Koolance - DIY Hybrid

I have been giving some thought to doing a Koolance - DIY hybrid design. I'm not to pleased with a $350 EXOS II design, but I want to use a modified 3/8 system. Seems like a lot of bucks for a pump, radiator, and reservoir. The only thing it has going for it is the auto shutdown feature for temps getting elevated.

I suppose I could build some type of circuit (like a pool alarm) that could monitor the coolant level and turn off the computer if coolant goes below a certain level. It wouldn't have to be too complex as I could use a switch that would go in the mb/power supply on-off headers. I assume this has been done before so how about some info or links to appropriate circuitry?

I wish to use my Koolance coolers on my CPU (CPU-300-V10) and the GPU-180 coolers for my NF4 and VGA coolers. My system so far is plumbed for a 3/8" CPU block and the 3/8" output splits into two (2) 1/4" lines. One line goes through the NF4 cooler and the other goes through both VGA coolers. The two (2) 1/4" lines then rejoin into a single 3/8" line for return to the pump, radiator, and reservoir.

So I have been considering an Eheim 1046 pump and a Typhoon reservoir for the internal components. I'm stuck on what might be an appropriate radiator and where to put it? Need some suggestions here.

You can see some pics and more info here: http://home.earthlink.net/~topnurse/index.html

Ok, as far as an auto-shutdown for elevated temps... most mobos these days have one already.

The coolant level sensor is un-neccisary. If it's a closed loop, the coolant cant go anywhere (unless you have a leak).

As for the pump, Ihave a dangerden ddc-12v (mcp350) and it is quiet, small, 3/8 native, and very well performing.

if you can find room for it, go with a double heater-core like a 77 bonnoville. If not, then a single heater-core will do just fine. Black Ice Radiators will work also, but have a little less performance than the heatercores. You can mount it in the top, in the back, in teh front, on the bottem (assuming you have your case on feet and air can go under the computer for intake) I have mine in the front. The top and floor are probly your best spots for a double.

Speaking of Koolance hybrids...









After the original pumps failed, everything concerning water-cooling was replaced, except for the 3 80mm fans on the top radiator. At one point I wanted to keep the original Koolance reservoir, but it literally disintegrated within a month of use, I'm guessing from Water Wetter. The system is now running 2 separate loops.

Point of the story - it's OK to mod your Koolance and make it better!

Very nice mod! I wonder what my EXOS I will look like when I get done dreaming up a way to make it run likes it's new baby, the EXOS II.

But for today I decided to bail on the EXOS II for my new SLI box. The thing was just getting way too expensive and many of the pluses were being subtracted out of the equation with not much being added into bringing it back into a price performance equilibrium. While I could see why they ditched the Colder quick disconnects they also ditched about $40 in costs as well. I'm sure with the slightly bigger reservoir and the increased pump they didn't add in another $40 so even if it did that was a wash. The change from three 80mm fans to two 120mm fans was also probably a wash as well in regards to expenses. The change in electronics is a "feature" change and doesn't do much to change the basic operation. The only real "chargeable" addition would be if it could aggresively compete with similarly priced competition units. So IMHO, a $150 surcharge didn't account for the change in tubing diameter to a hybrid 3/8 to 1/4" system. Well enough brow beating of Koolance as the market will be the ultimate decision maker

So let's get on to the fun of a Koolance/DIY hybrid. I really like the Koolance coolers and so I saw no need to change anything here. High surface area, high turbulence, and a gold plated block translates to me as "the right stuff" when it comes to dissipating heat with no problems. Since I am using a 3/8" CPU cooler that splits into two (2) 1/4" lines I see no problems in reduced flow or increased pressure. In fact I might guess that I will see moreflow as I will have a 3/8" going from the reservoir/pump side into an effective 1/2" on the return side which may drop the preload.

The stuff I picked out as compatible with my case/performance/price ratio from Sharka Corporation is as follows:

Eheim 1046 AC pump & rubber feet.
Typhoon single 5.25" reservoir
Black Ice Micro dual 80mm radiator
Danger Den black fill port
Asetek Pump control relay CTU01
Zalman flow meter
Innovatek Temperature sensor #500 504
All plumbing connections modified to have compression fittings
Pair of PCP&C 80mm Silencer fans

Even with the smaller radiator I expect to get somewhat better cooling over a Koolance "Professional" system as I will be increasing the flow rate from 2 Liters/min to about 5 Liters/min. The only drawback to this system is that it is not as plug and play as the stock EXOS system. Since I had already planned a major deviation of their system in that I wanted some changes in coolant and electrical routing to give a more cleaner and better looking implementation there are not much differences in the installation for me. Essentially the only major differences is that I will have to drill a hole in the top of the case for the fill port, make a bracket to attach the radiator to the outside back of the case, and rig up a temperature display to the front of the case.

The other changes will be mostly knick-nacky minor changes in plumbing and screwing down the reservoir and the pump. If you look at my personal website you can probably see that the hardest thing I will have to do is the internal plumbing as the brackets are a piece of pie! The total bill (not including coolers) was about $275, which was way under Koolance's proposed MSRP of approximately $350 for a EXOS II unit. Plus I get a unit that is actually much more portable and will in all probability have a considerably better cooling capacity. Not to mention it now has a much more cooler look as well

Top Nurse said:

In fact I might guess that I will see moreflow as I will have a 3/8" going from the reservoir/pump side into an effective 1/2" on the return side which may drop the preload.

Click to expand...

not really....

area (and thus flow rate) is governed by this
A = pi * r^2
where A = area, pi = 3.14..... , and r = radius

So, radius of your 1/4" tubing is 1/8, an dyou have two of them so:
A = 2 * pi * (1/8)^2 = pi/32 = .0982 square inches

One 1/2" line gives pi/16 = .1963

So with two 1/4's lines you have half the flow of a 1/2 line (neglecting resistance on the sides... wich will pan out in favor of the 1/2 as well)

Still, moving to one Black ice extreme or one single heater-core would have made your sysem sooo much better cooling not to mention quieter.

But, looks nice, be sure to make some pics

So, radius of your 1/4" tubing is 1/8, and you have two of them so:
A = 2 * pi * (1/8)^2 = pi/32 = .0982 square inches

Click to expand...

True...

But I said the wrong thing. I am using two 1/4" return lines back into a 3/8" system.

So:

3/8" ID translates to a .1875 radius ^2 = 0.03515 x Pi = .1104

.1104 - .0982 = 0.0122 which is negligible...

Top Nurse said:

True...

But I said the wrong thing. I am using two 1/4" return lines back into a 3/8" system.

So:

3/8" ID translates to a .1875 radius ^2 = 0.03515 x Pi = .1104

.1104 - .0982 = 0.0122 which is negligible...

Click to expand...

That's more than 10% of the cross-sectional area, therefore definitely not negligible.
You have shown you are more than capable of making your own entirely bad-ass Nurse-my-overclocked-CPU-back-to-life-bitches system..

There are a baker's dozen threads about the random entirely DIY system. Ignore them, and do what you think you should do, but I'll advise on a simple level.

Black Ice Micro II == teh no. It's tiny, it's cute, but it's useless for an A64, unless it's powered by four tornados.

Black Ice Xtreme (1, 2 or 3) is where it's at. These have the cooling capacity required by the A64, even an overclocked one. I currently have a Black Ice Pro II. It is the thin cousin of the Xtreme II. I am upgrading as soon as I get the $$$ to an Xtreme II [Holes are cut already in the LiLi, and the form factor is as big as I'm willing to go]

The radiator [and the pump, for that matter] are the two single most important parts of the loop. The Radiator dictates what everything else can be. With a MicroII, you're talking barely enough cooling capacity to beat high-end air cooling. A BIXII is more than capable of handling a Pelted A64. The Eheim 1046, if I recall, is a good pump. No complaints there.

Tubing diameter has been done and done again. If you have thin tubing, you must have a high end radiator, or the lack of flow will heat up the system too much.

I'll toss some physics in here (cause physics rocks, baby!) -- flow resistance is roughly proportional to the the fifth power of the tubing radius.

A length of 1/4" tube will have 7.6 times the flow resistance of a 3/8 tube; two parallel 1/4 tubes will have about 3.8 time the resistance of the 3/8.

HeThatKnows said:

I'll toss some physics in here (cause physics rocks, baby!) -- flow resistance is roughly proportional to the the fifth power of the tubing radius.

A length of 1/4" tube will have 7.6 times the flow resistance of a 3/8 tube; two parallel 1/4 tubes will have about 3.8 time the resistance of the 3/8.

Click to expand...

Could you elaborate please....

I suck at elaboration, so here's a link to the Hazen-Williams equation.

Oooh, they have a spreadsheet!

HeThatKnows said:

I suck at elaboration, so here's a link to the Hazen-Williams equation.

Oooh, they have a spreadsheet!

Click to expand...

Thanks for the link, I think.?.?

Fluid Flow & Pressure Loss - Fluid flow and pressure loss in pipes.
Hazen-Williams Equation - calculating Friction Loss in Water Pipes
Friction loss in water pipes can be obtained by using the empirical Hazen-Williams equation

The Darcy-Weisbach equation with the Moody diagram are considered to be the most accurate model for estimating frictional head loss in steady pipe flow. Since the approach requires a not so efficient trial and error solution, an alternative empirical head loss calculation that do not require the trial and error solutions, as the Hazen-Williams equation, may be preferred:

f = 0.2083 (100/C)1.852 Q1.852/di4.8655 (1)

where

f = friction head loss in feet of water per 100 feet of pipe (fth20/100 ft pipe)

C = Hazen-Williams roughness constant

Q = volume flow (gal/min)

di = inside diameter (inches)

Note that the Hazen-Williams formula is empirical and lacks physical basis. Be aware that the roughness constants are based on "normal" condition with approximately 1 m/s (3 ft/sec).

* Hazen-Williams equation on an Excel template

The Hazen-Williams formula is not the only empirical formula available. Manning's formula is common for gravity driven flows in open channels.

The flow velocity may be calculated as:

v = 0.40085 Q / di2

where

v = flow velocity (ft/s)

The Hazen-Williams formula can be assumed to be relatively accurate for piping systems where the Reynolds Number is above 105 (turbulent flow).

* 1 ft (foot) = 0.3048 m
* 1 in (inch) = 25.4 mm
* 1 gal (US)/min =6.30888x10-5 m3/s = 0.0227 m3/h = 0.0631 dm3(liter)/s = 2.228x10-3 ft3/s = 0.1337 ft3/min = 0.8327 Imperial gal (UK)/min


Man, I just got a serious headache looking at that mess o numbers.

HeThatKnows said:

I'll toss some physics in here (cause physics rocks, baby!) -- flow resistance is roughly proportional to the the fifth power of the tubing radius.

A length of 1/4" tube will have 7.6 times the flow resistance of a 3/8 tube; two parallel 1/4 tubes will have about 3.8 time the resistance of the 3/8.

Click to expand...

Similar to the conduction equation for vacuum if I remember right.

thin BAD! wide GOOD!

Math was never my forte...jeez I think I have a headache. Now where is that tylenol

I like to think of myself as a more practical type engineering type. Seeing is believing IMHO. So I think I will gather up a couple of gallon milk containers and fit them with the tubing I want to try and see how much of a difference using 3/8" through or two 1/4" will do. Will just use gravity so not as to add in any pump issues and punch a few holes in the bottom for air.

But according to the math upstairs it appears that three 1/4" lines beat the hell out of one 3/8" line. So....if I come off my 3/8" radiator with three 1/4" lines all going to different parts of the system do I come out better? IOW, one line to the CPU and NF4 chipdset, one line to the dual GPU's, and the third line to the hdd coolers. All would then return to the reservoir where a 3/8 line would take the coolant back to the pump and radiator. Seems this would work out great as the Koolance blocks all natively use 1/4" ID tubing.

Comments?

BTW, how come no one told me that the BI coolers are made out of copper? Started drilling the mount holes to use plastic rivets and I thought it was tacky that they didn't drill the holes already...self tapping screws suck big time. Where do I get a nice aluminum radiator?

Top Nurse said:

Where do I get a nice aluminum radiator?

Click to expand...

Quite a few oil coolers are made from Al. But I here they are more expensive and have less performance than cores.

I am unsure of the proper additive operation for this type of problem, it's probably inverse, but even still, it's still 7.6/3 =2.5 times more restrictive than a single 3/8 line.

You want copper. Copper dissipates heat better than aluminum. Copper is good. What other metals are there in your loop? If you're running DI Water, you're going to run into galvanic corrosion, unless you run Water Wetter or antifreeze along side it.

I just put a 1/2" ID system together yesterday. The flow rate is *disgusting*. With no loop on the head of the pump, the 1250 emptied the loop completely in under a second.

HTK, thanks for the linkage. Convinced me to go 1/2 on all my future systems!

Top Nurse said:

I think I will gather up a couple of gallon milk containers and fit them with the tubing I want to try and see how much of a difference using 3/8" through or two 1/4" will do. Will just use gravity so not as to add in any pump issues and punch a few holes in the bottom for air.

Click to expand...

With just using gravity and the weight of the water to provide the pressure, you need to be careful how you set up your experiment. Do something like marking two lines on your milk jug and timing how long it takes the water level to drop from one to the other. Have the upper line a short ways from the top, so the flow has time to stabilize and you have time to ready yourself and your stop watch. Have the lower line a decent distance above the bottom, so that there's still enough water weight in the jug to provide useful pressure. And have your air holes as big as you can, so the resistance from the air getting sucked in doesn't mess up the results.

So....if I come off my 3/8" radiator with three 1/4" lines all going to different parts of the system do I come out better? IOW, one line to the CPU and NF4 chipdset, one line to the dual GPU's, and the third line to the hdd coolers. All would then return to the reservoir where a 3/8 line would take the coolant back to the pump and radiator. Seems this would work out great as the Koolance blocks all natively use 1/4" ID tubing.

Click to expand...

Threre's no real way of knowing whether you'll come out better. It's sucks that nobody has come up with a program that will calculate the flow rate for a cooling loop. There's not even much information available on how much flow resistance a given waterblock or radiator has.

Have you bought all the parts you listed already? The 1046 is a rather weak pump, I'd like to talk into a DDC/MCP350 or at least a Ehiem 1048...

I am unsure of the proper additive operation for this type of problem, it's probably inverse, but even still, it's still 7.6/3 =2.5 times more restrictive than a single 3/8 line.

Click to expand...

But does the restriction really matter? I mean since I intend to use 1/4" tubing in some part of my loop what's the difference? It's just a given design consideration that has to go in, right? I'm just trying to mitigate the effect of the 1/4" coolers I will use. I thought about using some other coolers on the video cards, but I'm planning on 6800 Ultra's within the year and I had planned on using these Koolance coolers on them and they have 1/4" (6mm) id's as well.

One thing I noticed in the recent [H] review of the Koolance Pro cases was that while the Koolance product didn't do as well as the other products tested it didn't seem to change the ability of the cooler to properly cool an OC'd machine. Also [H] uses nothing but Koolance EXOS's in their test beds to determine how well a particular CPU will OC...

You want copper. Copper dissipates heat better than aluminum. Copper is good. What other metals are there in your loop? If you're running DI Water, you're going to run into galvanic corrosion, unless you run Water Wetter or antifreeze along side it.

Click to expand...

I think I have some nickel plated brass, some stainless steel, and the gold plated copper coolers plus the copper radiator. If DI stands for deionized H20 ie distilled H20 then I'm using it plus some stuff called Hyper Lube's Super Coolant in about a 10% solution. The Super Coolant has a nice puke transparent green color to it and has the right additives to keep the corrosion down. I also add a few drops of some anti-algae concotion I got at the aquarium store. It also seems to drop the temp a few degrees over the Koolance fluids.

Top Nurse said:

But does the restriction really matter?

Click to expand...

It depends on your goals. It's all a bunch of trade-off.

Once you've chosen your pump, you have a 'budget' of available pressure and flow. How do want to spend it? Restriction in blocks is usually okay (when it's used for turbulence), but is the convience of 1/4 tubing worth wasting a bunch of pressure to push water through it? Or is your flow already low enough (under 1.5 or so liter/min) that larger tubing will make little difference?

As a little example of the diffence it can make...the Eheim 1046 has a free-flow (no restriction al all) rate of 5 l/min, but can only push 3 l/min through a single meter of 1/4" tubing.

HeThatKnows said:
There's no real way of knowing whether you'll come out better. It's sucks that nobody has come up with a program that will calculate the flow rate for a cooling loop. There's not even much information available on how much flow resistance a given waterblock or radiator has.

Click to expand...

I guess this really is cutting edge stuff, heh? It's always the same for those who want to do something different...try it and see if it works.

Have you bought all the parts you listed already? The 1046 is a rather weak pump, I'd like to talk into a DDC/MCP350 or at least a Ehiem 1048...

Click to expand...

Already have parts though I think I will not use the Zalman flow indicator and instead opt for an Innovatek LCD unit to monitor flow, temps, and fans. Especially since, just for grins this morning, I dissected one of their in-the-coolant temp sensors. Now I know how to make them in any size tube I wish Have been looking at the Crystalfontz as well but that requires a lot more fiddling around to make it all work correctly.

Well...I choose the 1046 because it had the lowest flow rate out of the current crop of DIY pumps that were available. However, it has about 2-3 times the flow rate of what Koolance considers to be an optimal flow rate for their new Pro models that use a similar 3/8 to 1/4 hybrid that I am designing.

HeThatKnows said:
Once you've chosen your pump, you have a 'budget' of available pressure and flow. How do want to spend it? Restriction in blocks is usually okay (when it's used for turbulence), but is the convience of 1/4 tubing worth wasting a bunch of pressure to push water through it? Or is your flow already low enough (under 1.5 or so liter/min) that larger tubing will make little difference?

Click to expand...

Well I have an EXOS now that is rated at about 1 liter/min and it seems to cool just fine. I don't seem to have a temp problem in limiting my OC'd states. So I figured that adding some more coolers which would need more pressure would balance out with increased pressure of the 1046 pump.

As a little example of the diffence it can make...the Eheim 1046 has a free-flow (no restriction al all) rate of 5 l/min, but can only push 3 l/min through a single meter of 1/4" tubing.

Click to expand...

My father always told me that not going to math classes was a bad idea but I just wanted to be a nurse and while math skills are needed it's only basic algebra. So are you suggesting something bigger with my planned setup?

Added:

Forgot to mention that increased pressure might be a bad idea as the above mentioned video coolers utilize Koolance's Hydra-Pak soft cooler technology. I have been wondering whether an increased pressure would blow out the seals. Koolance didn't recommend using any of the typical DIY pumps because they said they couldn't guarantee the results <SOS>. I figured I was pusing the envelope with the Eheim 1046.

Hydra-Pak soft cooler technology? I'd never heard of that. Dammit, now I've got more surfin to do...

I admit that I'm madly in love the Laing DDC pump, but I'm not really trying to convince you to do anything. I'm into analysing the options and making informed trade-offs. If you're a 'do it now!' kind of person, you should just ignore me cause I'm a 'contemplate it awhile' person.

Hydra-Pak soft cooler technology? I'd never heard of that. Dammit, now I've got more surfin to do...

Click to expand...

It's a Koolance proprietary technology. Essentially you have a bag of water that lays against the hdd and absorbs the heat from the chips and the spindle. I doubt very seriously if it will ever work with the bigger is better philosophy of the USA scene though. However, all those with Aquaero, Innovatek, Koolance, and other low flow systems will be buying them like they was selling hotcakes at the county fair

I admit that I'm madly in love the Laing DDC pump, but I'm not really trying to convince you to do anything. I'm into analysing the options and making informed trade-offs.

Click to expand...

Actually I do like that pump! But in this particular instance I'm not sure it would work right. And I'm what they call a developer-creative type person. I usually inherrently know the right way to do something, but get hung up in the implementation phase in that I want it to be perfect

Been doing some research and have decided to ditch the CPU-300-V10 (3/8") cooler and just use the standard CPU-300-H06 block. I'm having a tool & die maker friend make me a coolant header that will have 3/8" on one side and three (3) of the 1/4" compression fittings on the other side. I have four (4) holes on my reservoir so now I will either have to add another hole or do a T fitting thing for my DD fill-port. Each 1/4" loop (CPU/NF4, GPU's, HDD's) will return separately to the reservoir. One of the reservoir ports will return to the pump via a 3/8" line and progress to the radiator.

This will make it much easier to utilize the Innovatek or Aquaero in-the-coolant sensors. I am going to monitor the temperature coming off the radiator and between the reservoir and the pump. Later down the road I may look at running sensors on all three coolant lines. It is difficult to make an informed judgement about individual cooler devices if you can't access this information. Then I can change coolers and really see the difference on device design.

Also have been kicking around the idea to move the pump off the bottom of the case. Partly it is a space issue, but I am trying to keep most everything in the mb removable tray so I can take it out with a minimum of fuss. The radiator is being attached to the back of the mb tray so that part of the solution is solved and it allows me to add bigger radiators if needed. The really biggest part of the case that is empty is the area directly above the CPU. So I think I will look into putting the pump and the reservoir (if possible) right above the mb. This would be a real bitch, but with my sub-frame design I think it will be doable. Just think...a completely self-contained cooling system that easily removes from the case with the twist of a few thumb screws.

That's neat! Water-balloons in your computer. I hope some reviewer pumps the thing up until it pops, so we can see how much pressure it really can handle.

When your tool-and-die friend makes your splitter manifold, see if you can integrate the temperture sensor. With it built into the manifold you'll have fewer connectors which is always good. I like the fact that you're measuring water temps -- all too often people ask if a bigger rad will help, but they never have the water temps needed for an informed reply.

Got a sketch or diagram of this frame idea?

Excellent idea! That is exactly the right place to integrate the sensor! The Innovatek temp sensor thermistor unscrews and I can have my friend just drill and tap another hole for it. One less thing to figure how to attach to the case somewhere.

I suspect that Koolance came up with the Hydra-Pak technology quite by accident. The pics I have seen bear a strong resemblence to the material they use to ship coolant over the pond.

Got a sketch or diagram of this frame idea?

Click to expand...

Go here to my site

Well I suppose I could have got all that stuff into the mb tray, but I would be able to hire on as a swiss watch maker when I was done. So I decided that it was time to re-evaluate my project. As they say "out with the old and in with the new" got me a new Lian-Li V1200B today. That case box is bigger than my gigantic suitcase! It looks like a very large carry-on in the flesh. So now I can get busy putting in all the new goodies I have been designing. Also get to do more mad scientist work at the skunkworks again building more sub-frames

Very nice case, though I am bummed about a couple of things though. First it was goodbye to the removable mb tray, but what the heck...it's a BIG case so no problems getting inside to look around. Second was that the pic on the outside of the box shows a lot of room between the front 120mm fan and the first set of HD racks. I was already drooling at the thought of popping in a 120mm radiator there and in the back of the case to boot. But they moved the HD cages forward and instead increased the room in the PSU area One other slight problem....no window! So I guess I get to cut my own and do it the old fashioned way...

So now I am deciding whether to take it back and get a V2000B instead. That case has twelve internal 3.5" HD racks (2 x 6) and I figure I could remove the front rack and still have space for 6 HDD's. That would open up a lot of space for the 2nd radiator. The other alternative is to move all the stuff back (on the V1200) to where it is in the picture on the box. However, that requires drilling out about 17 aluminum rivets and then re-riveting it all back in again. Guess I gotta make a call to the importer to find out why they changed the case design.

Even without any mods I can easily stick in a single 120mm radiator and many other places in the case beckon to be radiator modded. So how about some new suggestions here for the radiators now that I got a lot more room?

I also got a dual 5 1/4" reservoir so I can do my 3 separate lines AND the DD fill-port plus two holes for output in case I decide to ever run two separate loops.

I'm back!!!

Got tired of bagging on Lian-Li and decided to move forward with my V1200 Koolance/DIY project. I added more content and pics to my website.

Decided to take out the forward HD rack and make room for a future 2nd radiator. That leaves room for my current three Raptors. Does anyone really need more than 3 hard drives? One for the system and two for the RAID 1 right? PSU, MB, HD's, and reservoir are in. Now I got to design up some mounting plates to secure the radiators in place and get my tool and die maker friend to help me out

Top Nurse said:

One other slight problem....no window! So I guess I get to cut my own and do it the old fashioned way...

Click to expand...

TopNurse, you can get a nice bolt-on-window side panel for the PC-V1200B here

I was planning on buying one of his pre-modded Lian-Li cases to use with an Exos II or Aqua Computer setup. Or maybe I'll just go for the PC3-725BK. It should be about as good as the Exos II with the fans reversed. I thought it actually performed pretty well in the review given that the heat exchanger was cooled with hot air from inside the fanless case

I added more content on the V1200 project. Eheim pump ready to be mounted and wired tomorrow. Radiator and fan are now installed. New pics and material uploaded to my site.

ekovalsky said:

I was planning on buying one of his pre-modded Lian-Li cases to use with an Exos II or Aqua Computer setup. Or maybe I'll just go for the PC3-725BK. It should be about as good as the Exos II with the fans reversed. I thought it actually performed pretty well in the review given that the heat exchanger was cooled with hot air from inside the fanless case

Click to expand...

Actually I think if the fans had been reversed it would have raised a lot of peoples ears right up. For some strange reason no one thinks a low flow 1/4" system can compete with the big boys and girls....

I'm ordering an Aquaero VFD unit with a flow sensor. I already have Innovatek in-the-coolant sensors which I will use with it as well. Thinking about the $110 for the Aquastream setup as well. I guess this isn't the time to be quibbling over a few dollars after what I already spent, heh?

BTW, thanks for the windows link!

Added 2-25-05:

New pics up and mockup of plumbing. Waiting for Koolance hard drive coolers, Aquaero, and Aquastream hardware. Though it will run without the AC stuff I would like to get it in as soon as possible. However, I'm going to Florida for a while so this project and myself will be off the air for a while. Hopefully when I get back all the stuff will be here because I have to go back to work soon and make some more scoots