep for a quote >.>'

bobo5195 · 06-01-2006, 08:02 PM

There is still card --> air --> heat sink resistances. Cheaper just to buy a freezer and retro fit it. Condensation still might happen. Chemicals are going to be easier after a seal.

Personally i prefer the submerge approach as it seems easier (though FC77 adds to the price). should be able to get down to minus 50c easy.

the cheapest ive seen quoted for FC77 is $300 a gallon. Ive also seen little filler portions for about $80. Its going to be cheap for large orders but expensive for small ones.

Talcite · 06-02-2006, 12:14 PM

holy crap. $300 a gallon?

Is it a good thing to cool the entire mobo? Aren't there some parts that are sensitive to extreme temperatures?

Lol. retrofitting a freezer would be interesting... but yeah, there is a lot of condensation in a freezer, its just frozen =p.

Does anyone know of a formula or a chart to predict at what % humidity and what temperature, thing will start to condensate?

Brians256 · 06-02-2006, 01:44 PM

Quote:

Originally Posted by Talcite

holy crap. $300 a gallon?

That's still pretty accurate, last I checked. It's not cheap. Plus, it evaporates and you'll need to fill it up again over time.

Quote:

Originally Posted by Talcite

Is it a good thing to cool the entire mobo? Aren't there some parts that are sensitive to extreme temperatures?

Yes and no. It depends upon how low you take it. Also, if some bulk capacitors are leaky, you could run into stability issues as their capacitance decreases when the Galden/Flourinert leaks into them.

Quote:

Originally Posted by Talcite

Lol. retrofitting a freezer would be interesting... but yeah, there is a lot of condensation in a freezer, its just frozen =p.

This is why freezers have a defrost cycle and a drip pan for the liquified frost that had built up. Either you do it manually somehow (defrost manually or change out the dessicant) or you do it with some automated system (like refrigerators and freezers do).

Quote:

Originally Posted by Talcite

Does anyone know of a formula or a chart to predict at what % humidity and what temperature, thing will start to condensate?

Yes. This is standard HVAC stuff. Google for dewpoint charts. Condensation depends upon temperature, humidity and air pressure.

bobo5195 · 06-02-2006, 01:55 PM

Science word for dewpoint charts is psychrometric charts, i have a nice A3 in frount of me but can't give it out due to copyright. If you want to find out humidity and such you need wet and dry bulb thermometers. Google as they can be found

I have seen people run mobos in flourinert so it can be done. not heared of long term examples though. imagien FC77 is going to evapourate in a similar fashion to water.

Talcite · 06-08-2006, 10:38 PM

Anyways... I've decided i'm going to do it. Totally going for this thing, I'll machine it out of plexiglass and seal it. Then it can be a submerged/silica gel case. Really depends what my bank account is sitting at =p. will have proDesktop drawing in a few weeks. Need to wait for exams to finish up. I'll end up having like 8 items to CNC lol.

Brians256 · 06-09-2006, 01:30 AM

Good man! I love to see exotic things tried out.

Talcite · 06-11-2006, 11:34 PM

Hmm... I seem to have run into a problem. All of the dew point charts I've read are for above 0 celcius, and I don't know if i'm interpreting those psychrometric charts correctly, but none of them seem to be for below freezing either. Does a formula exist for a 30 deg ambient, and subzero contact surface dewpoint chart? And is it even realistic to expect subzero temperatures with a TEC?

Brians256 · 06-12-2006, 02:02 PM

What happens to water at 0C?

Talcite · 06-12-2006, 04:46 PM

oh lol... will frost have any effect on the electronics of the cpu or things like that? I actually can't remember if ice conducts electricity...

anyways, i drew out the rough sketch for the case today... It's going to have 2 seperate cooling loops with 2 seperate rads and pumps. Only the CPU and GPU will have TECs on them. one loop will be CPU and northbridge, other loop will be GPU, PSU, and internal radiator. Can you think of anything else that needs cooling? The internal rad is cooling for the HDDs. Maybe the mofsets? I don't know... give me some feedback guys =p

Shoebox9 · 07-23-2006, 10:24 AM

Not quite sure what you're proposing.

Yes, mofsets, RAM, etc will produce heat, but the FC77 & plexi-glass will take care of that. Lots of people have used open fish tanks with vegy oil (for silent computing) and the glass + oil surface seem to transfer a good 50w of heat away without any extra help. Plexi wouldn't be quite as good, but enough for m/b only heat. If you don't insulate the cold side of your TECs, this will cool the oil anyway.

Don't submerge your h/drives, and make sure you string your cables at least 1m above the tank. Less, and after 6m you'll have expensive oil dribbling out of your mouse and any other USB stuff.

It's all been done before- just use your head & follow the above tips! I'd use a sealed tank with external rads because the FC77 will evaporate slightly faster than water if you don't. :-(

PS. Most fans still work submerged in oil for a very long time (amazingly) so don't be afraid to add one for circulation if needed.

PPS Here's a link to some bored guy using FC40 & LN for a bit of fun. You could do better for long term use.

05-29-2006, 06:45 PM	#1
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	anyone ambitious? almost silent subzero case? I was thinking in the car today, and i was wondering if it was possible to create a TEC adapted or a LN2 system case. The idea is just to combat condensation, and perhaps a bit of noise as well. Why not build a case that encloses the motherboard, HDDs and the PSU? the rad and optical drives can be outside. Basically, the enclosed part can be airtight, and then put a hygroscopic material inside. Absorb all of the humidity, kinda like silica gel or something. The idea is if you remove enough moisture, there's no need to silicon your board, or dielectric grease the socket. The PSU would have to be watercooled, and the CPU, northbridge, HDDs, and vid cards as well. Mofset's too probably. This is a crazy idea, but it would mean that we could have almost silent cases, plus very fast systems. The input panel would have to be sealed with a rubber gasket or something. Rad hoses can go directly into the enclosure wall. Put a metal tube welded into the wall with barbs, just so the liquid can get through. Rad would have to be outside the enclosure, because it needs heat transfer. Optical drives too, because they need to open up. All the other cables like front panel, ide cables, power cables, could be sealed with silicon. PCI cards, rubber gaskets. Make one side of the enclosure openable, so the silica gel can be replaced periodically. am i crazy? or is it something someone wants to try? I'm still thinking about my waterblock, so i can't do this one.

05-29-2006, 07:12 PM	#2
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: anyone ambitious? almost silent subzero case? need to get heat out of the case as air is a good insulator and there are still things like mosfets which will heat the air. Make a close fitting case around it and fill it with FC77 by all means, that will work 2. Sealing will be a big problem

05-29-2006, 09:34 PM	#3
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Re: anyone ambitious? almost silent subzero case? Put a radiator inside the case as well with air flow through it, to absorb heat and transfer to the rad outside the case. __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

05-30-2006, 12:23 PM	#4
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? yeah. Sealing actually won't be that much of a problem. You don't need a complete seal. Only enough so the silica gel will be able to keep humidity low. However, a good seal will mean silica gel needs to be replaced less often. Hmm... rad inside and rad outside the case is a very interesting concept. What is FC77?

05-31-2006, 01:21 PM	#5
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: anyone ambitious? almost silent subzero case? everything leaks. expensive liquid and leaking does'nt go well flourinert (FC77 being the best for our purposes) http://solutions.3m.com/wps/portal/!...i_IcFQEAOIxDW4! Its water with flourine instead of hydrogen. Highly inert and looks and feels like water. could probably drink it and take a bath in it with some high voltage cabling.

05-31-2006, 06:33 PM	#7
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: anyone ambitious? almost silent subzero case? TEC chiller setup is way to advanced get some anhydrous copper sulfate and be done with it (or of course silica gel). Drip plate won't really work as cards will get in the way and act as a very expensive drip plate. Flourinert is as chemically unreactive as just about anything. The stuff your talking about is unstable enough that the blood stream can suck the oxygen out of it yet heavy enough a room temp. Plus free flourine in the lungs im guessing is not a good idea. Radiator + fan won't work so well. You chuck air out the back of a case you lose heat as your losing 20 CFM air or whatever. Thats a case full of hot air very quick. Best option in my option is use SLS to make the case and then put some FC77 or better (FC 40 or so) arpound to get rid of heat. In short all options lead to green backs.

06-01-2006, 07:09 PM	#9
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? anyone actually want to draw these out? 0.o how much is fc77 or 40? 3m needs you to call a sales rep for a quote >.>'

06-01-2006, 08:02 PM	#10
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: anyone ambitious? almost silent subzero case? There is still card --> air --> heat sink resistances. Cheaper just to buy a freezer and retro fit it. Condensation still might happen. Chemicals are going to be easier after a seal. Personally i prefer the submerge approach as it seems easier (though FC77 adds to the price). should be able to get down to minus 50c easy. the cheapest ive seen quoted for FC77 is $300 a gallon. Ive also seen little filler portions for about $80. Its going to be cheap for large orders but expensive for small ones.

06-02-2006, 12:14 PM	#11
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? holy crap. $300 a gallon? Is it a good thing to cool the entire mobo? Aren't there some parts that are sensitive to extreme temperatures? Lol. retrofitting a freezer would be interesting... but yeah, there is a lot of condensation in a freezer, its just frozen =p. Does anyone know of a formula or a chart to predict at what % humidity and what temperature, thing will start to condensate?

06-02-2006, 01:55 PM	#13
bobo5195 Cooling Savant Join Date: Aug 2005 Location: uk Posts: 400	Re: anyone ambitious? almost silent subzero case? Science word for dewpoint charts is psychrometric charts, i have a nice A3 in frount of me but can't give it out due to copyright. If you want to find out humidity and such you need wet and dry bulb thermometers. Google as they can be found I have seen people run mobos in flourinert so it can be done. not heared of long term examples though. imagien FC77 is going to evapourate in a similar fashion to water.

06-08-2006, 10:38 PM	#14
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? Anyways... I've decided i'm going to do it. Totally going for this thing, I'll machine it out of plexiglass and seal it. Then it can be a submerged/silica gel case. Really depends what my bank account is sitting at =p. will have proDesktop drawing in a few weeks. Need to wait for exams to finish up. I'll end up having like 8 items to CNC lol.

06-09-2006, 01:30 AM	#15
Brians256 Pro/Staff Join Date: Oct 2001 Location: Klamath Falls, OR Posts: 1,439	Re: anyone ambitious? almost silent subzero case? Good man! I love to see exotic things tried out.

06-11-2006, 11:34 PM	#16
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? Hmm... I seem to have run into a problem. All of the dew point charts I've read are for above 0 celcius, and I don't know if i'm interpreting those psychrometric charts correctly, but none of them seem to be for below freezing either. Does a formula exist for a 30 deg ambient, and subzero contact surface dewpoint chart? And is it even realistic to expect subzero temperatures with a TEC?

06-12-2006, 02:02 PM	#17
Brians256 Pro/Staff Join Date: Oct 2001 Location: Klamath Falls, OR Posts: 1,439	Re: anyone ambitious? almost silent subzero case? What happens to water at 0C?

06-12-2006, 04:46 PM	#18
Talcite Cooling Neophyte Join Date: May 2006 Location: Toronto Posts: 64	Re: anyone ambitious? almost silent subzero case? oh lol... will frost have any effect on the electronics of the cpu or things like that? I actually can't remember if ice conducts electricity... anyways, i drew out the rough sketch for the case today... It's going to have 2 seperate cooling loops with 2 seperate rads and pumps. Only the CPU and GPU will have TECs on them. one loop will be CPU and northbridge, other loop will be GPU, PSU, and internal radiator. Can you think of anything else that needs cooling? The internal rad is cooling for the HDDs. Maybe the mofsets? I don't know... give me some feedback guys =p

07-23-2006, 10:24 AM	#19
Shoebox9 Cooling Neophyte Join Date: Sep 2005 Location: Down Under Posts: 23	Re: anyone ambitious? almost silent subzero case? Not quite sure what you're proposing. Yes, mofsets, RAM, etc will produce heat, but the FC77 & plexi-glass will take care of that. Lots of people have used open fish tanks with vegy oil (for silent computing) and the glass + oil surface seem to transfer a good 50w of heat away without any extra help. Plexi wouldn't be quite as good, but enough for m/b only heat. If you don't insulate the cold side of your TECs, this will cool the oil anyway. Don't submerge your h/drives, and make sure you string your cables at least 1m above the tank. Less, and after 6m you'll have expensive oil dribbling out of your mouse and any other USB stuff. It's all been done before- just use your head & follow the above tips! I'd use a sealed tank with external rads because the FC77 will evaporate slightly faster than water if you don't. :-( PS. Most fans still work submerged in oil for a very long time (amazingly) so don't be afraid to add one for circulation if needed. PPS Here's a link to some bored guy using FC40 & LN for a bit of fun. You could do better for long term use.

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