Outside-air cooled PC

[Brian] came up with an interesting PC cooling setup. He lives up north where it’s chilly in the winter. Using a bit of dryer ductwork and he was able to harness the outside air to cool his box. The system uses a window insert along with a dryer hood to suck in the outside air with a PC fan. We hope the air is adequately warmed, as it is exhausted into the room. Join us after the break for more pictures of his setup.

Below are the duct fitting he used. This includes a hood for outside, a damper which closes with gravity, some scotch brite pad to act as a filter, a flexible duct, and a quick connect.

The next three images show the construction of the window insert. The filter and hood are on the outside of the board, and the quick connect on the inside. Although [Brian] shows a damper in his parts image, we’re not sure where that is actually installed. We think it would be best for this damper to have a servo actuator, as gravity won’t be enough to keep the wind from forcing it open.

113 thoughts on “Outside-air cooled PC

  1. “No, heating air will not cause condensation. Last I checked hair driers don’t spew water everywhere.”

    For some reason that comment made me laugh out loud… I got a picture of that in my head, I’m so easily entertained….

  2. 43 degrees farenheit how oh how can you possibly survive!?!?! Ok, where I live, its getting to -27C tonight (that would be -16.6F to you), and yes, when its cold, the air is very very dry. I put humidity into the house every day, but still have to touch the edge of the (metal) desk or (metal) bed frame before touching the computer. Every arc is usually about 1/2 inch (about 60kv). Repeat after me: Freeze dried! Air with moisture outside, upon meeting a metal pole, or a chain link fence freezes to it. Outside air is dry, inside air is dry. I use a thermalright cooler (Corei7-920) and keep the box next to an outside wall. Cool!

  3. Hm… Just get a slightly longer monitor, mouse, and keyboard cable. Then, put the entire box outside, sans case. Keep monitor, mouse, ‘board, and perhaps some USB extender cables inside. Problem solved!

  4. @Bill,

    Ok, I get the distinct feeling I’m being trolled, but here goes.

    Fact number 1:
    You cannot get condensation on a surface if it is not below the temperature of the incoming air.
    Fact number 2:
    You cannot cool something down using forced convection to a temperature colder than the air/fluid you are using for convection.
    Fact number 3:
    100% humidity is a saturation point, you cannot go higher than that.

    These are laws of physics.

    You state that somehow if you suddenly get an increase in outside air humidity to 100% from some arbitrary value you will get condensation. This cannot happen, since even if you are sucking in air at 100% humidity this is only being heated up, both in the ducting and then in the PC, lowering the dew point. The rate at which the humidity increases is irrelevant. The rate the temperature is changing IS relevant as Swighton mentioned, which is why I brought it up. The only way condensation could possibly happen is if the thermal mass in the PC causes the exposed component temperatures to lag behind in a sudden INCREASE in outside temperature (i.e. the PC is somehow suddenly colder than the incoming air), and this happens at a 100% humidity. This is highly unlikely since there are no really heavy components with a high thermal mass, except for the heat sink on the CPU which is several tens of degrees above temperatures which will see condensation anyway.

    I hope that clears it up for you. BTW I taught thermodynamics at a graduate level. feel free to bring up another argument, I’d love to see your response.

    I looked at your website, and I have a lot of respect for what you do, it just bothers me when people criticize someone’s elses ideas without grounds. I don’t have problems with constructive criticism, as long as its based on truth.

    1. I don’t have any thermodynamics background, Just 1 day of basics on humidity in an HVAC apprenticeship.

      To me the temperature and quantity of incoming air and the temperature and quantity of air in the PC and room would matter.

      Outside air is normally less humid than indoor air (Northern Canada) so the concern is that you cool the indoor air to it’s dewpoint.

      Think of rain clouds. A cold front meets a warm front. I’m guessing that only if there is enough cold to overcome the warm to cool it to a certain point it will condense and rain.

      In this setup there is both an abundance of cool air (outside) and an abundance of warm humid air (inside) so there will be some locations where the cool air is lowering the temperature of warm air to its dew point.

      So I would conclude that you would need to really insulate your PC very well ensure room air doesn’t leak into the case. As to not allow warm humid room air to condense on cool pc components or to loose heat to cool air inside the PC. Also you would need to monitor and mix cool air coming into the PC carefully so its not lowered to it’s dew point.

      Say the air in your case is at 5*C and your room temp is 20*C at 65% humidity. When it touches pc components at 5*C it loses heat and ends up at ~12*C it’s at dewpoint. This is why you turn you the humidity in a house down in the winter so that condensation doesn’t form on windows or where there in air infiltrating from outside.

  5. No trolling, and you are correct about the science but aren’t thinking about the possibilities.

    Read through the comments again. Now even Swighton admits it could happen. It’s not in the steady state, but by a transition in weather outside.

    Surfaces in the PC in direct path of the on-coming cool air that aren’t being heated by a component of the PC will cool to the same temperature of the outside air. A sudden change, (like it starts raining) would push humidity to 100% and could raise the dew point above what the the previous air temperature was (and the surface of the computer). Outside, this causes fog. In this computer on the cool surfaces, water condensation.

    You may have taught thermodynamics but I have worked in dimmer rooms in theaters that were outdoor air cooled and suffered this exact problem. The grills and walls the air passed by before hitting the hot dimmers would condensate every once in a while. That and the erosion by salt air and the designers regretted their decision for open air cooling.

    1. I believe you are right. You would need to carefully mix cool incoming air into the air already in the PC case>

      But especially have to be carefully of warm indoor air (20*C, 65% relative humidity) infiltrating your case and losing heat to cool components or cool air inside it.

      EI humidity on inside of the windows of your house when its cold out.

  6. Assuming the case is vented and will intake some room air… it’s not the cold air entering the case that causes condensation but the warm humid air circulating in your home that condenses against the cooler internals once the they’ve cooled below the room temperature.

  7. At college, due to negative pressure in the hallways, cold air is naturally pulled into the room. I just put my laptop “near” the window, and the temp drops significantly. No condensation issues AFAIK

  8. It should be illegal to run a clothes dryer vent to the outside in the winter. A dryer is as much as a quarter of the Btu’s of a furnace. Much more than a computer of course, but there is double indemnity. Air has to be made up. Would you want to run that little fan in a window blowing expensive heated air, outside? Of course not, but that is what is going on. Same with range hoods. The best example of bad design is a refrigerator, it has electric HEATERS inside it and is in a heated space in the winter! Many computer cases I have seen are full of the same bad design, holes everywhere including right next to fans, fans mounted against tiny holes, and intake above exhaust front and back. Hewlett Packard’s seem to do it right. One fan ducts into the CPU, coolest air right where it is needed most. Power supply exhausts above, very quiet. If you pull air off of the floor for intake with the duct and move enough it will work with out the waste of energy. Move the duct near the a/c vent in summer. With the window open there is a hard to seal gap at the top of the lower window and a serious break-in opportunity.

  9. @echodelta – There is a lot of poor efficiency heat transfer around the house and office all the time. You can’t always make up the difference.

    Example: I use a system like this to bring air into the small server room off my office in winter. I’d far rather duct the heat from that room upstairs to the bedroom or even out to the garage, but the volume of air I’d have to move through my office and into the server room to make the difference is too large to keep up with reasonably.

    In summer, I’d love to run the water as it goes through the pool pump across the condenser cores of my fridge, freezer, and window air conditions prior to going to the pool heater. That would far more effectively take the BTU’s out of those devices and put them into the pool water. Instead, the fridge & freezer heat the house that the air conditioner is trying to cool – and all the btu’s from all three get dumped by the AC unit to the outside.

    Unfortunately, for the time being, there’s no efficient way to manage the problem reasonably so we just have to GROW THE HELL UP AND GET ON WITH LIFE. Thank you.

  10. I think air moisture is the enemy… I realize it may not be a big problem but it’s certainly a factor outside the users control. I’m gone for a day leaving my computer on and it rains at noon. As that water evaporates it’s going to be pulled directly into that case.

  11. @echodelta said: “It should be illegal to run a clothes dryer vent to the outside in the winter. ”

    I’m picturing a new HBO series now…

    A new arrival at the maximum security Federal prison is being shown ropes by an old timer.

    “Over there, that’s Knuckles. He butchered his whole family with paring knife, but he’s OK for a game of cards. That guy behind him is Louie… Louie got popped after robbing his 100th bank…he’s the guy you go to for cigarettes.”

    “What about that dude over there?” The newbie asks, pointing at a somber-looking bloke with the shaved head and tats on his face, sitting by himself at a picnic table.

    “Oh, that’s Ted. You don’t *ever* want to mess with Ted.”

    “Why? What’s he in here for?”

    “He…” the oldtimer gulps, trying to steady his nevers. “He… vented his dryer to the outside.”

  12. I ran a setup like this in college. One exhaust fans, one intake fan sucking from a 14″ pipe to the window. It worked great. I was able to run all the fans slower and switch my gpu to passive cooling – while keeping it OC’d.

    Left it up all winter, even in the snow (I had a filter over the opening, so no snow went /in/ the pipe), it ran fine. No condensation at all.

    The air warmed from 20* to 60*, so cool but not cold. We had enough computers in the dorm room that it was always too warm anyways. Maintenance staff did come by to see wtf, but they were cool with it once they realized what it was.

  13. Yes he should make an inlet and outlet. Making the outside duct openings 16″ apart or so should suffice to cut down on short circuiting the heat exchange efforts.

    Bring cold air in to the case with one fan, dump it back out with a different fan, no interaction with the interior of the house.

    Also there is a chance he might be able to cut down on condensation, the naturally dry air from outside should dehumidify the interior of the case if it can be made a closed system…


  14. @JFS @Bill Porter

    How about we agree that under certain transient weather conditions, coupled with a mechanical configuration that has components of relatively high thermal mass (perhaps a hemebrew case with lots of machined stainless steel components which would hold heat very well due to their low thermal conductivity), that it would be POSSIBLE to form condensation.

    Trying to definitively say whether condensation WOULD form in the computer in question (since we know it is POSSIBLE, just not if it WOULD) is just guesswork without further details about the internal components of the PC in question.

  15. “It should be illegal to run a clothes dryer vent to the outside in the winter. A dryer is as much as a quarter of the Btu’s of a furnace.”
    You’re forgetting a key thing here (which many people here are touching upon) – Dryer exhaust is VERY humid, often so much that it will condense if it is routed indoors.

    However, I am surprised we haven’t seen more dryers with heat recovery systems (heat exchanger that heats intake air using exhaust air).

    As to the original hack: IMO no risk of condensation except in transient weather or extremely heavy rain. However, unless the PC is having problems when cooled using room ambient air (indicating the PC needs a cooling system redesign since it’ll fail when the weather gets warm), the PC’s cooling air intake should be taken from the room.

    If you take the PC’s cooling air from outside, it’ll heat up the air a bit, but the air exiting the PC will potentially still be colder than ambient. It will definitely be colder than it would be if the PC took room air from inside. This approach will make your room COLDER than if the PC were cooled with room-ambient air.

    HOWEVER: If you can solve humidity issues (Maybe a liquid heat exchanger?), in the summer if you take cooling air from outside and exhaust it back outside, you can eliminate heatload on your AC system. (If you exhaust air only, the AC system will have to cool the replacement air within your house/apartment.)

  16. I would stick a layer of dessicant on the floor of the case (assuming you have room), and have the outside airflow physically separated.

    Make a heat exchanger from two big heatsinks (you can buy them at JayCar or Radioshack or what ever) sandwiched together with minimal thermal paste between with ducted air from outside passing across one side and extracted air from the case in the opposite direction across the other side (and then returned to case).

    It should stop the issue of massive changes in case humidity due to external changes.

    Time to break out my Thermo/Fluid dynamics textbooks :p

  17. I have done this exact same thing except the dryer tube gos to the back of the computer where the air is sucked over a corsair h50 radiator. that computer is running folding@home all the time in the winter, and it has kept it running much cooler.
    It has run like that since the start of the winter and there is no sign of any dampness or corrosion anywhere.

    In the summer i will just turn the fans around on the radiator, so it blows the hot air outside.
    also the “outside” part of the tube exits under the roof of the porch, so no rain can get in.

  18. I did this like 8 years ago and it worked perfect. I hooked up a case fan inside the duct and there was no condensation. I used this system for about 3 years before I upgraded. The only downfall was that winter only lasts for about 5 months, so the rest of the time it was hooked up would be worst off in the summer

  19. Does this person live in an environment devoid of wild animals or weather? Or a dry, dusty summer month? I can see this going only one way in the long-run.


  20. (but strictly-speaking only because I know first-hand just how much damage can happen to an appliance once an animal nests in an exhaust airway — I don’t think it’s worth the cost of materials in the long-run compared to spending more on newer tech that generates less heat to boot.)

  21. This concept is becoming very popular with major datacenters, especially cooler climates. My former employer just built a massive datacenter that uses filtered (but not dehumidified – that one has been beaten to death here) outside air for cooling 6-8 months out of the year. They’ve won several awards for the design and are saving several million dollars a year on electricity.


  22. Over all, great thinking and good idea, and a nice clean setup.

    *But* you have great danger of things getting wet. a better idea would be to run the duct over the heatsink and then out, sort of like the old dells were set up

  23. Hard drives don’t like cold air and/or rapid temperature changes.
    Heavy condensation will form at the outer side of his box every time the temperature outside is lower than temperature inside the house. This might not be a problem if the case is plastic, but, it usually is not…

  24. Jesus you people are stupid and a bunch of fucking haters.

    I have been doing this for 10 years w/o a drop of condensation IN FU*KING SEATTLE!

    B*tch, whine and moan…. I’ll enjoy my 40f temps.

  25. Outside air in the winter is rarely humid, relative to inside air. In the coldest weather, its typically very dry. What he needs to worry about is condensation from moisture in indoor air. This would be easy to manage by making both an air intake and an exhaust on his PC and then running that exhaust to the intake of a HRV or heat recovery ventilator, where it is then used to ventilate his house. Basically, he can use his computer as an inline duct warmer. That way he saves money on his heating bill.

    An HRV is a great thing to have because it lets you have warmed, fresh outdoor air year round. They are mandatory on new construction in Canada because modern construction usues a lot of toxic products made with formaldehyde and other nasties. (Thats even more prevalent in the US where HRVs are almost unknown)

    HRVs warm incoming air with air being exhausted from your home. Throwing a computer into the mix makes total sense – it makes a good intake duct warmer.

    Using a computer as a intake duct warmer makes sense. (As long as its clean.)

    He should put a basic washable air filter on the air intake.

    Even on the coldest days the fresh air coming into my home, warmed by my HRV is in the low 60s or upper 50s. And its fresh air. Since my computer is situated not too far from my intake ducting I’m going to try the duct warmer idea.. it sounds very “cool”.

  26. Up here in wisconsin the condensation resulting from direct outside air could kill your computer rapidly. . .

    can’t say i have not thought of it though. As is my office is in a different part of my house w/ seperate heating. and I keep it quite chilly in the winter.

  27. @Bill – 43 degrees Fahrenheit isn’t COLD! Try -43 F without the windchill(so cold the LP gas wouldn’t vaporize as the boiling point of LP is -40 F). Then you can talk about cold. When it gets below 10, the humidity is on the ground in frozen crystals…

    @BoboftheNorth – Kudos from a Minnesotan!

  28. I tried this for about 3 months last winter. My initial concerned was condensation/humidity but upon the end of the project, my concern was how fucking cold my room got when the computer was off :).

    For those concerned with condensation/humidity I tried to examine that as much as possible. I never noticed any liquid or corrosion within my computer. I kept a hygrometer inside the case to compare it with the rest of my room and the outdoors. As expected, it was right in the middle of both most of the time. I’d be glad to post a couple pictures of my setup if anyone is interested, but I assure you, it was much more ghetto than this.

    Also, Im still using the same computer.

  29. Outside, the intake should be facing down using a doughnut loop pipe or double sink drain elbow to prevent precipitation from entering the pipe…..just an idea.

    for 4 season climates, reverse the flow in the summer to remove heat from the case to outside the home. This will lower your home cooling cost

  30. I’m a HVAC apprentice and I live up in northern Alberta, Canada. 1000km’s straight north of the Montana border. Its so cold up here LCD displays in your car will freeze.I’ve seen -53 celcius.

    I’m going to design something to do this. I like the idea of exhausting hot air in the summer too!

  31. re: all the comments about moisture. As long as you are only pulling air in during the winter months – well, what moisture? Once the dewpoint drops below 40, iced drinks don’t sweat anymore. Only potential complications are indoor humidifier use or other moisture-adding environment (boiling water, etc). And even then it’s only likely to build up on metal parts until they return to room temp. And since most of the metal parts are generating heat, even that’s not likely. Main danger of condensation would be in the change season, and almost any hvac scheme struggles in spring/autumn.

    I just opened my window and my system’s GPU temp went from 53 to 43. The air temp is approx 27F (-3 or -4C?) and the dewpoint is 22F (aka bone dry, not in the least worried about condensation).

    PS – always look at dewpoint and not relative humidity when concerned about condensation. 40 – 50 iced drinks start to sweat, 50 to 60 you need a coaster and you start to feel sticky, 60 – 65 is uncomfy, 65 – 70 you’re wringing out your shirt, 70+ you may want to talk to your local mad scientist for a set of gills. ;)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s