watercooling solution confusion...

CPU Cooling tricks and techniques.

Moderator: The Mod Squad

Postby Twisty » Sat Jan 21, 2006 7:35 pm

I see, looks good. You can guess the performance of the system if you have a Watt/degC rating for those coolers.

Think about the physics of the system, lets look at the heat flow
<Obviously this is idealised, a little heat will leak at each stage>

Cooling Block-->Coolent-->Radiator-->Outside air

So no heat is lost due to the volume of coolent, so it will make no difference to running temps.

However more coolent means it takes the system longer to warm up, and also better at absorbing transient heat loads.

There are other factors to consider more water will mean a larger storage tank which will radiate more heat to the outside (so actually lower steady state temps a bit).

One can further this. For each of the arrows (where heat is transferred from one medium to another) above there will be a temperature difference, I'll guess some values.

Cooling Block--3C-->Coolent--1C-->Radiator--10C-->Outside air

Suppose the outside air is 20C, that means that the cooling block is 20+10+1+3C=33C

Needless to say if you raise the temp of the air around the air around the radiator by 5C, the water block will warm by 5C also. This is why it is a good idea to have the radiator outside the case.

Changing the conductivity of the coolent will effect the values (just guessed) of 3C and 1C as this is where heat transfer to and from the coolent takes place.

Changing the radiator will effect the 10C difference between ambient and radiator.

If you are familiar with electronics this can all be done by drawing an analogous system using current sources and resistors. One could also draw up a nice system diagram in the laplase domain and plot a root locus drawing, OMG this stuff I am revising ATM is really affecting me, LOL.
I have now left the Building :tongue8:
Twisty
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1970
Joined: Fri Mar 02, 2001 1:01 am
Location: UK / Singapore

Postby ThEvil0nE » Mon Jan 23, 2006 8:19 pm

Very nice infos you have :mb_2thumbsup:: seems like you really know your stuff. (I can't wait to bore you with my questions) :twisted:


Anyway what do you think of my cooling design?

Image

As you can see, the water flow design is not like the typical water cooling setup. In typical designs, the pump's head goes directly to the blocks which I see it as a design flaw. The pump generates heat as it takes alot of running time and with that, you are not supplying the coolest water in circulation possible. My design on the other hand, water blocks are supplied with the coolest water in circulation. The T/Y connector/splitter ensures all blocks gets water directly from the radiator. This is also the reason why I am opting for a dual pump or as I have desided, a single more powerful pump.
Image Image
ThEvil0nE
Black Belt 3rd Degree
Black Belt 3rd Degree
 
Posts: 3937
Joined: Mon May 06, 2002 12:01 am
Location: New York

Postby Twisty » Tue Jan 24, 2006 5:25 am

I agree with you on the pump. The only thing to make sure is if the design of pump you are using is as good at pulling water though it as pushing.

I would shift the reservoir to the other side of the radiator. This will mean that the water going into the radiator will be hotter which in turn means that it will exchange more heat and you will get lower temps (this is because rate of heat transfer is proportional to temperature difference). Also the water will cool a little more as it is standing in the reservoir before recirculation around the system.

I am worried about the splitting of the tubes in parallel to the cooling components. It will be very hard to distribute the water as you need it, enivitably the resistance of each path will not be equal so each will not recieve a 1/3 of water flow.

There may be some advantages in running parallel but remember you are still putting in the same amount of heat into your system so no matter what you do here your coolent temps will remain the same.

Neither will you compensate for this by using a more powerful pump, higher flow rate just means warmer coolent entering your cooling blocks (less time in radiator) and the same temperature coming out (less temperature increase in block becasue of faster flow rate). IMHO a more powerful pump just creates more heat, and huge bore hoses are just awkward in the case.

I would just put them all in serial as per a normal setup, which allows one to priotise the order of cooling (i.e CPU, GPU, chipset), it isn't particularly important to have the chipset ice cold. If one must bifurcate then I suggest you use adjustable clamps on a couple of the hoses to adjust flow rates to each component, or even better make a T block with one entry and three exits with adjustable apertures.

Hope this helps.
I have now left the Building :tongue8:
Twisty
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1970
Joined: Fri Mar 02, 2001 1:01 am
Location: UK / Singapore

Postby ThEvil0nE » Thu Jan 26, 2006 12:50 pm

How about adding ammonia in the mix? Water cooling systems in general have used ammonia/water solution. 8O :? :o :? :roll:
Image Image
ThEvil0nE
Black Belt 3rd Degree
Black Belt 3rd Degree
 
Posts: 3937
Joined: Mon May 06, 2002 12:01 am
Location: New York

Postby Twisty » Thu Jan 26, 2006 2:06 pm

Really? Got any links?

I know ammonia is used in fridges, like the mobile type that is common in camper vans, but that is phase change, not just plain water cooling.
I have now left the Building :tongue8:
Twisty
Black Belt 1st Degree
Black Belt 1st Degree
 
Posts: 1970
Joined: Fri Mar 02, 2001 1:01 am
Location: UK / Singapore

Postby ThEvil0nE » Thu Jan 26, 2006 4:35 pm

I'm just reading along and yes you are right with the fridge.

For a more practical cycle see the icyball patent. This uses an ammonia/water cycle, ammonia is the refrigerant and water the absorber...

While my device is a departure in some of its principles of operation from standard ab sorption systems, in that there are no check valves employed by me, it operates through a similar cycle to all intermittent absorption systems, that is to say, I provide two receptacles, one a generator-absorber, which contains a suitable liquid, such as water, and the other an evaporator-condenser, which contains a suitable refrigerant, such as ammonia, with suitable connecting paths between the two receptacles so that when the generator-absorber is heated, the gases will pass over into the evaporator-condenser receptacle, by one path, and when the generator-absorber is permitted to become cool, it will reabsorb the gases, as they volatilize in the evaporator -condenser receptacle, said gases pass back and are absorbed in the liquid in the generator-absorber, with the result of absorption of heat (production of cold), within and about the evaporator-condenser receptacle.

I'm still trying to figure this thing out... the design and use of ammonia requires you to build the same thing, right? OR... would a small amount of ammonia give the same refrigerant effect if mixed with the larger amount of water and a small amount of anti-corrosive agent :? :roll:

Absorber/refrigerant pairs.
There a zillions of possible absorber/refrigerant pairs, many are still untested (or undiscovered) - none that I know of are perfect. The perfect pair would be cheap, nontoxic, non corrosive, environmentally safe, have a convenient working pressure, low absorber specific heat, high refrigerant latent heat and so forth. A continuous cycle requires a liquid absorber, an intermittent cycle can use either a liquid absorber or a solid adsorber.

The working pressure is governed by the refrigerant, the pressure will be what ever is required to make the refrigerant boil at the desired evaporator temperature and will be higher in the generation phase for intermittent cycle or in the generator in a continuous cycle. The ideal working pressure is probably lightly above atmospheric pressure. This is because a slight leak of refrigerant out of the system will degrade performance less than a leak of air into the system.

Some working pairs include, Ammonia/water, Ammonia/Calcium chloride(solid), Lithium bromide/water, Zeolite/water and Activated charcoal/methanol.

* Water makes a poor refrigerant where low temperatures are required. You obviously can't make a freezer using it because the refrigerant will freeze before your ice-cream does. Even at slightly higher temperature (say 4C) its vapour pressure is still very low. However it may be suitable for air-conditioners but even here the low pressure can be a problem. Well at least its cheap!
* Ammonia works at much higher pressures around 5-12 atm, this has its problems too. When used in a continuous cycle, you get a big pressure difference between the absorber and generator, this requires a pump to make it work. However adding hydrogen as a "capping gas" fixes this problem , I think this means you have a mixture of hydrogen and ammonia in what is labelled "low pressure vapour" in fig 4. The reactions are based on the partial pressures of ammonia not the total pressure of the mixture - so we can increase the total pressure to reduce the liquid level difference between the generator and absorber so that a pump is no longer needed. When Ammonia/water is used as a generator - water is released with the ammonia. This degrades performance and requires that intermittent systems be "reset" often by pouring the excess water back where it belongs. Ammonia is poisonous and corrosive.
* Methanol, higher pressures than water but still pretty low, a low latent heat, toxic, inflamible (explosive?), often contaminated with water. It is unstable at high temperatures when catalysts are present.
Calcium chloride when adsorbing ammonia increases its volume about four fold. The CaCl has to be processed into granules so the ammonia (gas) can penetrate.
* Lithium bromides main vice is that its very expensive.
* Activated charcoal is a better thermal insulator than we would like. It may also already contain trapped chemicals you have to get rid of before you can use it.
Image Image
ThEvil0nE
Black Belt 3rd Degree
Black Belt 3rd Degree
 
Posts: 3937
Joined: Mon May 06, 2002 12:01 am
Location: New York

Previous

Return to Cooling Tips and Tricks

Who is online

Users browsing this forum: No registered users and 1 guest