Using TEC as a condensation-attractor in a sealed system???

The forum for non-component-related silent pc discussions.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Using TEC as a condensation-attractor in a sealed system???

Post by shoebox9 » Sat Sep 03, 2005 9:02 am

Hi All,

As previously mentioned, I'm out to build a fast/stable/totally silent rig. TEC's have great appeal, but control & condesnation seem to be huge issues, & I'm not about to try home-waterproofing my system, & play till it pops.

I think I have a new idea. (At least in days of searching I haven't been able to find someone else with a similar system).

As a condensation control measure, I'm considering building my system inside an old disconnected bar fridge (because it's air tight) and attaching a 350w TEC (or 2 if need be) to a large (ie 12" or more square?) sheet of copper to attract/freeze the condensation out of the air. A water block, opposite the TEC on the back of the copper sheet, would cool a water loop for CPU etc. The water loop/s would be passively cooled by a large external radiator.

The water loop would be: external radiator -> internal res -> split a) -> hot side of TEC -> external radiator. Split b) -> water block on back of copper sheet for further chilling of water -> CPU -> external radiator. NB, GPU & h/drives would be passively air cooled by the TEC/copper sheet cooled air.

Next option: as above, but split b) straight to CPU -> external radiator. In this option , & CPU temps would be higher as TEC is only cooling copper sheet.

Third option: high powered TEC for copper sheet, medium powered TEC cooling CPU water loop. Both TECs cooled by external radiator.

I figure inside an air sealed system (with an external cooling radiator) a limited amount of water vapor exists which will be "attracted" to condense on anything colder than dew point. I presume this will occur fastest on the coldest thing, thus reducing available water held by this packet of air to condense on other bits inside the sealed box.

When off, the system will slowly warm and evaporate the water back into the dry air, ready for the next round of condense-here-first, the next day.
I don't care if the water vapor ultimately freezes on the copper sheet, this keeps it away from the CPU water block, etc.

Am I mad, will this work? I will build this setup shortly if no one can think of reasons why it shouldn't succeed.

Cheers,
Shoebox9

Bob_the_lost
Posts: 136
Joined: Sun Aug 21, 2005 1:52 pm

Post by Bob_the_lost » Sat Sep 03, 2005 10:59 am

Ironically i looked at this a while back, it's doable, but it's a bitch, look down the page a bit and you'll see the "is there a place for peltiers in silent computing".

Basically you'll have to worry about the efficency and the condensation, good luck.

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Sat Sep 03, 2005 4:06 pm

I may be misunderstanding-but it sounds as if you are going to route the same water to the Pelt's hotside,AND coldside as well as the cpu,the copper heatsink and an external radiator. Not good. You do not want the hotside in the same loop-system. The water would be getting heated-cooled and the total result would be the backside heat cancelling most of the benefits. 2 seperate WC systems would work-as would a WC loop of only the heatsink/cold side+CPU+reservoir,

Another idea is to WC just the hotside-which ,if I understand right-would make the coldside colder.

The plate/heatsink should be finned so the cold transfers better to the air in the box. Ideally the chilled metal heatsink offsets the heat put out by the computer components-so the inside temp is under 100 F degrees. You'd want at least a low speed fan in the box for circulation. You probably also want small air inlet exhaust vents,an exhaust duct above the CPU so warmed air tends to rise up and out-taking some humidity with it. The inlet duct should open near the fan,which blows across your cold plate toward the CPU/chipset etc.

The basic concept is real good-odd-but good. I don't feel the watercool is needed,but you at least want a big heatsink on the outside to dissipate heat from the peltier.

frostedflakes
Posts: 1608
Joined: Tue Jan 04, 2005 4:02 pm
Location: United States

Post by frostedflakes » Sat Sep 03, 2005 8:43 pm

Also keep in mind that it will be rediculously difficult (impossible maybe?) to quietly power 700w worth of TECs.

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Sun Sep 04, 2005 12:49 am

I really don't want any air vents, this would undo the whole idea of a single packet of air being de-humidified by a cold plate. Adding warm, water-rich air, is not the goal. Also, I'm talking of large TEC's, but only at low power, so they are more efficient. I'm not strong on the maths, but understand it's better to use 2 350's at 5v than one at 12v, if your setup allows such options (ie in-line, not stacked).

So back to the concept-

I currently have a h/drive sitting in the fridge. The drive is cold, but so far I haven't been able to find any condensation on it. Yet there's tons of condensation on the cooling plate at the back of the fridge- some times icy, some times it just beads or runs down to a collection tray. If the h/drive was warm/hot, surely this would work even better?

I guess I'm thinking of turning my case into a fridge. But what happens when I power down? As the drives, PSUs, etc, cool, and the water re-evaporates into the air from the cold plate, will dew form anywhere nasty then? I don't know is if I need a large chunk of metal inside, to act as a cold storage buffer, or a really thin cold plate, so it will warm up at a similar rate to the components.

Any ideas?

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Sun Sep 04, 2005 10:58 pm

We used to get Potato Chips that had a little dehumidifier packet in the bag-I think it's called silica gel. Not long ago I got a container of the stuff,about the size of a 1 lb coffee can. It's marketed to people who dry flowers. A sealed up case is gonna get moisture in it,you're not starting with a vacuum. My vent idea uses the CPU heat to evaporate the moisture and duct it out, That silca gel stuff might help. If the system is closed you'd want a drip tube,letting the water that condenses drip through the floor of the case/fridge into a tray to evaporate there. Most refrigerators have such a setup.

This is a radical concept. Try it on an old cheapo puter before you put your real rig in it.

How do you expect to mount CD/DVD drives? External USB 2?

A thing about having 2 undervolted 350w pelts-it may be overkill. You want to chill the air,or more accurately offset the heat output of the computer parts in such fashion that inner temps are in a nice safe range,but overcooling will add to the condensation problem. You are going to probably want at least some sort of fan in there,having one end hot and one end cold could have bad results. It seems that if the air circulated,it's probably best. I'd also plan on having the PSU somewhat isolated from this sealed chamber. AC current + water is a bad mix,sometimes lethal.
This would seem a good application for a fanless/passive PSU mounted external-or in a seperate,well vented compartment.

This whole idea may create more problems than is solves-but a lot of good ideas seemed dubious until the details got tried out...good luck.

Maybe I'll ponder turning a big picnic cooler inti a computer case?

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Mon Sep 05, 2005 1:07 am

While Silca gel pellet thingys can also be placed in an oven to remove mositre, and reused, I don't think they offer a ongoing solution to PC mositure reduction. They are a great "use once" solution to lots of problems however, and are used extensively in nappies these days.

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Update

Post by shoebox9 » Mon Sep 05, 2005 1:17 am

MY PLAN UPDATED-

* Cool the CPU (& maybe NB, GPU?) using water blocks, a pump, & a large copper/brass car water radiator, with no fans, attached to a wall. I'll use 2 radiators if needed, but plan to start with one.
* Use an old bar fridge as a case. Advantages- it is air tight (I presume), offers reasonably good sound dampening, and is free. Disadvantages- it is big, and I'll need to cut off the external gas cooling pipes & lie it on it's back, so it will fit under my desk.
* Cool the h/drives, PSUs, maybe NB/GPU, plus any other hot things inside the "case" by actively lowering the air temp using a 350w TEC (as opposed to the fridge's compressor based cooling system). The hot side of the TEC will be added to the CPU water cooling loop- to be passively cooled externally.
* Use the fridge's thermostat & a relay to control the air temp, by switching the peltier on/off. Peltier to be powered by a 3-15v 40A variable switching PSU (already ordered).

Now, the tricky bits-

1) How to cool the air inside the fridge, in a way that removes condensation from the air. As previously stated, my current idea is to create a second water loop, this time from the cold side of the TEC, and use copper pipe soldered to a large, thin, sheet of copper. There would be a drip tray under this, with a hose exiting the case. The sheet of copper would act as a water vapour condenser, in addition to being an air cooler.

2) I guess I'll need a fan (away from the cold plate so it doesn't disrupt condensation) so that localised over heating doesn't occur during the system warm up/case cool down period. Maybe NB & GPU should be in the CPU water loop to take some of the strain out of this part of the cycle?

3) What will happen as the system re-warms, after shut down? Could circut boards that weren't really warm anyway, now attract condensation as the air temp warms up faster than they do? Opps, hadn't really thought about this...?
Would adding a cold "bank" (ie large block of metal) inside the case to slow air rewarm help, or just make things worse during the cool down stage with no real benifit?

Shoebox9

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Fri Sep 30, 2005 3:16 am

Very fascinating concept. While there are likely easier ways to go,there is a great deal that can be learned if this is pursued,but remember-it's very experimental. I wouldn't stuff $1000 worth of hardware in it and hope. I don't yet feel sure that you can really rely on a plan that assumes the same air stays in the box. If you collect condensation,water vapor becoming water to get caught in a drip pan-won't that water then evaporate-becoming water vapor again thus perpetuating something you'd like to end?

Stuff like CD/DVD drives,PCI slots,PSU are not made to be airtight so I doubt there is total control of air seeping in and out. Aso-if you open the case for whatever purpose- Air changed.

In practice I am only guessing at what conditions you will have inside. It seems there should be some sort of compact dehumidifier that has a way to get extra moisture-humidity out of this closed environment.

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Fri Sep 30, 2005 4:49 am

In essance, what I'm describing is putting an entire computer in a TEC powered fridge, with siliconed holes for 240v cable/monitor/USB/Firewire cables. Normal fridges are both noisy & underpowered for this job, alas.

Yes, opening the "fridge" would change the air, but draining away condensate in a hose that exited the "fridge", with a tap at the end, would mean the water could be manually removed, never to return to that sealed volume of air. Doing this is probably just a bonus, and not essential.

While I am quite confident a fridge like system would work, I'm embarking on a simplier entry to water cooling, & extra low noise.

The plan I've finally settled on, is to put the old bar fridge I have outside, and put my hard drives in it, connected with 1m SATA cables through the concrete block wall. I've tested the bar fridge with drives working inside it, and have a temp range from 13C to 18C (18C @ 30C ambient). I will eventually also add a slowed car rad fan to the condensor pipes at the back of the fridge to help it along on >30C days.

Inside my office, I'll simply water cool with passive outside-the-PC-case radiators, so no fans. The pump will be submerged in a bucket, on the water's way to the radiators.

Down the track however, I'd like to revisit puting everything inside a silent TEC powered fridge. For now, I need to get my A$5-6k rig (incl monitors) working silently, & reliably.

BTW, This older styled bar fridge (with thicker insulated walls then brand new units) muffles the noise of 5 loud drives suspended inside, MUCH better than I expected. Almost impossible to hear. I was shocked. It made my large 3/4" wood + special-sound-foam + extremely ducted air venting effort, sound like a total waste of time. This encourages me that a small TEC powered bar fridge could be a SPCR winner one day, for high powered PCs?!

Cheers,
Shoebox9

PS I plan to run the PC 24/7 & will turn off the fridge (& then later the PC) before opening the fridge to remove/change drives, when I eventually need to do this.

ronrem
Posts: 1066
Joined: Sun Jan 16, 2005 2:59 am
Location: Santa Cruz

Post by ronrem » Sat Oct 01, 2005 12:49 am

The fridge to house a bunch of HD's is pretty COOL,HD's could even get mounted in plastic bags w twist ties to avoid any moisture-if thats even an issue-with a pack of that silica gel stuff in the bag. Then the main puter-if not too much of a heatmaker could get Ninja cooled or Watercooled and you can have power and no noise......now if you could squeeze it into a mini-tower.....

Myth!
*Lifetime Patron*
Posts: 151
Joined: Sat Oct 01, 2005 3:30 am
Location: Beds, UK
Contact:

Post by Myth! » Sat Oct 01, 2005 2:15 pm

does this mean we can pop round your office and find a fridge full of HDD's outside? :lol: put a coupla beers in too, there's a good chap :wink:

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Sat Oct 01, 2005 5:30 pm

Hmm,

Outside my office is a carport with some shrubbery, so you could sit in the shade & drink the beer, largely out of site from the street.

Maybe I should put some Christmas pudding in there as well?!

Security has crossed my mind. I plan to add a lock, and some sitckers indicating seaman specimen collection dates. (Only the brave will then bust in & drink the beer!)

Cheers,
Shoebox9

cotdt
Posts: 295
Joined: Tue May 10, 2005 3:30 pm

Post by cotdt » Sat Oct 01, 2005 7:12 pm

why all this work just for condensation? a bit of neoprene around the CPU does the trick. besides, you're not running the 350W peltier at full, but only at 40W or so.

i'm pretty sure peltiers can be practical in silent systems. it shouldn't be that complicated. =)

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Mon Oct 03, 2005 3:41 am

Sadly, I'm not using the TEC yet- just putting the h/drives outside, & passively water cooling the rest.

The only noise makers left will be the pump- which will be submerged in a large jar of water, & the PSU, which I'll drop in a tiny fish tank full of canola oil. So, total, utter silence (I presume).

One day I'll get back to the TEC's, but for now I just want to start with something simple.

Cheers,
Shoebox9

Myth!
*Lifetime Patron*
Posts: 151
Joined: Sat Oct 01, 2005 3:30 am
Location: Beds, UK
Contact:

Post by Myth! » Mon Oct 03, 2005 3:53 am

I wonder if you submerged a whole pc without fans in distilled water would it work ok and have enough....umm whats that word....where water rises because its hot and colder stuff replaces it....convection currents? well I wonder if it would stay cool. you would need to make sure there were no impurities getting in the water....so perhaps a kind of filter somewhere. Brita? lol
Last edited by Myth! on Mon Oct 03, 2005 12:11 pm, edited 1 time in total.

shoebox9
Posts: 76
Joined: Mon Jan 31, 2005 3:50 am
Location: Australia
Contact:

Post by shoebox9 » Mon Oct 03, 2005 5:11 am

I'm very attracted to the idea of submerging whole PCs (minus h/drives).

In this case however, there's sort of no need. I'll already have total silence. If I hadn't already bought a heap of water cooling stuff I'd be more tempted this time around, but as I have, I'd be in trouble with my wife if I suddenly don't need it all anymore...

Maybe next time!

Cheers,
Shoebox9

PS Distilled water is not a good idea, as lots of stuff will dissolve into it fairly quickly, making it conductive. There are many types of oils that work however- mineral oils, canola & sunflower oils, etc. The trick seems to either be using a vegie oil, or a mineral oil that's safe to breath the vapours of, in the long term.
Last edited by shoebox9 on Mon Oct 03, 2005 4:18 pm, edited 1 time in total.

Myth!
*Lifetime Patron*
Posts: 151
Joined: Sat Oct 01, 2005 3:30 am
Location: Beds, UK
Contact:

Post by Myth! » Mon Oct 03, 2005 12:21 pm

i had to look up canola oil....rape seed oil i spose :-)

yes impurities are the problem.... though it would be such a messy job changing parts if submerged in oil....and then you'd have trouble with the electrical contacts I guess. :roll:

Straker
Posts: 657
Joined: Fri Jul 23, 2004 11:10 pm
Location: AB, Canada
Contact:

Post by Straker » Sun Oct 16, 2005 4:28 pm

vegetable oils will go rancid, you don't want that.


not sure what the point of all this extra work is, but if you're willing to put holes in exterior walls, you could just bury a bunch of copper pipe (or a big metal tank) underground, no need for fans/radiators etc.

Post Reply