If you've ever worked or slept someplace that was air conditioned, you've already survived a dehumidifier encounter. Ain't nothing wrong with dry air; being in the desert is nothing like a bad hangover.

Turning off the heating is the worst thing you can do in a computer room? Huh. The upper floor of my house is heated solely by computers... :shrug:

You either have them on and the heat off... or heating on and them off. It is REALLY bad when you turn them all off go away for a few days and come back and power up directly... this is what my old school did and blew quite a few fuses and a couple of computers, because it was chilly and damp... stupid f***ers.

Maybe not... but some people are more suseptable to humidity(or absence of) than others. I can't stand dry air myself, still a point worth raising though. (even if it was a bit ranty - apologies to anyone offended.)

~ Boli

Interesting topic, and while I'd admit I don't have more than a passing understanding of all the physics that is involved in the formation of condensation and other atmospheric related things, I understand what I need to do to prevent it forming on my cold blocks and tubes.

A factor not to forget is where I live, (UK), and it rarely gets to 30C let alone above that. Mostly its 20C and 75% of the year that is artificially raised by house heating. Central heating tends to dry out the air a bit raising the due point as well. The chart shows the dew point at the humidity % at a certain temp and like I said it seems accurate to my personal findings using temp and humidity gauges. 100% humidity is very unlikely unless you are in a tropical rain forest although parts of the us can get near 100% I hear? Also rain has little to do with it as It's often high 30% low 40% when its rain in the summer here.

Like said there are so many specific variables to my plan the only way to know if this will work, and how well is to build it, just can't be worked out on paper or via equations at least not to my mind anyway.

I'm the opposite I hate "wet" air and find it uncomfortable when it gets above 60% humidity. We are all different as a girlfriend years ago continually complained of drying out on a trip to the US. The air-conditioning in the malls and cars, (we were in Florida), affected her badly where as I liked it full on ice cold and it didn't appear to dehydrate me, it was an excellent excuse for us to argue and fight, not that she really needed a reason.... :rolleyes:

What about isolating the case, and having an interior radiator cooling the inside air, so the delta between the air and the blocks is low?

Ora...
 

Er.

( Estás a apanhar do ar :D.

O Balde (pun) quer um radiador que arrefece o ar a entrar e reduz a humidade relativa porque algum do vapor precipita no radiador (fresco em função da temperatura ambiente). A caixa é isolada, e apenas tem uma saida de ar, a chaminé aquecida que provoca uma subida de ar quente (acima de ambiente) , e cria correntes de convecção que arrastam o ar mais fresco embaixo (que por sua vez já passou pelo radiador e arrefeceu a temperaturas abaixo de ambiente). Menos humidade, menos temperatura, menos propenso a condensações nas tubagens principalmente.

O que estás a querer dizer (que é diferente) é isolar totalmente a caixa do ambiente. E arrefecer o interior. Mas isso não é lá muito prático , especialmente que a quantidade de vapor de água ainda se mantinha no seu interior E para usares um radiador para tal tens de ter movimento de ar, quer seja por ventoinhas quer por convecção. E como usar ventoinhas está fora de questão , fica por criar correntes de convecção que não existem de maneira decente nesse tipo de montagem fechada. Tem de ser aberta. É para isso que se tem a chaminé, forçar movimento de ar com o frio/quente , sem auxilio de ventoinhas.

E se se conseguisse isolar a caixa e ser estanque, podia-se criar um vácuo (e sem ar não há vapor em suspensão, nem condensação). Mas ai é ainda mais chato, especialmente se se quer aceder ao hardware.

A teoria é porreira (e deveria funcionar), na prática nem sei se funciona correctamente. ;) )

Santanicoo

Because that would require a fan, or some way to move the air around, my system has to be totally fanless..... zero fan zone ;) Another possibility is to expand your idea to make the complete case a water box, similar to a central heating boiler. it could then be virtually sealed up and the air inside should remain almost as cool as the blocks / coolant temp...



TerraMex

Can you translate that last reply please :confused:

Já tinha pensado nisto uma série de vezes, talvez um dia me "passe" e trate do assunto ;)

Ok ok fanless :) . Just an idea.
The main problem i can think off is the chimney. The "thing that heats the air" would also heat the room, depending on the power of it.

I have one of these old fountains that I used to chill my old Celeron 300a and peltier setup. It worked really good, but that probably was less than 100w of heat to deal with.

It actually could get the cpu down to almost freezing, and the water temps were below freezing. I even had to add some automotive antifreeze to keep the water flowing as it would freeze plain water in the water lines. This was probably because they were only 1/4" ID, copper, and I didn't have a very good pump. I bypassed the temperature control switch to make the compressor run all the time.

I didn't get the exterior case, just the guts, so it looks rather ugly. I still have it in storage in case I ever feel like using it again. Pretty damn heavy and awkward. I would need to fabricate an exterior case/covering for it.

http://www.wsu.edu:8080/~i6735189/cp...rfountain6.jpg

It's just some of the things we've been discussing, but in portuguese , for better understanding by satanicoo.

It's not "something new". :)

I am in a similar situation as BladeRunner. My water is chilled to 10-15C by copper pipe in the ground. I get condensation on my blocks if its very humid and I then cook something like soup that adds a lot of moisture to the air (my kitchen/living room is one large room and my computer is there). When this occurs I've been avoiding condensation by opening some windows to vent the steam and cool the room down.
This chimney idea is very interesting to me. Its now one of the solutions that I may try out. I also like the idea of cooling a sealed case with a radiator and fan.
I have been considering raising the temp of my water to ambient by using a valve that would recirculate a portion of the water back through my blocks before it goes through my underground radiator. But its hard to give up the cold water.

Edit: After reading about surlyjoe's project at AOA Forums I'm going with the radiator and fan idea. He put a heater core in front of his intake fan in the lower front of the case. His chilled water passes through that first on the way to the blocks. But I'm going to put the rad after my gpu block for slightly lower block temps and easier plumbing for my particular set up.

fair enough. :)

Hmmm, I didn't make myself very clear on that one. I should have said that at the dew point, the air immediately adjacent to the condensor would be saturated, assuming ambient temp and relative humidity as stated in the chart. In essence there is a mini-convection flow, where the air next to the condensor cools to the dew point where it reaches saturation, moisture condenses out and the air flows back into the ambient space, where it warms back up....

Sort of, but it's not a question of Delta T, but rather that cold air at a low relative humidity(RH) has a lower dew point.

Correct, but not what I said... Assume a box full of dry air (0% RH) Evaporate water into it until saturation is reached. Measure the amount (mass) of water used. The amount of water needed to reach saturation will be a function of temperature, at lower temps less water is needed, at higher temp more water is.

Assume the box has been saturated (100% RH) at 20*C. If the temp is raised to 30*C, (and no more water is added) the relative humidity in the box would be less. If the temp is dropped to 10*C you will get condensation, and the air would still be at 100% RH.

Agreed completely, except that I'm not sure the rad would do well enough.

{Cloud physics discussion clipped, IMHO we were both saying the same thing, I just wasn't getting as deep into the details...)

Well, cooling without condensation will RAISE the RH, and bring the dewpoint closer to the temperature of the cooled air. Cooling with condensation will lower the amount of water vapour in the cooled air, but that air will still have a high RH (because of it's low temperature) as long as it isn't warmed back up.

From one of BR's earlier posts in this thread (highlight added)
Thus, 10*C @blocks, 12*C at rad assuming conservative values for his numbers. (not that it makes a big difference)

My assumption was also that there would be condensation, but that the air coming out of the radiator would be cooled and still have a high RH because of it's low temperature, which would NOT be as low as the cold side of the pelts. Thus the CPU area could still be below the dew point which would not have dropped significantly.

In my proposed modification; which locates the rad > airbox > chimney assembly below the computer box, and supplying the airflow to it; I am assuming that the air would be cooled and condensed by it's passage through the rad, just as you are. I am also assuming some drain or equivalent to get rid of the condensate. But I am suggesting that the air should then pass through the chimney so that it is warmed back up to ambient (or slightly above, but not grossly so - this would need some design tweaking) thus greatly lowering it's RH, and therefore the dewpoint.

This is exactly the same process that a regular de-humidifier uses, except that it eliminates the fans and other motorized bits per the Zero-Fan-Zone spec.

I would be tempted to draw more ASCII art to show what I mean, but the new rev doesn't preserve it properly.

Gooserider

Hi Guys

Just a quick question to stir things up. What is the Tempeature inside the case now? Is it above ambient? If so the rest of the components are heating the air as it passes through the case. Would this not lower the RH and the Dew Point?

Grayson

At the moment the PC is open and there is some warm up of the air in the pc. This is due mainly to no fans to move the air around, so it stagnates in areas. Natural convection will avoid this getting excessive, but convection requires a differential to work.

With artificial convection currents caused by the chimney effect this slow warm up probably wouldn't occur.

Sounds good but I really don't think atmospherics are that simple. If that were the case the only way the air could release moisture when it was cooled would be to condensate on everything around it, and that just doesn't happen. At times in the summer here in the uk, (and I use the term summer loosely), we can get big swings between relative humidity and temperature. almost always the humidity % goes down significantly when the temp drops.

Another point to clear up any Tec(s) used will have, and require the cold side to be insulated / isolated. This proposed solution is to prevent condensation on the block surfaces & tubes.

But.
 

Goose, your going well but ... you must drop the "saturated" stuff.
Doesn't exist when talking about atmospheric events. It's a wrong concept.
Confined gases/liquids can be/get saturated. The atmosphere (and including oceans) doesn't, it looses significance when talking about a open system of that magnitude. That's why it's a "relative" humidity.
It all about energy balance and transfer in the natural condensation/evaporation events. It's a constant process.

To give an example , sulphur expelled from a volcano as a vapor sublimates to solid state. The atmosphere wasn't saturated. The super heated vapor just lost energy to the surroundings and the energy balance was altered abruptly to the point it no longer had energy to maintain it's present state. Same goes to water vapor in contact with different surfaces at different temperatures , with different air temperature :) .

To give another example , in the outskirts of the Sahara Desert, you can have up to 100% RH , and no condensation whatsoever during the day. But in theory, it would be saturated and condensation would be forced (which does not happen). But as night falls, things literally freeze over. Lots of condensation occur (and ice) because of the balance shifted with temperature to the other way, surfaces dropped temperature considerably and energy is removed from the water vapour.

You can have just about any temperature / RH combination. And not have huge amounts of condensation or evaporation (uneven balance).
What changes (and we are discussing) is the Dew Point in those combinations when in contact with a temperature differencial (surfaces or air fronts).
I know it's a tricky subject.

I dont know how to make it more clear, language differences isnt helping either. Maybe it's my fault (probably). :)

A Dehumidifier re-heates the air to minimize temperature diferentials between the ambient air because it's not his job to maintain cool air. We just want less RH. :)

Anoher thing you're missing.
The interior of the case is not the main issue. The blocks and the tubes are because they are well below ambient, a few inches can have a delta T of several degrees.
Bladerunner assumes , if he can get the new ground rad, 8 ºC intake, even cooling pelts , it can go up to 12 ºC ... maybe 15 ºC on worse case scenario , not more that that I think. Equal air temperature of air passing (lets say those 15 ºC) and tubes will yeld no condensation because the delta T is minimum , even with high RH (if the radiator forces no condensation). Which if goes according to plan, will. Cool air, less RH, lot less risk. Of course, if things are tweaked properly. May assumptions. You dont need to heat up the air. I dont even think you should. It can shift the balance towards condensation if the delta T is high enough, given the cool surfaces of the blocks/tubes.

If , however, no insulation is used (i strongly advise against it), temps will depend on the cold side of the pelts. Things get even more tricky there. You'll have to know how low can the pelt go when crunching numbers , playing a game, etc. Some degree of control will be necessary, IMO.

You head him BladeRunner , make it happen :D.

TerraMex, I think we agree on the physics of what is going on, (at least I didn't see anything to argue about in what you wrote) but are approaching the problem from different directions. You are doing a threoretical physics approach, while I am taking a simple language approach.

For instance, I agree that 100% saturation doesn't occur in a natural solution (though it gets close in some places) however it is a good working approximations.

I think my description of what could happen is consistent with what BR describes, including using his own temperature figures taken from his earlier posts (I don't know where you get your numbers from, but they aren't the same.) However I'll admit that it's "guestimation based" rather than on any direct evidence.

Only partially true. A dehumidifier that didn't rewarm the air would have a very clammy feeling exhaust because the air it output would be colder but still have a very high RH because it was cold. I have experienced this with window AC units in Southern Louisiana - the air coming out is cold, but still so damp that it condenses into fog and feels clammy. (And a few minutes later the cooling coils ice up and freeze solid!)

What I am proposing in effect is to put the computer box at the output end of a fanless dehumidifier. What you and BR are proposing is to put the computer box between the condensor and the rewarmer.

Gooserider

Well, he stated clearly that an achievable scenario for the ground radiador would be 8 ºC (see previous posts). Also that he will be using peltiers. Now depends on the peltiers , but 12º isn't that far off for a heavy pelted WC given his ground radiator. And if the (auto) radiator comes before or after the cooling loop (preferably before to maximize the dehumidifier effect). Lots of If's like i've said before.


No, it will have the exact same RH of the heated air IF no condensation/evaporation occurs in the heating element and/or exaust of the Dehumidifier. It has already removed an amout of water from the air, the heater doesn't remove any more (not even it's job). Water vapour doesnt "disappear" with temperature shifts. It has to condensate , or water has to evaporate to change de current RH. Thus de Dehumidifier purpose. If the heating element is shut down, the air will exit cold, but with the same RH, but different dew point.

Same temperature can have a wide range of RH, and vice versa. What acts up on them is the enviorment.

( I'm getting the feeling you're reading the chart wrong, and getting a good deal of ideas from that. You need the RH % AND the temperature of the air in order to read the dew point chart. One without the other doesnt make sense. For all the stated reasons. At a given air temperature, at a given RH % , if the air is in contact with a surface (solid , liquid or gas) with lower temperature, condensation will form. )

Exactly.
But i already understood that. But again, the difference is the dew point. How cold are the blocks and tubes ? Dunno. One of key issues.
The difference is that BR's rewarmer has no comparison to a rewarmer from the dehumidifier. It's only purpose is to enforce convection currents which won't happen normaly given the cold air at the base. If the cold air isn't forced up (through the case) the system won't work (in BR's idea). Maintain cool air, as blocks and tubes are cool, passing through the case.

I suggest you take a look inside your refrigerator. Hmm, not much condensation at the groceries, right? But in the back you can see some sort of plate with plenty condensed water. The water condeses there because the plate is the coldest place inside the frigde.

If you make relatively air tight case with a small plate or block as the first part of the wc-loop then the same thing will happen. The humidity will condese at the plate instead of the cpu waterblock since the plate is coldest.

I guess somebody could have mentioned this before in the thread but it is only 16 C in appartment right now so it is kind of hard to concentrate.

Btw, I am not dyslectic, just from Norway

That's because the inside of the fridge gets to a preset temp, (around 5C), and so does the air trapped in there, along with the non heat producing food.

The reason you get condensation form ice on the cooling plate part, is because its much lower than the temp of the air and food in the fridge part. The dew point equation has just be moved lower down the temp scale. It's possibly reduced like you say by the fact a fridge is sealed and insulated for greater efficiency.



Admittedly I haven't got an air tight case, but I just don't think this would happen. You'd get more condensation on a colder part yes... but if other parts were below dew point, they will also form condensation too. I can't see it magically avoid those parts and divert somewhere else because it's colder.

It will work in the long run if you maintain a stable coolant temperature and humidity inside the case.
You were planning to do something like this I assume. This will maintain a rather stable water temp throughout the year. This guy's setup varied 4,3 C over a time of 2 years.

We all agree that the first part of the wc loop (the plate) will be the coldest part and will extract humidity when it's temp goes bellow dew point.

The next step in the loop will not extract any humidity as long as the plate extracts the humidity fast enough to lower the dew point. And that is possible if you have; a slow water temperature change and slow air exchange.

But you must also insulate the case or you will get condensation at its outside.
Therefore it is more common to insulate the tubes and water blocks.

The 8*C number was the minimum, with a range of 8-10*C. Since we are talking about Summer humidity, I'm assuming the upper #, but it doesn't matter that much, since changing just shifts all the temps ~ the same amount. Also per BR's spec, the rad comes AFTER the blocks / pelts, which per his temp numbers the blocks will be on average 2-4*C colder than the Rad. (I don't know why we keep arguing about BR's #'s on each post, they don't change... :shrug: )
I agree the water vapour isn't dissapearing, but the air temperature is changing. As I have said before, if one has a box of air at a given temp and RH, then cools the air (but not to the dew point), the RH goes UP. If instead, one heats the box, then the RH goes DOWN

If the air cools to at (or below) the dew point then condensation will occur, but ONLY enough to lower the RH to the point where the dew point of the now cold air is below the ambient temp, but the RH of the cooled air is going to be effectively saturated so long as it is cooled.

One of my key assumptions is that the air coming through the rad will be cooled to something close to the rad temperature since the airflow will be low. (If the airflow were higher, then you would still have condensation on the rad, but not as much air temp change so more dehumidification effect would be occurring) I also make the assumption that the air temp will not change significantly between the rad area and the block area.

If the air coming through the rad is close to the rad temp, then it will have a high RH for air of that temperature. When that already cold, humid air hits the slightly lower temp blocks, then the blocks might be just enough colder for ADDITIONAL condensation to occur.
I agree it's a key issue. If they are above the dewpoint of the cold moist air coming off the rad, then there is no problem. If they are below, then there is a problem. My opinion is that since per BR's spec, they SHOULD be colder than the rad, the risk is considerable.
In the design as BR proposes, I agree, the rewarmer doesn't do anything to the RH in the computer chamber, which I see as a problem.

In my proposed alternative, the rewarmer still drives the convection currents, it just now ALSO serves to effectively lower the RH (and thus the dewpoint), hopefully enough to the point where further condensation on the cold blocks will not occur.

---------------------------------------------

Actually Skoddelos, if the air is humid enough, you WILL initially get condensation on all the stuff in the fridge, especially if it isn't a self defrost model.

After that, you have a semi-sealed box, where the water vapour condensing on the plate is effectively removed, lowering the RH of the air in the fridge below the dewpoint of the groceries, but not that of the plate. Since it's now drier, the water that was condensed on the groceries will evaporate back into the air, leaving just what condensed on the plate.

Gooserider

You are mixing some things up.
But i get what you're saying.
You're not being clear at some points. As I had to decipher parts of it.

If I miss my guess, you're just talking about the point where Rh is related to the quantity of water vapour that can be held by the air at a given temperature.

That's all nice, but the actual water vapour remains the same if no condensation or evaporation occurs. The air can carry more or less vapour, but if none of those two events occur, the real quantity of water vapour remains the exact same.