...right here.

-Ed

Wow, is that a nasty setup!

The waterblock is nowhere near flat or smooth. The fan is only blowing through the middle of a much larger radiator, and it's right up against it so you're going to get lots of turbulence noise that you wouldn't get if you used a shroud. Looks like someone threw together a watercooling kit to hit a price point rather than with any other consideration in mind.

that is hideous, how did it get 7/10? i would have given it minuses! and no im not being harsh! Look at that surface on the waterblock, its like its been lapped on the pavement outside!

I'm pondering watercooling: try get some wins with my TBred now and be prepared for a future A64. Saw that kit a while ago: for me the killer is the noisy (unquantified) pump. The consequence is it seems to really improve on good aircooling you need to spend £160 rather than £60

But OK - some criticism of the following thoughts by Seal (who has been swimming in it for a while ) and others are welcome:

1) You want the coolest possible air going over the radiator: hence a lower front intake mount is a good option. (The fact that the radiator exhaust air then warms your case is fairly academic as your target heatspot(s) - CPU & GPU - are well cooled anyway.)
2) You want the coolest possible water hitting the CPU, hence the flow should go from the radiator to the CPU (then back to resevoir, then pump etc).
3) 1/2" tubing (ID) is better, 3/8" or 10mm is OK. Avoid smaller. (that kit is 8mm IIRC)
4) You must have a shroud between the fan and the radiator.
5) Avoid joins, t-junctions, elbows cause they are lethal to the effective pressure
6) The waterblock itself will not make as much difference as the items above: sure the latest/greatest nozzled/pinned is better but a well made 'cheapo' (that hasn't been lapped on the pavement) will get me going.
7) For quietness you want Eheim. (there is another thread going on this but this is my executive summary!)

Any hints how to achieve this for £60 are welcome

Hehehe, i bet youve been reading some watercooling forums somewhere else...

1. You dont HAVE to mount it in the lower front, many people here have very effectively mounted it on the rear with exhaust air going over it. Also in many setups you will see the rad at the top of the case where the cd-drives go. That also works very well for them even though theory says hot air rises so air above the case is often a little warmer.
2. Not really, you also do not have to have pump -> rad -> waterblock, in theory this will only gain you about 0.1-0.3 degrees C temperature wise because the water is not heated up by the pump by more than 0.1 degrees c. firstly. because the water is traveling so fast, secondly because the pump would have to be VERY hot to transfer a significant amount of heat into the water, this was confirmed by one of watercoolings most senior figure "Cathar" famous for his very scientific testing approaches and waterblock designs which have been bought by dangerden! (the cascade and white water block is his design)
3. Yeah dont even think of getting smaler than 1/2" or you will change in the future.
4. Yeah thats ideal to prevent the dead spot where the motor mechanism is on fans. Provides better cooling by a good 1-2 degrees
5. Erm, only hardcore overclockers that need the very best temperatures say this. I run with a T and have had an elbow in my loop before and have one currently in there, i believe that the 1-2 degrees i may gain from not having a T is not justified because not having T's or L's can make your watercooling setup very messy with excess tube, also it isnt justified for me, as when you have full load temperatures of 40 degrees C on a fully silent setup... Overclocked! i dont need that extra 1-2 degrees and will happily substitute the extra cooling for neatness.
6. Ive always prefered low restriction waterblocks although alot of the new blocks for maximum cooling temperatures have impingement nozzles. I personally am against this as to make it work you need a pump with a very high head/pressure like at least an eheim 1250. I dont want a 1250 because theyre quite big, and vibrate a LOT more than i could handle!
7. Yeah eheim 1048 is the most reliable/reputable among watercoolers, plenty enough power!



Thats my current setup, im going to re-do it soon so it looks like my old setup below.



My old setup, as you can see it looks a lot tidyer. Also note Pump -> Waterblock -> Rad. Also note, one elbow and one T.

Someone else's quote is that "Rules are for fools to obey and wise men to diverge from at their peril" (more or less). So I can accept other locations, and notice many other all-in-one kits are coming as CD bay installs. But strikes me that to get the biggest bang for the buck out of your radiator you want your coolest air (wherever that may be) going through it. My view is CD bay installs are a consequence of space/layout decisions rather than optimun cooling?
Again this is a bang for the buck 'rule': I wasn't concerned that much about the heat the pump is adding to the water but just seems logical to me the coolest water in the loop is the stuff exiting the radiator?
Your T is saving you a resevoir (AFAIK which is not a lot) which seems a reasonable compromise. But your 1/2" tubing looks near rigid - a complete PITA for routing? And presumably the source of the "make sure the waterblock sits flat on the CPU" stuff (and your elbow on your waterblock)?

I ment the mod projects where the rad is at the top cd bay (or just behind) with a grill cut out in the top of their case and a fan attached.

I would like to state (you probably already know this) that most of your pump's positive pressure is going to the T connecter, and is wasted. Putting it before the pump would be better. Possibly by a significant margin, allowing fan speed to be reduced.

A radiator would be best mounted outside the case completly, in it's own wooden box with sufficient dampening.



If the temperature has enough difference to majorly affect performance (3°C) then a new pump or more efficient tube system may be a good thing to get.

------

My problem with that review is the review itself. There were no temperatures to compare the achieved temp to.