Sign in with
Sign up | Sign in
Your question

Water Cooling Core 2 Duo E6600 & 8800GTX

Last response: in Overclocking
Share
June 17, 2007 3:21:30 AM

Hi guys, I'm finally gonna ditch my CPU and GPU HSF and go for some nice water cooling, however, I've never dived into this market before and I'll be quite honest with you, I have no idea where to start, who's better or who's worse.

I've built plenty of systems for myself and for customers at my shop, but have never been requested to do water cooling, so forgive my ignorance. I'm just starting to overclock my system and want to throw my stock coolers in the nearest drawer.

What can I get for around £100 that will cool the CPU and GPU quite well (and probably the northbridge too)? I have no idea who's best, what's good and what's bad - all I'm going on is common sense at the moment and I'd really like some advice as I don't anyone else who has experience with water cooling.

As far as kits go, I can't find many around my price range that supports my 8800GTX. I'm not going to be overclocking my GPU, but if I'm going to invest in watercooling, I want the GPU cooled too, because it doesn't exactly idle at a cool temperature and load temperature would be nice to be cooler.

So yeah, any advice would be greatly appreciated! Thanks in advance!
June 17, 2007 10:25:19 AM

Look for something from swifthech. Because you have no experience i suggest going for a kit solution.
June 17, 2007 3:00:09 PM

I am in the same place as you re experience and am doing my first WC system this week. All my research led me to a custom system; a kit is NOT a good idea. I found the research to be a lot of fun - ok I am a hopeless geek :) .
For a wealth of good information try this link. http://www.xtremesystems.org/forums/forumdisplay.php?f=...
Enjoy
Related resources
June 17, 2007 3:38:25 PM

I'm doing the same research for when I upgrade to the q6600. Below are the part that I have selected so far. Initially I don't think that I am going to do the cooling for the 8800, as I may buy a new card when gen 2 Dx10 come out. Swiftech does make a kit in your price range, but most vender's will help make sure that all your parts will work together.

Radiator HW Labs Black Ice Xtreme III Triple 120mm Radiator - Blue
1/2 ID $61.95
Pump Swiftech MCP655 Water Pump
1/2 ID $76.99
CPU Block Swiftech Apogee GTX Extreme Performance block
1/2 ID $72.95
GPU WB Swiftech Stealth Extreme Duty G80 VGA Waterblock
$104.99

I'm still researching the pump switch, reservoir, and in-line temp sensors

From my research cooling the north bridge and south bridge with water alone yield little help in OC'ing vs. air. Adding a tec to loop seems to help.
June 17, 2007 5:55:21 PM

I would go with one of several full coverage water blocks that are out there for the 8800 series they will do a better job of cooling the whole card.
June 17, 2007 6:06:13 PM

Hey guys, thanks for the input. I've just been looking at Zalman's products and was thinking of getting one of the Reserators with a 8800GTX block.

What do you guys think of them?
June 17, 2007 6:30:07 PM

I have my system watercooled. I would really recommend Koolance

www.koolance.com

http://www.tomshardware.com/2007/03/28/a-beginners-guid...

E6600>
http://www.koolance.com/shop/product_info.php?cPath=29_...

OR (this is what I have)

http://www.koolance.com/shop/product_info.php?cPath=29_...


8800GTX>

http://www.koolance.com/shop/product_info.php?cPath=29_...


I Really happy with Koolance they make real good quality product and they work great and are well designed.
June 17, 2007 6:50:37 PM

Slight problem, their shipping costs a bomb to the UK. Are there resellers in the UK?
June 17, 2007 7:25:24 PM

Yeah, check eBay for a few stores that resell Koolance gear here. Its not cheap though... 300 quid for the base unit alone (Assuming you're going for the Exos 2).

Looks good though, and I might aim for one someday, when the E6600 rig is sorted. Let me know how it goes, and remember... NEVER mix copper and alu water blocks... galvanic corrosion is NOT your friend! Stick to all copper, or all alu, and you shouldn't have much of an issue. Koolance blocks are copper with a gold plating, so those are ok for most systems. There IS a review on these forums somewhere by someone who's used the gear, do a search for it.
June 18, 2007 3:23:13 AM

OK, listen. Im gonna go ahead and tell you a few companys that you would be wise to steer clear of when it comes to the world of Watercooling.

Thermaltake
Zalman
Koolance
Corsair
Aqua Computers
Innovatek

You might be thinking, " 8O wow, that was just about everything i was looking at!". Well pretty much all of those offer ready-to-go systems that wow people who dont know any better. They offer poor performance, and some even include Aluminum and Copper in their pre-designed loops. That is not good. As Cable mentioned, it will practically destroy your loop. No matter how much anti-corrosive you put in there, it will only slow down the process, while at the same time hurting performance.

Newer Koolance blocks include Anodized Aluminum tops. They claim to have protective layers to prevent corrosion, but i hear its so cheap all it takes is a scratch to inhibit corrosion. As for the gold plating, im not sure as to how that effects performance. Im pretty sure that their rads are Aluminum too, so thats not a good idea either. Pretty much its not a good idea to have anything Koolance in your loop.

I like some of ByDesign's parts. Its pretty much a choice between the MCP655 and the MCP355 (with modded top by Petra or Alphacool). Those are the 2 most common, the most extreme you can buy is an Iwaki RD-30, but those are very expensive and require a seperate power supply for full use. I like the GTX, and they prove to be superior (bowed) on Quad Core, but i think the Fuzion will catch up with the release of their Accelerator Nozzle. If you live in the UK, i would probley go with a Thermochill. I hear they are easier and cheaper to get over there than they are over here. The PA series is optimized for lower CFM fans, so you wont need extremely loud fans to get supreme performance. If you are in a jam for cash, go with Swiftech rads. They are best for Price/Performance. If you are cooling a 8800GTX and a Q6600, i would def. go with a triple rad. As for a GPU block, i personally like EK, but the Stealth is really nice for the 8800 line. You could also get a MCW60 (w/ RAMsinks) and be able to use it when its time to buy a new graphics card.
June 18, 2007 11:25:14 AM

So the Koolance 8800GTX cooling blocks didn’t drop the temps on my cards at idle by 27C, it’s just in my imagination, because you know their stuff is junk. :roll:
June 18, 2007 1:38:36 PM

Soon I'll be doing my next watercooling rig, I've been putting it off for a year or so. My last two were Koolance cases and I'm looking for something that has similiar potential for cooling the entire system.

In my systems I cool the CPU, Northbridge, GPU, and (some) hard drives. I'm not married to Koolance and will be happy to swap out parts for better stuff but to be honest there is very little out there to make me jump.

I'm not sure what WW is talking about when he means to steer clear of the Koolance cases because my oldest case is the p2-c from 4 years ago and I've not had problems with corrosion or high temperatures.

Over its lifetime I've upgraded the cpu and video cards several times.

I've upgraded the cpu blocks to the koolance 300 series and they lowered my temps on all of my cpu's I've put in those systems. The only complaints I've had is changing or topping off the coolant in the p2-c and the pc3-400 systems is a pain-in-the-rear and that the reported cpu temps are about 5-8 degrees off, but even with that and the fact that I keep them in a relatively warm location the temps are still lower than the air cooling systems I put in the (much cooler) basement. They start complaining when thier sensors hit the 45C degree mark so the fact that they easily keep the cpu under this an pretty good accomplishment when the average room temp is ~27C [okay I like it WARM] with all the extra heatsources in the loop. When I take it to a more moderately cooled room the temps rarely break 35(C).

Some of the systems WW mentions I've seen and they simply don't offer the power to cool an entire system like a Koolance (integrated) case does.

The only reason I haven't pick out a 3rd koolance case is that I'm not in love with some of the newer ones. Though I have been tempted to cut the top off of a CM Stacker 830 and pop a kit in.

As for CPU blocks, Danger Den looks good, but still I'd love an apples to apples shoot out on them vs. the new Koolance ones.

Having seen thier cool-pack technology for HD, I'm not going to even pretend there is a competitor in area especially since I run a lot of WD drives that run HOT.

[Edited because I got my answer on BlackIce vs. koolance radiators in cooling capacity.]

The worst thing people say about Koolance is the corrosion. My personal experience tells me that isn't a factor over the life of the case so next argument please because you think after four years I'd see some with all the moving of the cases and upgrading of the parts I do.

To be frank I'd love to see a shoot out

One with various cpu blocks in a hot loop. This one I have seen sorta done but usually its just a CPU and occsionally a GPU and its rarely consistent.

[Paragraph deleted because I found what I was looking for. Can't wait to see the end of this shootout.]


It would need to be a loop where you've got to cool the cpu, gpu, northbridge, and 1-2 raptors (for heat). In a warm room and a cooler room. That would stress the systems out.

That way we could get an idea of which cpu can deal with the heat (inside and out) better.
June 20, 2007 4:01:57 AM

Quote:
So the Koolance 8800GTX cooling blocks didn’t drop the temps on my cards at idle by 27C, it’s just in my imagination, because you know their stuff is junk. :roll:


Wow, it looks like i stepped on somebodys toes 8O . Before considering that statement, i must ask what other blocks you have used on your GTX, also, what are your temps now? I never said that Koolance products dont work(congratulations it works!). I did say however that they are weaker performers. If you truly want, buy an EK FC block, a Stealth, or MCW60 and see how your Koolance block stacks up agains those.

kcrush, i really wonder why you WC your HDs. I have 2 Raptors in my main system, and I have never had any problems with them overheating. Just curious.

Also, the corrosion thing is realy only a concern if you mix blocks from other companies. Like if the OP says "Golly gee! I think ill buy this Koolance GPU block and this D-Tek Fuzion!" that would be a disaster a little ways down the road. Koolance uses a gold plating, that is there to prevent corrosion. But if someone got a Koolance GTX block and a Brass/Copper rad, then you better believe there would be corrosion. It sounds like you have never used anything but Koolance, and would be the reason as to why you had never experienced corrosion. Ask yourself this question? Why does Koolance use sooo much Alu in thier loops? Its not the best performer....The gold plating can only hurt performance because it has a lower thermal conductivity than copper...so why? I really cant think of a reason.

Also, i dont really know where you get the idea that the stuff i recommended isnt on par with Koolance. Its true, that what i recommended isnt an entire kit, and you will actually have to put it together yourself, but everything i offered would outperform the Koolance counterpart you offered. Please, show me the individual pump, rad, and waterblocks that you would recommend

If you are talking about the Hydra-Pak thing from Koolance, then you are right, there really is no competitor. However, the reason for this is because there isnt a significant need to WC harddrives. Even if you do, you dont get any better performance out of them. As for myself, i would never let any Hydra-Pak into my loop simply because they are prone to puncture and leaks.

As for CPU blocks, as i said before the Apogee GTX and Fuzion are the 2 best for performance/price at the moment. The Danger Den blocks are really, really old, and they dont hold up to todays standard all that well. However IMO, the absolute best performance block is the Little River Storm G5, but those are really rare and expensive.

Also, can you link us to this CPU Waterblock shootout? I would love to see who comes on top. Also what did you think about the whole BlackIce vs. Koolance deal?
June 20, 2007 12:36:20 PM

First of all this is what I said in the first place;
Quote:
I would go with one of several full coverage water blocks that are out there for the 8800 series they will do a better job of cooling the whole card.

Now the reason I recommend full coverage is because those little stick on heat sinks just don’t work, the memory on this card gets really hot. You will also note I didn’t specify any particular block, I said one of several. As for my temps they can very a couple of C depending on the room temp, but they are very consistent, 45C first card and 51C the second card at idle. The different temps between cards is due more to the fact that the second card always ran a little hotter then the first no matter where I put it.
I know there are some people that seem to hate Koolance for some reason and others that don’t like them just because they hard someone say something negative about them. All I know is that their stuff worked for me and that’s all that’s important.
June 20, 2007 2:47:52 PM

I have built 3 rigs with water cooling and I can tell you from my experience that a pre-built kit is not the way to go unless it is a kit from Swiftech or Danger den. When I first built my rig it was all air and then later I decided to try my hand at water cool. After allot of research I went with Swiftech and love it. My CPU temps stay at or just above the room temp all the time. After I did mine a friend of mine decided he was going to do the same thing but bought a kit from ThermalTake.... His temps did not really change all that much... a few degrees at best. So I tolled him to buy the same thing that I had and wallah! He now runs the same temps as me.

The third one that I built I decided to try some cheaper Koolance waterblocks and was disappointed although it could just be because I am expecting the same performance as what I am getting from my Swiftech block.

I suggest going with 1/2" hose and no less if you plan on cooling your CPU, GPU & NB.
Here is where I buy all of my stuff: http://www.sidewindercomputers.com/

Also, I would suggest NOT using a reservoir and use a "T" line instead it greatly improves the ability of removing air from your lines. Reservoirs may look cool but they just do not work very well.

If you would like more information or help to not make the same mistakes that I made let me know and I will be glad to suggest parts and techniques
cdl1701@yahoo.com
June 20, 2007 3:06:38 PM

Quote:

kcrush, i really wonder why you WC your HDs. I have 2 Raptors in my main system, and I have never had any problems with them overheating. Just curious.


Because once while checking out my system I noticed that without a lot air blowing over those suckers some of them can get extremely hot. Granted even with my non-WC HD's I've never had a problem, but I just feel better that the extra heat is going somewhere else as a preventative measure. Plus it adds to the cool factor. :D 

Quote:

Also, the corrosion thing is realy only a concern if you mix blocks from other companies. Like if the OP says "Golly gee! I think ill buy this Koolance GPU block and this D-Tek Fuzion!" that would be a disaster a little ways down the road. Koolance uses a gold plating, that is there to prevent corrosion. But if someone got a Koolance GTX block and a Brass/Copper rad, then you better believe there would be corrosion. It sounds like you have never used anything but Koolance, and would be the reason as to why you had never experienced corrosion. Ask yourself this question? Why does Koolance use sooo much Alu in thier loops? Its not the best performer....The gold plating can only hurt performance because it has a lower thermal conductivity than copper...so why? I really cant think of a reason.


As for the alu? Well I have car guys whispering in my ear going, "Aluminum good." :) 

I'm always looking for alternatives which is why I read various forums and look at the few reviews I can actually see. I have wondered what would happen if I replaced the aluminum rad with a black ice, which is why the current shootout between the three companies is interesting reading. And I've heard lots of good stuff about DD blocks so again I read and watch.

I'm the type that does a lot of research before I commit unless I am not given enough time for a build. Which is good for me because your comment about mixing them and getting corrosion is the type of info I look for that isn't published. That just tells me I can't mix those brands. Fine, if its an all or nothing type deal, not a problem.

Quote:

Also, i dont really know where you get the idea that the stuff i recommended isnt on par with Koolance. Its true, that what i recommended isnt an entire kit, and you will actually have to put it together yourself, but everything i offered would outperform the Koolance counterpart you offered. Please, show me the individual pump, rad, and waterblocks that you would recommend


I base my ideas on other reviews that give temps that don't come close to what I acheive at home. I've seen some of the reviews do just cpu and northbridge or cpu and gpu and thier reported temps are about equal or higher than the ones I get with a koolance pc3-400 case with the CPU block 300-h06, HD block HD-50, northbridge cooled by GPU 180-v-06 and GPU cooled by GPU 180-L-06. Yes I am aware they all use 1/4" and the difference between 1/4" and larger tubing sizes and thier effect on cooling.

I know the koolance cpu sensor is off by 6-10 degrees (depending on whose review), but as I've said before in a warm room that sensor reads 38-40C and in a cool room rarely breaks 35C.

I've not seen a review that did all of those blocks, so cut them some slack since they are cooling fewer components than I am. So when the numbers from the reviews don't come in significantly cooler or when the reviewer mentions the system also has a temperature discrepency and the numbers come in the same as what I see on my home box I am kinda dissapointed.

There was one system out there whose name, I can't remember, that used basically refrigerator parts that would do the trick. But if I did that I think my land lord would kill me over the rising electric bill. :D  Oh and the ones with the huge freaking external radiators are out because of space issues, technically for space issues, even koolance Exos series is out. TEC coolers are out for the same reason the refig case is out, though the hit to the electricity bill is most likely less noticiable.

Example if I see a review held in a room I consider cool 20-23C and the reviewer comes out with cpu temps of 37C...that's equal or less than what I'm getting in a cooled because I'm adding the heat of the hardrives to the loop.

Granted my old system is using a chip that doesn't generate as much heat as some of the ones being tested now, but my next WC rig isn't for that chip its for a future chip/GPU combo that will possibly run hotter than what they are testing. So when I see a system review just do a CPU cooling I wanna see reported temps much less than 35C (going under the assumption that the actual number is higher), if its doing CPU, northbridge and GPU I'm more likely to accept it as okay.

Quote:

If you are talking about the Hydra-Pak thing from Koolance, then you are right, there really is no competitor. However, the reason for this is because there isnt a significant need to WC harddrives. Even if you do, you dont get any better performance out of them. As for myself, i would never let any Hydra-Pak into my loop simply because they are prone to puncture and leaks.


Its not about HD performance, more about keeping the ambient heat down. Those things radiate enough heat to raise the ambient temperature of the case and with my non-WC HD also radiating heat with no direct airflow except that coming from the radiator or air being pulled via the powersupply it can get quite warm inside the case, which reduces the efficiency of the system as a whole. I don't worry about the hydrapack leaking (no more than any other component) as I try to avoid cases with sharp edges. Since the HD are sealed a leak would only land on another hd and that smell would be more than enough to make me hit the off button quick during testing when they are most likely going to leak or just after a move. I do dry runs for my stuff.

Quote:

As for CPU blocks, as i said before the Apogee GTX and Fuzion are the 2 best for performance/price at the moment. The Danger Den blocks are really, really old, and they dont hold up to todays standard all that well. However IMO, the absolute best performance block is the Little River Storm G5, but those are really rare and expensive.

Also, can you link us to this CPU Waterblock shootout? I would love to see who comes on top. Also what did you think about the whole BlackIce vs. Koolance deal?


I'd love to see a full CPU waterblock shootout, mostly I get individual reviews with different testing methods which at best will test 2-3 blocks or 1 block and air cooling so I can't compare them directly to each other I can only compare them to the one control I have and that is my system.

I find the Black Ice vs. Koolance debate very interesting. With the latest report everyone focused on the 5m/sec air velocity, I was more interested in the numbers from the 3m/sec since its a lot closer to anything I'd put in my system at full blast during a hot summer day.

The only thing I don't like is the call of fanboi's from both sides ranting. If you don't like the results. Show me yours! The constant rallying to one side gets tiring and makes getting real info difficult. But I'm in this for long long haul so I have to dig through.

The latest Koolance test was meant to be an extreme example and I think that was lost on a lot of people. Between the 3m/sec air flow and the in and out water temps that is what I cared about most in that test. You can have a technically superior product that performs not as good as a comptitor in real world testing. And when it comes down to it, the performance in the real world is what counts.

Now its HWLabs turn to come up with a test that showcases the same three radiators to impress me.

Of course that brings up the topic, what would be a neutral fair test case for a) the various radiators and b) cpu/gpu blocks? One whose results could be applicable to the novice and the advance WC user (you may never satisfy all the fanboi's so I'm leaving them out).

Most tests are only barely applicable to me because my machine has more loops than most and has a higher ambient temperature. However a test geared toward me would be over kill for a novice WC user or one who just wants CPU/GPU cooling and a cool room.
June 20, 2007 4:29:43 PM

i have been running 2 exos systems for over 2 yrs, 1 over 3. with 0 problems.
the gold in the cpu blocks is good. my only issue now is with a Danger den 8800 gpu water block, which is copper and not coated. we are waiting to see if any problems arise from this. and my temps are ambient about 75' F, cpu never over 100' F, n/b 98' F, my 8800 gtx o/c 650/1060 is idle 120' F, load 140' F. i am currently running both exos on this rig 1 is video the other does cpu and n/b.
June 21, 2007 6:44:40 AM

Surf, your loop is going to have a tough, short life ahead of it....I thought i established earlier that the gold is not good, because it isnt as good of a heat conductor as copper.

kcrush, stop listening to the car guys! :x lol. Aluminum is good for alot of reasons in cars, lighter etc...but Copper is a much better conductor of heat, and will improve performance. Also, as recently tested by Cathar (if you dont know who that is, you havent done your homework), tubing sizes doesnt actually have all that great of an impact on temps. This is found here

http://xtremesystems.org/forums/showthread.php?t=147767

I have to ask also, what are you cooling? Because those temps on a really old setup doesnt say very much for Koolance. Custom building is the way to go. Period. Swiftech kits are pretty much custom built (minus the custom part, but you know what i mean). I can tell you this from my experience. My first experience w/ WCing was a TT kit. My what a mistake that was. I got better temps on air, which proves my theory, High End Air>Low End Water. However this time around I went custom. Im in the middle of a build right now (that means im on my laptop), and i will post my worklog and temps after i get my Quad on July 22.

Here is a small WB shootout, just between the major performers. Nobody really wanted to risk Al in their loops, so thats why you dont see Koolance etc.

http://xtremesystems.org/forums/showthread.php?t=131297

Fuzion and GTX on top almost everytime.

If i were you, i would put a 120mm fan on my HDs, and drop the HD blocks(it will improve your temps). Also why is a rad blowing into your case? Colder air in, warmer air out.

Also, dont say you have more loops, because from what i can tell you dont. You have a simple single loop. There are a few things that really pissed me off about the Koolance test (that they had to of realized).

1. They purposefully used Thermochill's Rad optimized for lower CFMs. I would have liked to have seen the results if it were an HE120.2.

2. They used unrealistically high CFMs, which cannot be replicated by even the most powerful fans.

3. The water that they used was not too far off from boiling. If you ever have water that hot in your loop, you are in some deep trouble. (Also most tubing isnt rated for temps that high)

4. Also, alot of their showings of other companies specs is absolutely wrong, and can be proven wrong by science and mathematics. (Cathar proved them wrong in a discussion on XS)

I would really like to see BillA, or Petra(Alex) (both really honest, good, thorough people)put some of these arguments to rest. A comparison of CPU blocks(on IHS, non-IHS, Single, Dual, Quad Core), GPU blocks (GPU only vs. FC), and Rads (real-world test on low and high CFM fans).

Also, unlike what cdl said. T-lines do not get rid of air faster, more like 50x slower than Reservoirs. Reservoirs work fine, and as long as you stay away from that Danger Den Bay one, you will be OK. (i personally am a fan of EK reservoirs.)

Kcrush, if you really want to do your homework on WCing, check out XtremeSystems, you will learn ALOT.
June 21, 2007 2:20:10 PM

Quote:
Surf, your loop is going to have a tough, short life ahead of it....I thought i established earlier that the gold is not good, because it isnt as good of a heat conductor as copper.


And what about gold's lower thermal conductivity? In a computer water cooling system, this is inconsequential when dealing with water block plating that's only microns thick. Koolance feels that the benefits of gold plating are clear.

Quote:
kcrush, stop listening to the car guys! lol. Aluminum is good for alot of reasons in cars, lighter etc...but Copper is a much better conductor of heat, and will improve performance


Yes I have it in my Mustang and in my gaming machine and it works good for both of them.

Quote:
tubing sizes doesnt actually have all that great of an impact on temps


True its head pressure and flow.

Quote:
Because those temps on a really old setup doesnt say very much for Koolance.


Go back and reread what he wrote.



This proves what?

Quote:
Kcrush, if you really want to do your homework on WCing, check out XtremeSystems, you will learn ALOT.


Yes this site has a lot of good stuff on it but it’s not the only source on the internet.
June 21, 2007 3:06:22 PM

WestWarrior thank you for your concern.
i am a machinist, and the work i do goes into space. we gold plate everything for corrosion resistance. in space since there is no atmosphere, there is little to NO radiant heat transfer, its all conductive(i.e. the side away from the sun on the space shuttle) just like in a W/C setup. my only concern has to do with my danger den 8800 gpu waterblock, which is unprotected copper. which i may send out to have gold plated, just sneak it in to the next batch of parts we send out.

also i have used koolance parts since i started w/c. its been 4+ yrs with 0 problems. but i guess i better knock on wood now.
June 21, 2007 3:22:48 PM

Overall I prefer to remain neutral in the one-side vs the other arguments. In this case Koolance gave thier test, lots of people pounded Koolance, but they have yet to provide a counter test. I'm sure that will come in the future and I will watch until then.

Interesting, that's the first time I've seen that test on the tube sizes on the effect of cooling. I'll add it to my reference materials.

Everyone knows copper is better conductor of heat than gold, however, its does do a good job regardless. And less face it all copper coolers are not created equal, some are really good, some suck, and some all fall within a good average.

And I will point out that the shootout you mention underscores one of the major problems in the Aluminum vs. Copper radiator debate.

No small indepenents will do a head to head, so the "question" only gets "answered" in theoretical explainations or big company "tests" like the one Koolance did. All of which are soundly denounced by the "other" side for this reason and this reason and that reason.

The main reason they avoid using the Koolance products is almost always given to be corrosion. Which if I must remind after four years and two machines I have seen none of. I also point out the corrosion argument doesn't fly if you are doing a test loop running for a few hours.

Given that I must presume that either we are dealing with elitist pride, simple anti-koolance unreasoning hatred (which given some of the rants I've seen from people who have never tried it, rises very high on the list), or straight up fear that someone has come up with a solution that breaks thier expectations. Their tests while very informative all smack of ignoring the elephants in the room.

Frankly, if I had the time and money I would just go test them all myself, even the ones that are in theory inferior. Until it is done the debate will rage back and forth.

I've seen numerous tests on various sites and they are informative as long as they stay in thier comfort zone. I have been perusing the Xtreme systems site and have gotten lots of valuable information and some ideas to try for my next build. Every now and again I check up for updates on this guide.

As for the fans blowing ito my system that was the way the Koolance system (2-in and 1 out) set them up. A while a go I tested the system with all 3-fans blowing out, temps went up when the air in the room was warm. I tried all 3 blowing in didn't help either.

One of the rules for my builds is that the only fans in the case are those on the radiator and powersupply since nearly everything else that generates heat has a waterblock. Yes I could put a fan in or I would plug in the one that is at the back of the case..but that would defeat the purpose.

Also since I use it as a demo machine for friend and family to show off skills before I build thier machines (which reminds me I have to go and redo the wiring again after swapping out some hds). A machine that doesn't sue the big bulky fans in everyone elses case has a huge appeal factor. And every now and again they find some new giggle factor like watercooled ram chips and guess whose got to install them? (And I still have to work/game on this machine). A small price to pay when someones little darling needs a new computer and they drop 2-3 grand in my lap to make a glorified websurfing machine whose one workout will be WoW.

I'm sure many consider it unreasonable or dangerous, however, those are the rules. It also keeps dust down to a minimum without resorting to dustcleaning meshes on the intakes.

My loop is Rad, split to CPU/Northbridge or GPU/HDs, rejoin, resevoir , Rad. A more accurate description would be I have more blocks in my loop than the average wc.

Which I base on a lot of reading tends to stop at CPU/GPU or CPU/GPU/Northbridge or CPU/GPU/GPU.

When I say WC loop, I mean from leave Rad to going back into Rad reguardless of the splits.

Old machine
Athlon 3000 OC, 6800GT OC, Northbridge, 2- 150 WD raptors, 2-WD 250 HD, 2 Gb Ram The only thing that doesn't have a wc block on it is the ram...which will be fixed in the next build...maybe. The koolance system its on is rated for 400w (according to their site).

New machine will be a C2D, I've assembled a C2Q machine before and frankly there isn't much point at the current time, even 2 extra cores can't speed vista up, which is eventually what it will be running.

Video cards are going to be WC, but since there will be low airflow that leaves voltage regulators and ram spitting out lots of heat so they need to be in the loop.

ALL HD will need to be in the loop.

new giggle factor water cooled ram chips.

Its going to spend most of its life in a 23-25C room. It is the last room to cool down in the summer, first to heat up in the winter, mostly because of the computer spewing heat all day. The room is mot a matter of bad or good choice, it is the only choice.

Basically the machine is going to be showpiece and my day to day machine and it will have to work in "unreasonable" enviroments. So when I see "tests" for custom loops that show high temps in a "reasonable" enviroment and the cpu is the only thing in the loop, thats an automatic fail. Of course that's why I've been putting this off for so long.

Quote:

I would really like to see BillA, or Petra(Alex) (both really honest, good, thorough people)put some of these arguments to rest. A comparison of CPU blocks(on IHS, non-IHS, Single, Dual, Quad Core), GPU blocks (GPU only vs. FC), and Rads (real-world test on low and high CFM fans).


As would I, but the test would need to include all the rads (including "hated" Koolance models). You can cite all the chem/physics you want and be 100% correct in the matter but until you put it through the same real world testing you put the other manufactuers stuff through there will also be that hole in any argument that they can exploit.
June 21, 2007 4:36:26 PM

The kits are nice for first time users,or people that want something quick to set up.But with some reading you can get enough info to setup your own water loop that provides better performance at the same price point.
Some good info to be had at overclocker.com or the xtremesystems link provided earlier.
If you read through the posts you'll find many who have used various kits and generally don't recommend them to people for this reason.
June 27, 2007 2:39:02 AM

hey there

if i read correctly, you're going to split your loop b/t cpu/northbridge?

realize that water takes the path of least resistance, meaning the more restrictive cpu block will get very little flow.
!