Choosing The Best Pump for Your WaterCooling System

1- The pump doesn't care what cpu you put in the loop so it's a non issue. The radiator will care since the more heat you put in the loop, the harder it is for the radiator to remove it. However, since you only have a CPU, this radiator is ok but if you want more thermal capacity, get a MCR320 or a PA120.2 if you want to keep the same overall dimensions. This one match the MCR320 with one less fan.

2- It's still unknown what is the thermal characteristics of the Q9450 but we can assume it's cooler than the current Q6600 due to the smaller die (45nm). If we find the best for the Q6600, we can assume it will also work for the 45nm version as well. With this in mind, only the Apogee GTX and the D-TEK Fuzion will be able to cool quads fine but wait for reviews of the Enzotech SCW-1 and the Thermalright XWB-1 on quads before taking a decision. I'm in the same position as yours and waiting for the radiator to become available will let me take a wait and see position.

3-The difference between the MCP350 and MCP355 is that the latter one is a 18W version and the former 9W. More power mean it will have a higher head pressure. This is necessary when you have a loop with a few blocks and more blocks mean more restriction. A more powerful pump will be able to push water more easily thru restrictions. However, there might be a reliability issue with the higher powered pump but it seems to be caused by a custom top. For just a CPU loop, the MCP350 is fine, especially if you pair with a Thermochill PA120.2 (this radiator is much less restrictive than the MCR220/320).

Hope this answer your questions. Veterans, if you find a error with my explanations, feel free to correct me since I'm also a n00b in WC, only saying what I learned by reading around here.

Hi ya,
I was looking for recommendations from anyone who has used a Iwaki md-20r on the best way to mount this beast. I was planning on using some old mouse mads to limit its vibration. But thought perhaps some one had come up with a better solution.

thanks!

Jax

I've 2 options: the 1st is to get MCP655 and the 2nd is to get Laing DDC2 12V 18w and Petras Tech DDCT-01s Delrin Pump Top. is that Petras Top compatible with the pump I want to get? Is it difficult to attach top on the pump? and the main question: what is better: DDC2 with Petras Top or MCP655? My loop is something like this: pump >> PA120.3 >> D-TeK FuZion >> MCW60 >> MCW30 >> reservoir.

The top is compatible with any DDC and it's not hard to swap tops. Just unscrew 4 screws, swap and screw it but don't overtighten too much or you might kill the pump later.

As for what's best, the DDC2 with top is better than the MCP655. However, on the flip side, more will prefer the latter due to more reliability. You can't get wrong with both anyway.

Xilikon, thanks. By the way, what is the reason that Petras top makes DDC2 pump better?

http://www.pc-cooling.de/Watercoolin...12V+Pumpe.html

If you don't speak German translate the page here: http://babelfish.altavista.com/

L'enFer,

Xillikon (and a lot of other people) know more about this than I and can probably explain it better. But see if this helps:

The OEM tops -- Alphacool, Petratech, OC, etc -- work better than the stock tops for a number of reasons. Three are more important than others, they are dynamically related, all having to do with the inlet side of the pump.

First -- bear in mind that a cooling loop is a closed loop. "Wave" type artifacts are introduced into the flow by the pump itself, the blocks, the loop geometry and the tubing, and eventually get back to the inlet side of the pump. This prevents the pump from loading liquid smoothly, and magnifies and transmits the artifacts, which return to the pump, and so on. By increasing the inlet's cross-sectional area -- creating a mini-reservoir, if you will -- the height between wave peak and trough is lessened and the effects of the artifacts are mostly eliminated.

Second -- The idea that the inlet side "sucks" liquid is not actually true. Pumps of the kind we're discussing here work best with positive pressure on the inlet. The increase in cross-sectional area is also an increase in volume -- which in turn is an increase in mass -- which translates into positive pressure.

It's intuitively clear that these things -- smoothness of the flow and positive pressure at the inlet are related.

Third -- Another intrinsically related inlet concern is flow laminarity. In this case, "laminarity" refers to the tendency for the center of a linear flow (i.e., mostly going in one direction, and moving through a tube) to move faster than the outside. Again -- the problem is largely resolved by increasing the cross sectional area and volume. The volume increase is proportional to the cube of the dimensional change, the area increase is proportional to its square, and the circumference directly proportional. The "outside" of the flow is defined by the circumference, while the center is defined by the area (static) or by volume/time (dynamic). However you look at it, a lesser portion of the flow is effected by laminarity in an aftermarket top. .

Something to remember when you're designing your loop -- The actual internal diameter of most high-flow g-1/4 fittings is actually 0.366", just a skosh less than 3/8". Using a 1/2" instead of 3/8" barb on the inlet side of an OEM top won't help you much, if at all.

Hope this helps,
BDL

Quote:

---

Originally Posted by boar_d_laze

L'enFer,

Xillikon (and a lot of other people) know more about this than I and can probably explain it better. But see if this helps:

The OEM tops -- Alphacool, Petratech, OC, etc -- work better than the stock tops for a number of reasons. Three are more important than others, they are dynamically related, all having to do with the inlet side of the pump.

First -- bear in mind that a cooling loop is a closed loop. "Wave" type artifacts are introduced into the flow by the pump itself, the blocks, the loop geometry and the tubing, and eventually get back to the inlet side of the pump. This prevents the pump from loading liquid smoothly, and magnifies and transmits the artifacts, which return to the pump, and so on. By increasing the inlet's cross-sectional area -- creating a mini-reservoir, if you will -- the height between wave peak and trough is lessened and the effects of the artifacts are mostly eliminated.

Second -- The idea that the inlet side "sucks" liquid is not actually true. Pumps of the kind we're discussing here work best with positive pressure on the inlet. The increase in cross-sectional area is also an increase in volume -- which in turn is an increase in mass -- which translates into positive pressure.

It's intuitively clear that these things -- smoothness of the flow and positive pressure at the inlet are related.

Third -- Another intrinsically related inlet concern is flow laminarity. In this case, "laminarity" refers to the tendency for the center of a linear flow (i.e., mostly going in one direction, and moving through a tube) to move faster than the outside. Again -- the problem is largely resolved by increasing the cross sectional area and volume. The volume increase is proportional to the cube of the dimensional change, the area increase is proportional to its square, and the circumference directly proportional. The "outside" of the flow is defined by the circumference, while the center is defined by the area (static) or by volume/time (dynamic). However you look at it, a lesser portion of the flow is effected by laminarity in an aftermarket top. .

Something to remember when you're designing your loop -- The actual internal diameter of most high-flow g-1/4 fittings is actually 0.366", just a skosh less than 3/8". Using a 1/2" instead of 3/8" barb on the inlet side of an OEM top won't help you much, if at all.

Hope this helps,
BDL

---

To be honest, I don't know more than others. I'm just stating some facts that I learned from reading all the threads and the stickies. BDL explained very well the reasons but let me sum the reasons : smoother flow, inlet straight in the impeller instead of a 90 degree bend causing a restriction and the bigger diameter of the holes compared to the stock top (the stock top is actually much smaller than the aftermarket ones.

Thanks guys.

Hey. I was just wondering if this pump is enough for my watercooling setup? I am going to cool CPU and two GPUs with a 3x120mm HW-Labs GT Stealth 360 radiator and a 1x120mm Swiftech MCR-120 radiator.

The pump: http://www.petrastechshop.com/ladd1mcin12p.html

Btw, 2nd post here at XtremeSystems. :)

And sorry for my bad English.

Quote:

---

Originally Posted by LockBreaker

Hey. I was just wondering if this pump is enough for my watercooling setup? I am going to cool CPU and two GPUs with a 3x120mm HW-Labs GT Stealth 360 radiator and a 1x120mm Swiftech MCR-120 radiator.

The pump: http://www.petrastechshop.com/ladd1mcin12p.html

Btw, 2nd post here at XtremeSystems. :)

And sorry for my bad English.

---

You should be okay as far as flow is concerned. I would reccomend at least getting the Petra's top with the DDC though. I would be more concerned on having a CPU and 2 GPU's w/ those rad's, depending on the components.

So it's to few radiators for cooling all the Hardware? Do you have any recommendations for me? My case is quite small, so I don't have much room to work with. It's a Lian Li PC-V1100 btw. :)

It depends on the components. If you are cooling an E6700 and 2 8600's then you are fine, a Q6600 w/ 2 8800's is another story. I'd reccomend at least a PA120.3, which you may be able to fit on top of the case (w/ some cutting).

You should get all the info, components to cool, case and budget together and ask the question in a seperate thread instead of posting in existing ones. You would be more apt to get a faster and better reply.

guys, suppose that i have block on a q6600 ,a 8800 gtx and the northbridge, but a lesser pump like the mcp350... there will be considerable heat and restriction, but what happens if i use 2 radiators ,maybe 120.2's to help deal with the heat, and hook the radiators up in parallel with each other so that the radiators present less of a restriction? with less restriction in the radiator side of the loop maybe its enough to compensate for having to pump through 3 blocks... im assuming that the radiators dont need high flowrate and turbulence to work well and the longer the water stays in the radiator the cooler it gets...

Isn't there an easier way?

Great thread !!!

Quick question, I have an Apogee drive with a broken barb, so I am going to put a new top on it essentially making it a mp350. I am only cooling the CPU(q6600) and nothing else, I have one radiator with a 120mm fan. Which top does everyone recommend? I am looking at these three:

http://www.performance-pcs.com/catal...ducts_id=22700

http://www.performance-pcs.com/catal...oducts_id=4506

http://www.petrastechshop.com/peddcudeddcp.html

I like the idea of a clear top from a looks standpoint, but I am more interested in the performance. Thank you in advance for any input.

Utnorris

Well, you can't go wrong with a Alphacool or Petra top :)

How come no one has tried one of their AC pumps. I mean those brass stainless steel pumps look sexy.

Hey guys

I never used Water but i want to start, i have a Thermalright Ultra Extreme 120 and i'm very pleased with it but i want to move to water, as a start i'm searching a good kit i saw some from swiftech but i want your help to recomend me.

You recomend my a kit ? or buying the separate parts ? if so wich parts ?

Sorry to bother but in my coutry there's no water cooler parts, so i have to pay a lot to import them and the cost will be high, so i have to be precise in wich i order

Thank you very much
Best regards for all :up:
Sergio

Check out the kits at Petra's. These aren't some pre-made crap, they use really quality components. Petra does ship overseas but you have to contact them directly.

Quote:

---

Originally Posted by Nate P.

Check out the kits at Petra's. These aren't some pre-made crap, they use really quality components.

---

+1 for Petra's Kits!!!!!! :D

Thanks guys, wich block is best for CPU the D-Tek or the Swiftech ?

Cheers
Sergio