Looking for examples of sealed custom cases to deal with condensation

[Liblinski]

Hello,

I want to be able stick a Mo-Ra outside during next winter and/or use plate heat exchangers with my tap water (5c) during the hottest days of summer. I am looking for successful examples of controlled atmosphere cases to deal with condensation.

As of now, I am thinking about drying the air of an air-tight case with a pair of low powered TEC coolers. I know I can't cool the PC with TECs, that is not what I am seeking. I just want to trap the moisture of the inside of the case on the cold side of the TECs and remove it. The volume of the case being pretty small, so the TECs could remain pretty low power. Fans would be circulating the air in the enclosed case and directing them to and through the TECs' cold side heat-sinks. Still pondering what would be the best way to funnel the moisture out when turning the case off...

If you have an example of an even marginally successful try at a controlled atmosphere case please tell me about it!

 

[Liblinski]

Or maybe I could also learn from some successful refrigerated cases?

 

[EarthDog]

Sounds like a nightmare for little returns... but cool as hell.

I don't have an example, sorry. I'm looking forward to seeing what you come up with!

You don't see too many reefer cases because they typically don't work well (units aren't meant to cool constantly).

 

[Aleslammer]

Using the heat exchanger idea probably wouldn't have a condensation problem as the hot side will never be as cold as the cold side. Playing with a chiller 5c most days not a problem and have seen a spread from 0c to 8c through the year where I live.

 

[Liblinski]

@Aleslammer, That's very interesting. I was under the impression that on the hottest days in my area (central Europe), maybe a week or two a year days total, would potentially be the most humid too. Especially indoors. Making the use of 5c-6c water unsafe to cool my PC.

@EarthDog, The build itself is pre-supposed to be rather effortless once running. Do you mean the temperatures I am going to be running run at (derived from -10; +7c ambient) are not worth any particular effort? Initial effort, including the present feasibility "study" and the efforts to plan it carefully, don't bother me. They're part of the fun. But if the end result is bound to be worthless or close to worthless...

 

[EarthDog]

Im saying that a fridge/freezer isn't meant for sustained loads like that and will burn out. The end result isn't 24/7 feasibility if that is what you are going for.

 

[Liblinski]

@EarthDog, as initially explained this is definitely NOT a freezer build. I am totally aware that these kind of build are based on misconceptions and fail to deliver.

The goal is NOT to have a controlled temperature inside the case. Rather, the goal is to achieve a controlled dew point temp and let the external sub-ambient water-cooling do its thing. Simply put, the goal is to just DRY the atmosphere case not to cool it. Drying the air until the dew point is so low (or that required air temps for condensation are so high) that condensation isn't an issue anymore. If with a TEC cold trap and a bit of silicate I can get my humidity rate at around 1-0.5 % and that the coldest point in the loop is 5C the case-ambient would have to climb to 95-115c to see any condensation. That's not happening.

But if you guys think using -10c to 7c coolant with no noise and, supposedly, no worries is not worth the hassle, then I have scrap this idea indeed. I never used a chiller or any kind sub-ambient cooling.

 

[Liblinski]

Ok. What I need to know and ask from you is now clearer to me. So let me reformulate and ask again.

1) How much is it worth to be able to run -10 to 5c coolant temperature for daily use? Most demanding task probably being Gaming and then 3d CAD design, then the occasional 3d render and then the occasional video editing (4k60,<20min vids). @Aleslammer How has it been for you to use that chiller at 5c in a daily use? Worth it over custom water-cooling?

2) Are there any particular failure points in freezer builds that are not linked to the lack of cooling power and its underlying misconception?

Thanks again!

 

[EarthDog]

1. Worth it is up to you. Running such a thing for 24/7 operation just doesn't seem worth it. That said, I have no idea what fpu this is going under. I think it would be more worthwhile with an Intel than AMD. The yields you'll get from increased clockspeed versus time/effort to setup and maintain, isnt worth it to me. On either side, I dont imagine but another couple hundred MHz more.

 

[Aleslammer]

Cheap portable AC unit, duct the intakes and the cold exhaust sides between the AC & case, hot air out the window, not going to say anything about noise and cost to run. Do have a AC to dismantle for parts if it seems to work to you're satisfaction and possible for a recirculating sealed unit or resale.

Found this if you can find a used one and see how it works, did see something about cooling zones in the case but nothing in depth. Might be a review out there some where.
https://www.techspot.com/news/45231-digital-storm-intros-sub-zero-cryo-tec-cooling-system.html

Saw pics of a computer in a fridge that seemed to work, but as I remember was a commercial unit from a bar and can't find the original posting and back in the S775 or S1366 days.

I don't run my daily on the chiller just use it to bench with, to touch a little on improvement over the chiller to a good water system at least on the new AMDs helps but nothing to write home about. Don't know about Intel haven't played with the new offerings yet.

 

[Liblinski]

Thanks for the comments Aleslammer.

I think I remember the same pic of a commercial fridge you mention. Too bad I can't seam to find it. Cheers man!

 

[Ben333]

Use the same sub-ambient water that you'll be feeding the CPU (& etc) and run it thru a small rad (80 or 120mm?) inside of the case. Place a fan on it ... drip tray below. Maybe do this with a second cheap-o pump. Then you can de-humidify the case before cooling down the waterblocks.

 

[Liblinski]

@Ben333

That seams indeed like the easiest way to do it. Great idea. Just gotta be careful about how much watts the components that are not in the loop will pump into the case.