|
|
General Liquid/Water Cooling Discussion For discussion about Full Cooling System kits, or general cooling topics. Keep specific cooling items like pumps, radiators, etc... in their specific forums. |
Thread Tools |
07-09-2004, 09:32 PM | #51 | |
Cooling Savant
Join Date: Jan 2004
Location: Alberta
Posts: 631
|
Quote:
Bill, on the issue of stuff that is tolerable in the "real world" and high end cooling, is Swiftech working on any thing with heatpipes? |
|
07-10-2004, 01:47 AM | #52 | |
Cooling Neophyte
Join Date: Mar 2004
Location: Denver, CO
Posts: 66
|
Quote:
One of the things that is easy to forget as a computing enthusiast is that most people, even in technical professions, haven't a clue how the stuff works. You've got hardware, software, and wetware in order of reliability, and any tweaks to the hardware requiring software monitoring for failsafes dramatically increase the chances for failures, and once you bring the wetware into it, failure is assured. Probably the toughest part of w/c is the potential for catastrophic or cascading failures. That alone would require it to be much more reliable than other hardware since otherwise it would be inducing failures of the other hardware. Hell, know why business desktops don't have blow holes? Because if you can spill on a $15 keyboard by golly you can spill into the blowhole of a $1500 system. Here's a twist on the thought of the thread. Imagine for the moment the article is symptomatic of the arrogant meglomania of the enthusiast community. Perhaps the problem isn't that there are crap products, but that innovation is being squelched by the enthusiast community's self-important rants about how many crap products there are. What is hurting w/c is enthusiasts that blast the "barely adequate" solutions and advocate high flow large tube systems with bigger pumps. It should be obvious with a little thought that reliabilty would be higher and maintenance lower on a well designed lower flow smaller system. The very products the enthusiast community loves to bash are the ones that are required to move w/c forward, we need to help them mature not ridicule them - build better minivans and economy cars instead of denigrating them because they aren't Formula-1 or NASCAR. |
|
07-10-2004, 04:23 AM | #53 |
Cooling Savant
Join Date: May 2004
Location: Portugal
Posts: 179
|
As for smaller pumps that fit in bays:
Timmer Coolpump Typ 700 12V DC powered by OASE Preiswerte und kompakte 12V-Pumpe mit bewährter OASE-Technik. Die Pumpe läuft ruhig und zuverlässig und kann einfach an einem freien Anschluss des PC-Netzteils angeschlossen werden. Technische Daten: 12V DC, ca. 5W, 50cm Kabel Förderleistung max. 8 Liter/min Anschlüsse: G1/4 Innengewinde or this one: compacttube für Eheim compact 600 With this one you can even see the water flow Fördermenge: 150-600 l/h (per Schieberegler einstellbar) Förderhöhe: 1,3 m Leistung: 11 W, 230 V |
07-10-2004, 05:04 AM | #54 | |
Cooling Neophyte
Join Date: Jun 2004
Location: Illinois
Posts: 60
|
Quote:
|
|
07-10-2004, 06:39 AM | #55 |
Cooling Savant
Join Date: May 2004
Location: Portugal
Posts: 179
|
LPorc : "It should be obvious with a little thought that reliabilty would be higher and maintenance lower on a well designed lower flow smaller system."
Well, I think that right now you don't have any reliability either in a low or in a high flow system. Of course you can have the same security in a big car as in a smaller car. It's just cost driven. So far, no one considered important to design a self-contained WC system, or having parts with MTBF. The industry is in a standstill as it is at a turnpoint. If Swiftech is to introduce such a specification (new pump) we will see more professional attitudes towards the evolution of this market. |
07-10-2004, 08:54 AM | #56 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
nice to see posts such as LPorc's, but agree with Jag that size per se does not preclude reliability
nice pumps Jag, is their mfgr 'into' reliability testing ? (and ISO 9002, that kinda stuff) AA go listen to a gear pump no product dev comments |
07-10-2004, 09:05 AM | #57 | |
Cooling Neophyte
Join Date: Mar 2004
Location: Madison
Posts: 99
|
Quote:
|
|
07-10-2004, 09:21 AM | #58 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
why are we ignoring the elephant in the refrigerator ?
look at the new Mac, every difference between that system and a DIY system reflects the application of fairly basic engineering - possibly considering reliability, cost, and performance in that order. and that Hitachi unit |
07-10-2004, 09:48 AM | #59 | |
Cooling Neophyte
Join Date: Jun 2004
Location: Illinois
Posts: 60
|
Quote:
Wrong, self-contained units do exist. The Aquagate being one of the most recent, with more hopefully on the way Strange that whenever a newbie asks about these units, there's at least one "pro" water-cooler who's happy to say that such products suck and that the noob should instead fab a system, from as many non-computer differentiated parts as possible. And this advice is usually offered without asking the newbie anything about his intentions for the unit, or anything else for that matter. Innovation doesn't work if few people buy things during a product cycle. And no, near as I can tell the industry isn't at a standstill OR a turning-point. What they are doing is casting about, trying to find a solution that will enchant the masses and make money. If you want niceties like MTBF ratings on WC products you’re going to have to wait until enough people are buying things, finding their weaknesses and complaining like crazy about them. |
|
07-10-2004, 11:15 AM | #60 | |
Cooling Savant
Join Date: Jan 2004
Location: Alberta
Posts: 631
|
Quote:
Anyway, I strongly believe that the reason that watercooling is being held back (and will be in the future) is heatpipes. Nothing more complicated than that. Heatpipe heatsink designs can be improved by a lot. Every heatpipe heatsink that I've seen has two TIM joints on each heatpipe, as well as the one from the processor to the heatsink. Why must that be there? Why has no one hollowed out a small hole in the bottom of a copper plate (evaporator stage) and made a heatpipe from that, or, hell, cast an aluminum base around heat pipes? That would probably drop the thermal resistance by 0.05C/W or something around there. Also, why not use a radiator rather than a heatsink? Get a 160*80mm radiator, attach pipes from the base and have the actual fluid flowing through it with a vacuum of the proper level inside of it? That should eliminate the pump and those TIM joints as well as expanding the surface area vastly... I think the reason we complain about poor units is because they don't offer anything. The Aquagate, when quiet has a thermal resistance of 0.31C/W... BARE minimum with a 3.06 or higher by Intel's standards. When turned fully up, it's 0.23C/W and about as loud as a Tornado... Where's the improvement? Why should we buy this? It costs $200 whereas an ALX 800 costs $15 and a Tornado costs $10. Hmm... Equal performance and less hassle for 1/8 the price? |
|
07-10-2004, 11:36 AM | #61 | |
Cooling Neophyte
Join Date: Mar 2004
Location: Denver, CO
Posts: 66
|
Quote:
But then again, wouldn't the same excellence of engineering concepts taken and applied to the smaller, lower pressure product also improve it? I.E. construction values provide different results for different loads and scales. I suppose if you start getting into problems where the smaller scale becomes more difficult or smaller enough that the properties change (going nano anyone?) there is an issue. |
|
07-10-2004, 12:04 PM | #62 | |
Cooling Savant
Join Date: May 2004
Location: Portugal
Posts: 179
|
Quote:
Right now for example you make a good? block, use a car radiator, an aquarium pump and presto, you're in business. Does that appeal to an organization? When I say standstill, I seems that everyone is watching the others, but at the same time convinced that it's all ok for now, as if the business will run by itself, like you post "wait until enough people are buying things". Only it's up to you to show the products you sell are ok, not waiting for something to happen, then react. When someone trys a more professional approach, the costumers will see a difference and they will show their preferences. This market will divide itself, and you either have the ability to compete or will be left behind. To BillA No info on that at the moment. Last edited by Jag; 07-10-2004 at 12:14 PM. |
|
07-10-2004, 12:14 PM | #63 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
deleted post
Last edited by BillA; 07-10-2004 at 12:15 PM. Reason: wrong thread |
07-10-2004, 01:22 PM | #64 |
Cooling Savant
Join Date: May 2004
Location: San Diego, CA
Posts: 202
|
Mass-market is already happening
What about the new Apple G5 computers? Apple I'm sure plans to sell at least tens of thousands of them; not a number to brush off. And they have water-cooled components out of the box. The Apple propellerheads must've done some very thorough testing on those things before taking that much of a financial risk.
I'm sure it works, and must have some benefit compared to air-cooling or they wouldn't have done it. Also, the notion that typical users really have no idea whats in their computers applies to Macintosh users more than anybody. I've said it before, and I'll say it again: If computers are going to keep getting faster than new cooling solutions will have to be introduced into the mainstream. The physics of shrinking chips just demands that...no way around it. So mass-market is going to happen with watercooling I think, probably after heatpipe HSF's are mainstream to the point of commoditization, which is happening now. As far as enthusiasts go, I'll return to the car analogy, there are drivers and there are hot-rodders. You see a lot of little Honda Civics out there, a "mainstream" user, but every now and then you see one painted lizard green, running NOX with a six-inch exhaust tip and a airplane wing glued to the trunk...thats a "enthusiast!" Take whats stock and make it go faster...thats what overclocking your computer is basically. Its already happened with air-coolers. Time was in 'puter world that using copper in your heatsink, or a 80mm fan, was extreme bleeding edge geekery. These days thats the stock Intel reference cooler. Crack open a Dell XPS and you will find a nicer HSF than anything that existed, even for the enthusiast, four years ago. The same thing will happen with water-cooling. This all assumes Moore's law holds true for a while on the process side, and it looks like there are snags cropping up for everybody in the business at 90 nanometers. So if Moore's law stops, I have no idea of the future regarding coolers. All bets are off! |
07-10-2004, 01:34 PM | #65 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
HAL-9000
did not (really ?) mean to ignore your question re origin may I gently suggest you buy one ? all will be known "and must have some benefit compared to air-cooling or they wouldn't have done it" possibly, that would seem to be the case though but it is also product differentiation, for marketing |
07-10-2004, 01:42 PM | #66 | |
Cooling Neophyte
Join Date: May 2004
Location: NC
Posts: 47
|
Quote:
The A64 top that you sent me already cracked I was having problems with air in the system I can home one day and the top of the cascade was cracked. I don't think it really matters how thick the acrylic is, it just cant handle pressure points. I didn't even have it that tight. I believe that for reliability an all copper block or brass top at the least is going to be necessary. I do like the idea of the swiftech block thats all sealed. Only problem is if it ever gets clogged it will be a pain to clean it out. As for budget watercooled systems, its funny you guys are talking about this as our company is trying to find a decent sub$200 system that is decent quality and performs well. I have been telling them these cheap kits are going to be more trouble than they are worth. I hope they listen. I really dont want to handle all the upset people with crappy performance or reliability, especially the people that have leaks and have to hear their warranty on their hardware isn't good because of water damage. I don't think watercooling is anywhere near the price-point vs performance/quality it needs to be for it to be sucessful in a normal consumer market. People are so cheap, I don't see them paying more than $100 for a watercooling kit unless they are a real enthus, which probably account for about 5% of the computer users now at best. Plus dealing with the fact that 90% of problems today are caused by the consumer and not the product, doesn't lend support to the idea that they can even handle a watercooling installation. Bottom line, if you can program your VCR you can probably install a basic watercooling setup, unfortunately that leaves about 80% of the population in the dark |
|
07-10-2004, 02:10 PM | #67 | |
Cooling Savant
Join Date: Jan 2004
Location: Alberta
Posts: 631
|
Quote:
infinity9: I think you should just advise them to avoid the whole water deal at all, unless they've actually got a good reason to go for them. Prescott heatsinks are nearly as good as water performance wise (well, close enough for more commercial tasks) much cheaper and safer as well. Bill, how does one get a MTBF of 50,000 hours when it's only been in development for 2 years? Also, is the big version the MCP 1200 that was referred to a long time ago (stated to be too expensive for the vast majority of users, the power wouldn't be worth the price) ? |
|
07-10-2004, 02:48 PM | #68 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
extrapolation of existing designs with long term data, shorter term testing to confirm correlation with other data
std stuff the term of price/performance is incomplete, add noise, size, and - of course - reliability slice it up as you wish, what a consumer does with each purchase and what the predominance of cheap crap sales is saying don't blame the suppliers (drug war come to mind); this country is quite fond of buying junk - and moving on, to buy some more not called The Cathedrals of Consumerism for nothing |
07-10-2004, 03:41 PM | #69 | |
Cooling Savant
Join Date: May 2004
Location: San Diego, CA
Posts: 202
|
I actually just upgraded...to an Asetek Antarctica
Quote:
(1) I found the Antarctica for $39.99, about the same price as the MCW6000 series goes for, at jab-tech.com. The only "catch" to such a cheap price is I had to screw the top on it myself...boo-hoo. So price wasn't an issue. (2)The Antarctica performs more or less the same as a MCW-6000 series, at least the 3/8" ID version. (3)I can take the Antarctica apart and clean it, which I can't do with the MCW-6000 series very well, as demonstrated by phaestus. (4)I like quick connects, and with the 22600 kit and the Asetek, everything in the cooling loop is quick connect., which use to be the deal with Swiftech all round. The MCW-6000 is the first Swifty block I have seen in some time that doesn't use quick connects. (5)The Asetek product is aesthetically more appealing than the Swiftech. The only drawback is the acrylic top, which I don't normally trust. But the Asetek's is unusually thick, and I haven't seen reviewer or user feedback regarding problems, so I took the dive with the Asetek after all was said and done. My previous block was a MCW-5002, which I may adapt for something else. I've always liked the MCW-5002, but the Swiftech doesn't cool as well as the latest WW clones such as the Asetek (or the MCW-6000 for that matter). Buying a Euro component was weird for me. I've almost always bought Swiftech stuff, and have always been a big fan of the company. I think I own around seven Swiftech HSF's of various flavors, including a pair of MC370's, which was the first enthusiast cooling product I ever bought, all those years ago in the Abit BP-6 Celery days (sniff, sniff). The MCP600 pump is the best pump currently out there for watercooling...period. Perhaps that C-Systems thing can give it some competition now. One of the reasons I've always liked Swiftech was that the stuff you buy from them is made on top tools by top people. Thats why I'm curious about the origin of the MCW-6000 series. As far as the Apple thing goes, I think they really needed WC'ing as an engineering solution more than marketing. After all, they only put it on their fastest G5, not the others, probably due to cost difference. Last edited by HAL-9000; 07-10-2004 at 03:56 PM. |
|
07-10-2004, 05:33 PM | #70 | |||||
Cooling Savant
Join Date: Jan 2004
Location: Alberta
Posts: 631
|
Quote:
Quote:
Quote:
Quote:
Quote:
|
|||||
07-10-2004, 05:44 PM | #71 |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
AA
cool your jets every consumer will evaluate a product's 'feature set' differently HAL-9000 perhaps I did not understand your question; I designed and prototyped the MCW6000/6002 right here in Signal Hill then helped work through the inevitable teething problems |
07-10-2004, 07:38 PM | #72 |
Cooling Savant
Join Date: May 2004
Location: Portugal
Posts: 179
|
Plot for Half-moon design
26-04-2003 = thread: http://forum.vapochill.com/showthrea...5&pagenumber=3 |
07-10-2004, 08:24 PM | #73 | |
Cooling Savant
Join Date: May 2004
Location: San Diego, CA
Posts: 202
|
Quote:
I like quick connects because I have become accustomed to the convenience with them in Swiftech parts ironically. Also, I tweak, I take stuff apart and put it back together again...a lot. For instance my current case is an Addtronics mid-tower, but in the very near future I am acquiring a Lian-Li V1100 and will have to transplant the whole water-rig to it, and re-cut all the tubing (when you see the inside of the V-1100, you realize it is the water-cooling case to have IMO). Quick connects will make that process easier for sure. And yeah, I have seen issues with acrylic tops on blocks before. I know folks who have crackes theirs after applying to much pressure etc. The only reason I posted what I did is because unregistered suggested I buy a MCW-6000, and replied that I had just bought a block. The MCW-6000 was the other "contender" for my WC dollar, and I chose the Antarctica for reasons I outlined in the post, that's all. And unregistered, to clarify my question I am curious where the blocks are actually manufactured. I am sure all kinds of prototyping goes on at Signal Hills that you can't talk about (damn!). But every waterblock or heatsink on Swifty's website except the MCW-6000 stuff has a little "Made in USA" icon on their respective product pages. So I am curious where they are manufactured. Another reason I am curious is the only outfit I have ever seen make a forged pin-heatsink is Alpha Novatech with their Microforging process. Does Alpha have a hand in making MCW-600x's? Where are the production units made? |
|
07-10-2004, 09:08 PM | #74 | |
CoolingWorks Tech Guy Formerly "Unregistered"
Join Date: Dec 2000
Location: Posts: 2,371.493,106
Posts: 4,440
|
Quote:
|
|
07-10-2004, 11:00 PM | #75 | |
Cooling Savant
Join Date: May 2004
Location: San Diego, CA
Posts: 202
|
Quote:
|
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|