OK, managed to find some more information on it. Water-X CPU Water Cooler Few things that worry me about the design. 1. The main reason watercooling is great for use in a pc is that you can efficiently transfer heat from a small die to a large radiator. With this design, you are not getting the heat to a larger area than with a normal heatsink, yet you are introducing another potential loss of efficiency with the water stage. 2. Both waterblocks and radiators perform more efficiently with higher flow rates. Performance drops down to a very low value as the flow drops to zero, effectively, heat must diffuse through the water. However, with this design, the pump is very small and will not be able to provide very high flow rates. This will greatly affect the efficiency of both the radiator section and the waterblock section. In addition, I'm not entirely sure I like the idea of having a mag drive pump so close to the processor. 3. The unit combines an aluminium block with a copper radiator. They mention that they include a special cooling fluid. It is likely that they will have to load this full of chemicals to prevent galvanic corrosion. This will affect the thermal properties of the cooling fluid as well. Also, these additives don't prevent galvanic corrosion, they merely slow it down to a point where it shouldn't be a huge problem. But it will still occur. This is fine with conventional watercooling as you can change the water regularly, clean your blocks and so on before you get a build up. (or just use copper throughout!!) However, this unit is sealed! 4. Only the best designed waterblocks, when coupled with a good, fairly large radiator and a good pump, significantly outperform HSFs like the big Swifty ones. Yet I find it hard to believe that they can manufacture an efficient waterblock into a unit that costs less than most good waterblocks alone. As you may have guessed, I have my reservations. I think I'll stick with conventional watercooling. 8-ball
As I predicted, it does not perform very well. http://www.hartware.net/review.html?id=282 performs at 10 degrees C above the SLK800 (with equal dBa fans) on an XP2000+. Note the site is in German. 8-ball PS, I would still be interested in a second opinion, though I think someone has clearly tried to design something which would appear to be attractive to those not in the know. AKA: a con.
Check these links ... one review of the sink and a second link showing how it works : REVIEW SHOP EDIT : read first, then post. The reviews the same as the one posted above Cheers [DE]FreD_S
"Let the buyer beware" About covers it. edit: I knew I'd seen this before, and not long ago. Taking pre-orders @ www.1coolpc.com/waterx.htm Anyone interested????????
Reading the build up that site is giving this thing forces me to choose between two options. 1. Those guys at that web store don't test things they sell. 2. Those guys at that web store don't care what they sell. I think this is called a no win senario.
They are con-artists. They say themselves the unit is self contained, but they make it out like its a benefit! Think about it, most water cooling rigs have a reservoir and a radiator, often with a 120mm or 2 80mm fans pushing air through it, yet this thing seemingly removes all the heat from the fluid with a few measly copper fins and a 60mm fan, which is so sandwiched between the two layers that its already tiny CFM is choked. Its a gimmick designed for impulse buying.
Chambochae, Agreed. Sad thing is that while they won't fool those who are really into puters they will reap a quick profit off those just getting into building. A real n00b reading that web store's BS will buy in a heart beat. Thing that web store should think about is that those so burned will never be back. And it is a sure thing I won't be buying from that site. If they will screw the n00bs............ I hope those at sites like this one all over the web, take note and shun that site. Disgusting.
Ok peeps well I finally got a replacement, and once my "spare" processor arrives in the next few days I will be testing it ! But I have fired it up, takes 2 12v supplies, one for pump one for twin 60mm's and its bloody loud !
I know, I know its not efficient........ BUT The idea is different you hav to give it that at least! (or do U know different). There is no Technical reason why the "problems" 8-bALL has so correctly highlighted could not be re-engineered..... water cooling solutions will become smaller and more effiicent I liked the pump connection to the mobo idea even if the application of the design is misguided.... Its different and promotes debate on the subject. I look forward to the results TiGerMan Keep us posted
Water cooling can only get so much smaller. Allow me to explain why. Heat exchangers have an efficiency rating at a given flow rate, written as C/W. This implies the temperature difference required to move x amount of heat. Currently the best performing block on the market is Cathar's white water. At reasonable flow rates, this can give a C/W of around 0.17. The radiator will also have a C/W ratio, which will vary depending on the flow rate of water and the cfm of air. consider a processor putting out 90W. We will assume no other blocks and that the pump is not introducing any heat into the water. Also assume that the temperature of the water is constant throughout the loop. Before anyone goes mad, this is a very reasonable assumption, as long as you have a reasonable flow rate through the loop. When considering temperatures you will achieve, you HAVE to start at your final cooling media. This is AIR. You will have air of a particular temp being blown through your Rad. You will have to dissipate 90W of heat from the rad, so the water will have to be sufficiently warmer than the air that the temp difference between water and air divide by the rads C/W for that flow and cfm is equal to 90W. Now consider that as the processor is producing 90W, the waterblock will have to remove 90W of heat. This will only happen once the die is sufficiently warmer than the water that the temp difference divide the C/W for the block at that flow rate is 90W. This determines the temp of the die. So if the rad gets smaller, then it's C/W value will increase. As such, the water will have to be that much warmer than the air for the rad to be able to dissipate 90W. Now consider the block. In order to keep the costs down, the design of the block will be relatively simple. This will equate to a higher C/W value for the block at a given flow rate. Also consider the flow rate. Reducing the flow rate, particularly to the levels which a pump small enough to fit into this kind of design would be capable of, will greatly increase the C/W of both the rad and the block. Put simply, water cooling introduces another heat exchanging step, when compared to a regular HSF. However, it can only overcome this by using a radiator of a considerably greater surface area than would be allowed around the socket. The main problem will be producing a water block sufficiently efficient that using a rad so small would still produce a system with an overal C/W lower than a conventional HSF. It probably could be done, but it would cost several HUNDREDS of pounds, and that is no joke. As a materials scientist, I can tell you that it will not become any cheaper to manufacture anything of this efficiency for many years. Then you have to think about what you could do with a HSF if you were prepared to pay that much for a cooling solution. I can tell you right now, that the way to improve cooling as a single unit attached to the CPU is by the use of a vapour chamber base. The reason water cooling is effective is it can remove all of the cpu heat from a very small actively cooled area, as shown in Cathar's water block, which features very small fins only over the core. With conventional HSF's, the problem is the spreading resistance of the base. It is difficult for the heat to spread to the outer portions of the base. In order to do so, a temperature gradient will be set up, from the heat source to the edge. With current designs, this means that the core will still be very warm, as the outer parts aren't doing much. A vapour chamber base would have an effective thermal conductivity of up to 20,000 W/m/K compared to 400 W/m/K for copper. This will allow all of the fins or pins on the heatsink to be effectively used. This will greatly increase the efficiency of HSF's. In fact, it has already been done, but only as an OEM solution. If it were manufactured to the same quality of engineering as current top of the range HSFs, the performance would be incredible. Conventional water cooling would not allow much of an increase unless you were using chilled water or peltiers. Sorry for going on so long. If any of it doesn't make sense or I have made an incorrect argument, please let me know. It's been a long day. 8-ball
More information for you. Here we have a kind of mini-review in a forum. Still not looking good. review Happy reading 8-ball
Ok peeps here's a few updates. Firstly I got an email this week (spam) saying look at our new product, which was said heatsink. So I told them I already had one, and as far as they are aware they are the only manufacturers, and the model I have is a pre-production model. So anyway I got my spare proc this morning. Its a P4 1.8, and fired up a fresh install of XP on my albatron px845PE Pro 2, with 512mb of 2700 OCZ ram @ 266 (wanted to test at stock). I installed the intel official heatsink 60mm fan version, and did some testing (mobo, cpu): IDLE: 26°, 28° LOAD 28°, 42° That was with about 40minutes in both states. I then installed said heatsink at got the following results.... IDLE: 29°, 28° LOAD: 29°, 39° As we can see there is a slightly difference in the motherboard temps, I belive that is down to my central heating kicking in. That is why the mobo temps with the standard h/s are slightly lower, as heating probably kicked in between the two HSF test. So as you can see the watercooled hybred, does offer a slight advantage over stock hsf, but is it worth it ?! In my opinion no. I will be writing this up in full in my site review but bascially to summarise: Pros: Looks Cool Unique Reduces temps slightly Cons: Hardly worth paying for anything more than intel hsf Does site in mount, wobbles around Huge magnet on top of CPU ??!!?!?!? Bloody noisy
yeah, its always reccomended to keep magnetic drive pumps(which produce loads more EMI than fans) away from components. I didnt think this would perform well at all, like mentioned a intel hsf can almost beat it(so a swifty could kill it i bet)and just adds the risk of a leak.
