Which Pump?
 

Im currently about to build my first water cooled rig and was wondering what pump would be best. I have a Danger Den RBX, Zchipset, and maybe in the near future a GPU block as well. I currently have a Eheim 1260 but the thing is damned huge and I remember hearing Cathar said the heat from it would overtake any pros the high flow would give. I was thinking of getting a Swiftech MCP600, but would like to hear any other suggestions for a really good small-medium sized pump. Thanks! :)


-Ricey

I'm not sure if they fixed the MCP600, but my suggestion (as I have stated many times :D) is the danner/supreme Mag 3... it just rocks for the money, and is very quiet. It will definitely run the setup you are using, plus the additional GPU block easily.

Mag 3 puts out a tonne of heat too... I would still advise it though, second to the MCP 600.

Would the heat output affect inline usage? Or just submersed?

Inline transfers very little heat to the loop, submerged transfers all the heat to the loop. It is only 35 watts.

oh, to clarify somemore about my setup I have a Athlon 64 3200+. Also have 1/2" tygon tubing, water wetter, and a black ice extreme radiator. After reading a bit i was thinking about something better than water wetter and the BLX but im not to sure which items are better than these. If anyone has anymore suggestions, let me know please, first time with watercooling as well :confused:

got to have the mp600 awsome pump and the best service i am have recived so far and it pumps like mad

Heater core. Not all of the heat, only around 22 watts go to the water while submerged, because some of the energy put into it is turned into mechanical energy.

The MCP600 is now fixed and known as the revision 2, I went and bought me one from http://www.sidewindercomputers.com.:D

I was told the MAG 3 puts out quite a bit of heat compared to the Swifty which only consumes 9v, which might affect performance.

I was thinking of going pump>rad>blocks>res>pump, would the MCP600 handle it if i ran it though the rad first?

Also, i wanted to add a UV additive into my system, but I'm afraid that most coolants (like water wetter) has color already. Is there a good clear coolant out there i can use?

disregard the UV additive part, found out it works with water wetter and most other coolants :p

Also thanks for your suggestions, planning to build this sucker by friday, so keep those suggestiongs coming in :D

I'm pretty sure the radiator preforms best at the end of the loop, where the water is going the slowest. It's just common sense that; the longer the water stays in the radiator, the cooler it will get.

I kept reading a lot about how water wetter stains tubing, so i think i'll decide to go with Airpsirits hyperlube mix except without antifreeze. About 85% distilled water + 10% hyperlube + 5% sanitizer. Will probably add blue UV additive to it as well.
Anyone else agree with nonphixion about radiator at the end or should i keep it in the front of the loop?

The mag 3 consumes 35 watts, but remember that not all of that goes into the water: some of that is dissipated in the air.

Also, concerning the RBX, has anyone in the forum tested and posted the results for the different nozzles?

the water doesn't go any slower at the end than the begining of the loop - it's a closed hydraulic system....

water temperature will vary - that's why people often run pump>rad>cpu>(other blocks)>pump - so the CPU block get's the coldest water, and the rad the warmest...

I'm afraid that your common sense must be broken. Perhaps some reading on the topic of water-cooling will help fix it some. :)

Maybe if you'd point me in the right direction... I could educate myself on the topic :D

I'm just getting into this whole watercooling thing, and I guess... the more you know, the more you don't know... or something :confused:

yea, first time here too, so if my questions seem dumb to some of you sorry about that, just dont want to screw up my first dive into water cooling :p

Ok, i decided to get a Swiftech MCP600 and go with 85% distilled/10% Hyperlube/5% sanitizer. Gonna get this stuff within 2 days. Anyone have anymore suggestions? :)

Well, I can give you a quick answer right here. For more info, post a question on a different topic.

Water cannot slow down in one part of a water loop compared to another part, because water does not compress or expand. So, the same amount of water is flowing per unit time for every part of the loop. Another way: each section of your water cooling loop has the same number of gallons per hour flowing through it.

Thought experiment. If it was flowing slower in the radiator, that would mean that water was piling up behind the radiator waiting to go through. Thankfully, that doesn't happen.

Now, flow rate (gallons per hour) is NOT the same thing as velocity. You can have water going at 100mph or 1mph while using the same flow of 1 gallon per hour. If the tubing is really thin, you have to push it through at high velocity to move that gallon through in an hour, while pushing one gallon through a foot wide tube takes much less velocity.

overclockers.com has a good article that cover the basic principle involved in watercooling. It's a good reading.

Radiator efficiency is, in fact, not only based upon flow rate, but of size, surface area, airflow and the like. It is best at a certain flow rate, and is worse both before and after that, for the surface area, size, and the airflow of the radiator. Speed can change in the loop, but it will probably be minimal.

Where do you take that from? As far as I know (mostly from my thermal transfert class), the efficiency of an heat exchanger (a rad for exemple) will not have a maximum, but rather an asymptotic evolution pass a certain point. (i.e. the efficient will still improve but with a very slow slope). That is, looking at the rad alone.

Also, what do you mean by speed? Velocity? If that's the case, then how can you say that the change will be minimal within the entire loop? Especially with all those wbs using impingement (sp?) plate.


P.S. not sure asymptotic is a word, might have just made it up from french! :P

Sorry I wasn't thinking about jet impingement. http://www.overclockers.com/articles481/ The cooling peaks, and then goes down.

Helps to look at the correct graphs:

http://www.thermal-management-testin...issipation.htm