page 1 of 3
This is going to be a story of a love-hate relationship between the reviewer and a rather exceptional product, VapoChill.
Basically, VapoChill is a mini-Air Conditioner for your CPU. In the first stage, a compressor pumps pressurized liquid freon to a vaporizer (a copper coldplate) mounted on the CPU, which then absorbs heat from the CPU, thus cooling it.
Contact with this heat source brings about a boiling evaporation of the freon coolant which starts the second stage of the cycle, wherein the coolant, now heated into a gas, passes through a condenser (a radiator cooled by a 120mm fan). Leaving the radiator/condenser, it travels back to the compressor and begins the cycle anew. All of this comes under the heading of, “The Vapochill, Vapor Phase Refrigeration System.”
The term “Vapor Phase”, of course, refers to the freon’s two states – liquid and gas. VapoChill claims a -15 C to -20 C operating temperature for this unit. The VapoChill comes with three other features worthy of mention:
- The more interesting and unusual of the three is a HEATING element which attaches to the processor just above the copper tines which insert into the motherboard. This is to prevent condensation.
A novel approach to say the least, and not just a little responsible for very high system temps we will talk about later. I feel that properly dealt with this feature dose no harm and likely does much good.
- The Second is a “PCB” which holds the system in “Reset” while booting and so allows CPU temps to drop to acceptable levels before attempting to boot to High FSB speeds.
This turns out not to be as useful as one might think. This is because The ASUS motherboard- and probably others, requires a “stepping” double boot.
For those who don’t know: This is where one is unable to attain one’s highest stable FSB setting by just booting to it. One first must boot to a lower FSB, then enter windows and wait for Peltier or Vapochill temps to drop then restart your computer, reenter your BIOS and go for your highest settings.
This occurs in systems which don’t use Super-Cooling as well. But there one does not need to enter Windows before re-entering the BIOS. You can simply wait until just before Windows starts to load at startup, hit the restart button and re-enter the BIOS at that point.
- The third point of interest – the VapoChill will work with just about every CPU out there. VapoChill provides upgrade kits should you wish to switch processors. All-in-all, a very versatile cooling system. Asetek is currently working on a VapoChill coldplate which will run on 370 motherboards.
All of this comes neatly packaged in a computer case from HELL!
I’ll have a good deal more to say about this later. For now, I will try to tell the story of my encounter as it unfolded:
When I first contacted Minotaur, the US Distributor for Asetek,
the Danish company who manufactures VapoChill, I received a warm response. They went so far as to state that they were about to contact me and ask if I’d like to review a VapoChill! (God I love this job!)
I called Joe up and told him of my of my plans and as always asked for his advice. After a lengthly discussion we decided to buy the Asetek with an 800 MHz processor to go with it. We also decided to have Minotaur install the 800 MHz Pentium FC-PGA CPU and Asus Slotket into the clamshell case which houses and insulates the CPU from condensation.
We did this as installation can play a big role in how everything from a stock fan to a water-cooled Peltier operates and cools. The VapoChill is no exception.
This would also allow Minotaur to take it’s best shot at protecting the new CPU from condensation. Minotaur was eager to accept the challenge. I also sent Minotaur my ASUS P3V4X mother board, wanting to give it a second chance at running chips above 700 MHz, in a super-cooled and highly overclocked environment. Unfortunately history repeated itself (see my Swiftech articles); the P3V4X Could not handle the AGP at anything above 1040 MHz.
However, this proved to be inconclusive. This is due to the fact that this particular 800 MHz chip was later found to be limited in its overclocking ability in an ASUS P3B-F motherboard as well. Something I believe has nothing at all to do with the VapoChill Cooling System.
But rather, this particular 800 MHz chip; chips are simply not created equal! This tends to be the case more and more as a given CPU, say the 700 or 800 MHz, ages. Intel gets a more exact fix on what a chip needs to be able to do, to be completely reliable, and begins reducing the overclocking overhead of the chips it sells.
A chip that can perform reliably is moved up a bin. Not just my opinion, several system makers have told me this – but in the realm of opinion nonetheless.
One week later I found a large box on my door step! The case was fairly attractive with five 5 1/4 in. drive bays and four 3 1/2 in. drive bays. Sad to say three of the large and one of the small drive bays are taken up by the VapoChill unit.
But of course I knew beforehand that this was to be the case “:O} Before I started her up, I’d have to install my Motherboard, RAM and peripheral cards. So I got to it! I sat down and read the manual for any quirks this new techno-application might have. I read through the case docs in a flash (a single sheet of paper, half of which was in Danish).
I went back to my work area (a converted-to-my-use-parlor) and pulled the front cover off the case. It was a bit tight fitting, but hey, it was new. I got the bottom half off and gave another pull and off she came! I’d just broken every-single-connection- between the case and the cover! It rained plastic bits and pieces! Not my happiest moment in computing. What went wrong! I took a closer look.
Unlike any case I’d ever owned, this cover was meant to slide up then pull away. There’s no outward sign of this and no mention of it in the Docs. The only way to see this arrangement is to remove the front cover, which must be done in order to remove the case cover!
Ah! The case cover: A one piece top and sides arrangement which slides back and then lifts off….part way, until the side latches (made of extremely thin and weak aluminum strips glued on) snag the inside top rail, stopping you cold.
One has to spring the bottom of the case cover’s sides outward and tilt the cover to clear the vertical latching strip of the inside case rails. Let me just say putting the case back on is even more fun. Did I mention sharp edges every where to increase your delight?
AND that’s just the case’s cover! Tyson at Minotaur went to some pains taping a rubber covering strip along the more prominent edges inside the case – for which he has my thanks! But he must have spent hours covering them all.
page 2 of 3
Ever hear of a tool-less case? Well, VapoChill has not. All drives must be screwed into place – no snap in slide rails for your drives. To accommodate the cooling unit above, the Power Supply Unit is mounted in front of the CPU, covering the motherboard.
Though the CPU and RAM slots remain in the clear, you must remove the power supply to access the CPU or remove the motherboard. AH! The motherboard! This case designer was not content with the usual case pass through for your motherboard hookups – Mouse, keyboard, printer etc. Nope.
This fine case comes with its own backplate which requires one to remove every single securing screw/bolt from the mother board in order to fit the mother board connectors through the back plate slots, then re-screw them through tiny holes drilled for this purpose in the backplate. I guess they anticipated motherboards trying to escape from such a case.
This honey comes without a single case fan, not one! So much for case cooling. To be honest, I have a very hard time understanding VapoChill’s thinking on this. Do they believe that the CPU is the only component in need of cooling?
Especially in a case housing an advanced super-cooling unit whose sole purpose for being is extreme overclocking. Case temps at 73 F room temp? 36 C!! No wonder their truly fine cooling unit pulls only -15 C at idle and +5 to -5 C under load.
The System was every bit as quite (37 dB) as everyone said it would be – not surprising! Not a breath of fresh air was stirring! While I’m on the subject of the inexplicable: The cooling unit itself doesn’t exhaust its radiator’s hot air out the back of the case, as it sits some five or so inches from the case’s rear,
where two series of perforated holes form two round indentations, one above the other, for two 80mm fans. Unfortunately the are no holes drilled to accommodate the mounting of these non-existent fans.
Tyson does try – he drilled and mounted a fan to exhaust some of the hot air from the radiator.
Now for something really different, so strange Tyson couldn’t believe it until he had checked what I told him against another unit that had just arrived:
VapoChill setup their radiator fan to exhaust away from the back of the machine into the congested area surrounding the compressor near the front of the machine.
I was in anguish! Both Jon and Tyson (brothers) bent over backwards in helping me acquire and setup this unit. I told them the type of review I’d have to write and how sorry I was this was so.
Both men told me to tell it like it is! (This I have tried to do.)
THEN they asked me to try and make it better! To help them improve a product they had faith in but realize was still in development. Well HELL – somebody had to do it”:O}
I ordered two 120mm Sunon fans and cut some holes. First, I removed everything from the case except the cooling unit, which I mummified in saran wrap and wet-damp towels.
Out came my grinder. I drew a 5 1/2″ inch hole in the front bottom of the case and cut it out. I wanted to use a 5 1/2 hole saw, but at $50.00 for the bit, I couldn’t afford one. So I ground away.
I’d let my fans cover my sins. Then I cut out the two 80mm fan indents in the top rear of the case, turned the existing 120mm radiator around so that it blew through the radiator and out toward the back of the case.
I had a dead 120mm fan (wrongly placed fingers make short work of these babies) from which I removed the two remaining blades and motor. This was attached to the inside of the case opposite the radiator (Super-glue to tack and Shoe Goo to attach permanently).
I positioned it there so that the external fan mounted upon the other side of the hole would suck air from closer to the radiator exhaust.
Then the bottom front fan was installed.
To this was attached an extended scoop carved from a CoolWhip plastic container – seemed appropriate at the time. I did so because the front plastic cover is two layered – the top layer is at an angle creating a air pocket, a real source of noise! This Picture doesn’t show the scoop – ran out of film.
page 3 of 3
That Power Supply had to go! It completely blocked the free circulation of air throughout the case. But where? Then it dawned on me: It was about the same as the water reservoir in my water cooled Peltier set up! And where was my reservoir installed?
Why behind the drive bays against the back wall of the case. I tried it for fit. Serendipity! I simply built up several layers of Velcro striping beneath it and such was the fit that this held it up out of the way.
But how to power it? However clever its location, I still needed to get a plug to it. No way could a simple hole be cut near the front, besides the fact of the one piece case cover preventing this, the cord would be laying all over my work bench.
Then I noticed a small rat hole sized perforation in the bottom rear of the case. I grabbed onto my tin snips and hacked out a plug sized hole to snake my power cord though to the front of the case.
Later, I velcro’d it to the front side on the bottom of the case, then I used Shoe Goo to surface the rough edges of the hole to prevent any electrical mishap.
Now what about that BX chip and my RAM? Using Velcro, a small CPU fan was attached to the side of the CPU Clamshell blowing air directly upon them. Video card Ram? I Velcro’d a smaller fan to the bottom of my drive bay so half its output flowed across the top and half across the bottom of my card.
I left two PCI slot covers off to accommodate exhaust.
So was it worth it? What did all this buy me?
Well, overclocking my 800 MHz CPU to 1120 MHz at a V-core of 2.0 volts, I dropped 14 C in system temps with an ambient room temperature of 22 C. CPU temps went from between +6 to -6 C to -3 to -16 C under 100% Seti@home load. CPU temps dropped from -15 to -26 C at idle.
Along the way I “rounded” my ribbon cables and placed them in electrical conduit (Black tubes seen in the pictures) and did what I could to wrap my-now-too-long power supply wires up and out of the main air flow.
But cooling came at a price!
This was one loud, obnoxiously loud, computer. About the time I realized that a computer that seemed loud from the next room (20 feet away) was indeed too loud, Joe saved me once again!
There on our front page was a brief little ditty on installing a rheostat to slow and thus quiet fans. After installing them, the noise level dropped to a sensible 51 dB on “A” weighting and 56 dB on “C” weighting (measurements taken at 18 inches from the front of the machine). Of course, this is a bit louder than Asetek’s specified 37 dB (I was unable to check this with my Radio Shack Sound level Meter as the quietest it can measure is 50dB, but at three feet away you could hear nothing”:O})
I would like to add that slowing my 120mm fans and thus quieting them down cost me nothing in cooling performance. I found that on hotter days, more fan speed and therefore louder fans are required, but at 70F room temps, they are slightly louder than say a stock Compaq or Dell computer.
This brought my modifications to, what was for me at least, a successful conclusion.
However I experienced a great deal of frustration in the process. First and foremost of which was the fact that such a really, very fine cooling solution should be housed in an awkward and arguably dangerous case.
Furthermore, this case offered NO inherent cooling features and was so poorly thought out and designed. Yet, The VapoChill unit uses a fixed length copper tube to connect to the CPU, making a replacement case difficult to find and hard to satisfactorily modify.
This is something Asetek will have to accomplish for us. The consumer lacks the resources to try a dozen or more cases in order to find one worthy of the VapoChill unit. I understand the difficulties and expenses involved in seeking out a better solution. TOUGH!! That is no reason to STICK-IT to the consumer. Has Asetek considered the countless hours of dissatisfaction multiplied by the number of units sold?
I’m Told by Minotaur that Asetek has, or will, implement some improvements to this case. A 120mm fan…hole…no fan! And a special sharp edge coating paint to reduce the number of small cuts. And perhaps a re-positioning of the cooling unit so that it can exhaust it’s hot air out the back. However, as seen from the pictures of the “New Case” supplied by Minotaur, only the protective paint has been implemented at this time.
I’m sorry, but this seems to me like making a silk purse out of a sow’s ear! Thisis especially exasperating considering the purse contains real gold!
No amount of touching up will ever make this case worth owning.
Asetek, as a customer who really appreciates what you have done right by delivering a truly excellent cooling product, I ask you to right what you have done wrong. Forget all that stuff about price points. Your customers are overclockers – who else even thinks about cooling?
Overclockers seek access to their cases on a daily basis. Every time I have to open this box, I ask myself “How could a company that can produce such a fine and efficient a unit that it can run on a 300 watt power supply not know they are hurting their product with this case?”
I expect to hear from VapoChill customers. I expect some will come to your defense, but many more will echo my complaints. I will post those responses, pro and con.
“Is this case so bad I really should stay away from buying a VapoChill?” That’s not a question for me to answer. But would I buy this unit again?
YES!! I would; but I’d gripe all the way to 1120 MHz while doing so. I’d gladly pay an extra hundred bucks for a really decent case. If VapoChill can’t swing one on that budget, it’s doing something wrong!
In Part 2 I will run bench marks and compare VapoChill to the finest water-cooler Peltier setup I have had the privilege to use (OK so it’s the only water-cooled Peltier setup I’ve ever used – it’s still one helluva cooler!).
I would like to thank Minotaur for their exceptionally kind and helpful treatment of me while I struggled to understand and properly use this new system. But most especially, their honest and forthright answers to all my questions. Not once did they shirk or try to side-step any of the issues I’ve raised in this review. Jon, Tyson, You were as good as your word…AND THEN SOME!
You can reach me at my E-mail address or in the Forum. I welcome your views – diversity and opposing opinions are as welcome as agreement and quite often more helpful.
Email from Asetek:
“Hi Dan,
We have read your review with great interest and we have nothing against critics; it’s the best way to improve our products. But just as you are disappointed in some areas with the VapoChill I must say that we are twice as disappointed that a “serious” review site doesn’t even bother to contact the manufacturer of the hardware before drawing any conclusions. I know you have contacted Minotaur, but of course Asetek knows more about future plans, R&D and changes than a distributor.
As a serious guy you should have contacted us and mentioned the details, and then we WOULD have had a fair chance of answering or at least comment these things BEFORE you wrote all your self made conclusions. I’m not asking to read your review in advance or anything, but I think it’s common ethics to ask a manufacturer for future plans, comments etc. At least our other reviewers has done so, and if you take a look on leading hardware sites, it’s common business to get comments from the manufacturer and include them in a review.
Obviously you don’t have any ideas or information about why we use this case. So I will try to explain things!
The current case is actually not a cheap case, and as you wrongly describe the case manual is not in Danish neither the half of it. Together with the case and separated from the manual – we know most our customers doesn’t read the manual – we have placed a sheet of paper describing in BOTH English and German how to disassemble the case. This is really an easy job and no tools are required, and we simply don’t get how you can destroy the plastic clips. You are sure you did read the piece of paper, or did you just start to pull the plastic clips? Anyway we have not had one single customer complaint about this.
Regarding screws etc. we have chosen a case where everything has to be connected with screws with purpose as the VapoChill is designed to last several years, our experience tells us that screw connections last longer. (BTW. have you ever considered that mounting a sliding device on a CD-ROM-drive requires four screws – just the same as mounting the CD-ROM drive directly in the case. Of course with a sliding device you can remove things, but honestly how often do you do that after all?)
As for the lack of cooling fans in the case, again it’s a choice from our side. Of course we could have provided extra cooling fans for the case, but then the product would be more expensive (and many users don’t need them) and as you correctly describes the current customers are Overclockers, and honestly speaking we cannot even satisfy 5% if we pre mounted fans. We have done that before, but then people complained about that they wanted other fans, or they already had the fans from earlier setups etc. Therefore we have made the choice that our customers can decide for themselves which fans etc. to use.
And now to the heart of your comments! The PSU is mounted vertically in order to get space for the hoses containing the coolant. You write that you have to remove the PSU in order to remove the CPU. And so what? I mean if you want to change your CPU in a VapoChill you will need screwdrivers anyway to remove the CPU insulation kit. We really can’t see the problem in removing four screws, as you still have access to AGP/PCI-cards, HDD connectors and memory with the PSU mounted.
This is also why we don’t use another case, as in all cases (we know of) the PSU is mounted horizontally why there is no room for our cooling unit. Then you can say that we could have made the same solution as our competitor, but firstly our philosophy is, that copying others ideas will not bring us anywhere, and secondly we don’t like the idea of having a second case only for the cooling system.
Has it ever occurred to you, that we actually have thought about this??? I can tell you we have! As you describe it’s NOT possible to make a hose for refrigeration, which is stretchable and or more flexible than the one we use today. May I point out, that at least until now our competitor has used stiff copper tubes, with has absolutely no flexibility at all. In air conditions in cars today polymer rubber hoses are used with the result that the coolant is diffusing through the hoses, why the system has to be refilled periodically.
As we don’t want to place the bother on our customers to refill the system (which requires an authorized AC technician why it’s expensive) we have made a system which is 100% hermetic sealed, and STILL he have managed to develop a flexible stainless steel hose. Actually it’s quite an achievement! And this is where we lose a lot of respect for overclockers.com as instead of just drawing your own conclusions (as you obviously don’t know much about refrigeration ) you should have contacted us and asked if we had any comments on this.
Unfortunately this also means that we cannot just go ahead and grab a case on the shelf and mount our cooling device and then bend, stretch and fit the hose to the CPU but rather use a special designed case. If you have any good ideas about this, which the whole AC industry hasn’t been able to develop, I’m very interested to hear more about it, as we would implement it right away!
The conclusion to this is that between other things, this is a question about physics, which cannot be changed on the fly. Then you can ask yourself “Is this attitude really good enough?” Of course it’s not, and that’s why we have worked on an improved model of our current case, which will provide better airflow both in the bottom and in the top, one more bay for CD-ROM drives and some other details. If you had bothered to ask, you could have received material about this under *NDA, just like other reviewers have done.
Has it ever occurred to you that developing, drawing, testing and manufacturing of a new case takes time? Actually we have worked on the steel part of the case for more than half a year, it’s soon to be ready, but the front plastic bezel will not be ready before Christmas. Anyway we will start to ship the new case (the steel part) with the old front plastic bezel and then when the molding tools are ready offer our customers an inexpensive upgrade. And as you describe in your review we have already started to do a special coating on all steel parts of the case, which removes 100% of the sharp edges whatsoever.
If we wanted to we could have developed a whole new case, this is really expensive and the place to put the costs would be on the customers. Our unit is expensive as it is, and we don’t want to make it even more expensive. You might not agree on this, but our customer analysis made by overclockers and submitted by overclockers don’t want higher prices.
We hope after you have read this mail you will re-consider you future strategy about reviews, and post this mail partly or fully on your site.
Thanks,
Per Adamsen R&D manager,
André S. Eriksen CEO”
*For those of you who do not know, “NDA” means “Non Disclosure Agreement”. This is an agreement between two parties that specifically states that information shared between the two is priveleged and not to be disclosed unless certain conditions are met; for example, a release date.
Joe Citarella’s comments on Asetek’s email:
Andre raises a number of points that go to the heart of how Overclockers.com conducts its reviews. I want to make our approach crystal-clear:
We are not here to be sales assistants to the manufacturers. We are here for YOU, not them.
We write from the perspective of someone buying an item off the shelf, because that’s what YOU do.
We don’t let manufacturers co-author or guide us in our reviews. We find THAT unethical. The product should speak for itself, as it is, now. If you buy the product today, you don’t care what they’ll do with it in six months. You want to know what it’s like NOW, without special help.
We pull no punches around here. We write things as we see them. YOU don’t get special privileges like signing NDAs or special access to engineers when you buy the product.
While I fully understand the choices Asetek made, we have to judge the product based on whether or not it is likely to satisfy you. Asetek may think moving a power supply to change a CPU is a trifle not even worth mentioning; we don’t feel the same way.
We are more than willing to print additional material following a review, as we are doing here, but if YOU bought this product as we did, I doubt that understanding why things were done would seriously change your perception of the product.
This doesn’t make us too popular with manufacturers. We don’t regularly get handed goodies from manufacturers; a lot of what we review, we buy, as we did with the Vapochill. We are fortunate that the advertising revenues we receive from annoying ad banners are enough now to allow Overclockers.com to selectively purchase items off the shelf and test them.
We do not receive monetary compensation or retainer fees from any manufacturer and never will. That’s too high a price for YOU to pay in “let’s not offend the manufacturer” reviews. The four principal writers get up every morning and can look at themselves in the mirror without any problems AND THAT WILL NEVER CHANGE.
Email from Bob:
“I don’t see the purpose of buying anything like the VapoChill…what are you going to do with it? Being that AMD has finally slammed the door on overclocking, and you know Intel is sure to follow very soon, there will be no overclocking beyond jacking up the FSB speed. You will just have one big overpiced heat sink. Yeah, jacking the FSB speed up will produce a little more speed, and a little more heat, but nothing that a good 40 dollar Alpha copper heat sink wouldn’t fix right up.
Why should we (anyone) spend the money on this type of equipment knowing that they are sure to become usless in the coming months? I for one am not going to shell out 500+ bucks for something that I know is not going to be useful in just 2 or 3 more months. I think that this issue, FIRST AND FOREMOST, should be the one dealt with by the overclocking community as a whole.
Have YOU given any thought as to what you are going to do? Even getting the unlocked CPU’s, the next generation CPU’s are going to quickly outperform today’s extremely overclocked CPU’s. Seriously, this site is going to take a major downward turn with this new generation of locked CPU’s – there isn’t going to be much to write about. I would like to get your take on this subject, and please feel free (I insist) you forward this e-mail to the other members of Overclockers.com and see what they have to say about it.
Best Wishes
Bob – “RA3 for Q3A is DA BOMB!”
AKA’s VonKahn, Sizziling_Dung, The_DeCrapatator, Freddie_FuddFu**er
P.S. — Not that I am talking bad about your site or anything…cause this is where my browser calls home 😀 I love this site, but I think that this question(s) need to be pointed out. Please feel free to use this as you see fit.”
Email from Jesse:
“I just read your review of the VapoChill and I am in total agreement with you. I would gladly pay extra for a better case…if I can come up with the initial arm and leg for the stock VapoChill, $100 more would not be a problem.
I have been saving for the VapoChill for a while; as a result, I have been reading everything I could about the unit. OcShootout mounted their VapoChill in a modified Addtronics 7896A case. They promised a how-to guide but I have yet to see one. If you still have your VapoChill….then, could you PLEASE TRY TO MOUNT YOUR UNIT IN THE ADDTRONICS 7896A CASE!!! I know if it is possible you can do it.
By the way, if you cannot find the Addtronics 7896A, the Supermicro 750 or 760 should be the same. Addtronics makes it for Supermicro. I am begging please try man…if I am going to shell out this kind of money I don’t want to be stuck with a crappy case.
Thanks
Email from Sander Sassen at Hardware Central:
Hi Dan,
I agree! Sent mine back without even reviewing it!
Sander Sassen – Siteleader at HardwareCentral
Visit us HERE.
Email from Mark:
Man that was a good article! Thanks for the honesty, and telling the
facts! Maybe they will use another case from now on?
Email from Michael:
The BigTower of the Asetek System is made by this company:
Schaefer
maybe it is this one:
It’s a German company located in Dresden (where the copper-Thunderbirds come from)
Just to let you know… Bullfrag
For a case I chose the
Supermicro 760-A.
The only major modification required is enlarging an opening in the top of the chassis. This is necessary to slide the compressor into the case, since it will not fit unless the entire frame is completely disassembled. A jig saw and high quality metal cutting blade make easy work of the task, but it does leave some sharps edges to contend with.
Mounting a power supply is a bit awkward, but there’s plenty of room to work with. I found a couple existing holes in the upper rear lined up with the screws in my power supply. If you decide to use the power supply that comes with this case you may find it more difficult to mount, as it is rather large due to the redundant cooling. I found the 400 watt Emac to be a good fit, while offering plenty of power.
The 760A comes with several small Sunon fans, but I elected to cut two holes in the case – one on the bottom left side for intake, and another on the top panel for exhaust. A pair of 120mm Nidec fans from a Liteon FS020 move plenty of air, without sounding like an F-15 taking off. I also used the drive bay vents that came with the original unit.
The only major problem is mounting hard drives. The 760A is really disappointing in this respect. I believe it only has one bracket for two drives (what were the folks at Supermicro thinking when they designed a server case with room for only two drives??). When I did the conversion I was running a four drive array, with a fifth drive for my OS’s. To fit all of this I ended up using the brackets from the Liteon case. This is the only area where you may have to do some tinkering to find a solution that works for you.
I took a few pics of the case. As you can see I wasn’t after the
best cosmetic look – I just wanted to get the cooling unit moved because I was drooling to see how well I could overclock with it 🙂 Right now the case is pretty well gutted but you can still see how the compressor fits.
I should also mention that I modified the CPU mounting hardware a tad. It came with their latest enclosure, but I’m sure you’ll agree that mounting the CPU to the cold plate, while ensuring that you make even contact, is a pain.
Instead of running the screws through the back of the PCB into that plate, I first glued the metal plate to the copper evaporator, then glued the screws to the plate. This way all you do is slip
the CPU over the screw posts and you have both hands free to put even
pressure on the slug while tightening the nuts down on the back of the PCB.
I also found out that high quality thermal grease is a must. Some compounds contain water based silicon and the water freezes every time you run the system. When this small amount of water freezes it pushes the cpu away from the cold plate, resulting in uneven cooling.
Everything looks fine at idle, but under a load the CPU won’t be cooled properly, and ultimately you can’t overclock as much. Oh and I guess you’ve heard the case tilt trick by now, but Asetek doesn’t approve of the lower temps it creates simply because they’re not sure if the current insulation is adequate.
I bought some silver compound from a european store – global cool I think it was called. I later found out that coolingstore.com has some called “arctic silver” that supposedly works pretty well. A guy in Sweden read my complaint about Vapochill temps gradually rising over a period of time and pointed out that water based compound can cause the expansion problem I mentioned before.
In my case, I would install a P3 800 and it would overclock to 1120 no problem – then over a period of about a week the average temps at idle and under load would slowly rise by as much as 20 F – and eventually 1120 was no longer stable. I would take the clamshell apart and remount the CPU with fresh compound and it would be back to normal for another week or so.
This guy said cheap grease will expand when it freezes and slowly pry the slug away, as well as working it’s way out of the contact area. A weeks worth of on/off freeze/thaw cycles and the CPU is no longer being cooled properly. Asetek agreed that high quality grease is necessary and they’re working on offering the “good stuff” in future kits.
I noticed in my dealings with them that they go on the defensive, rather than take criticism constructively. I told them before that obviously this is a product in development, and they should welcome criticism as a chance to improve the unit. I suspect if they made some good changes they could give Kryotech a run for their money.
I’d actually like to see that happen, because I had a big argument with Kathy Hemby (Kryotech customer relations) about a year ago. She said the Vapochill would never be comparable to their units. I think they’re missing the boat not offering a more flexible solution, not to mention being able to buy a bare bones kit. Well at any rate, let me know where their response is – should be interesting 🙂
Asetek gave a me a bunch of s**t when I reported my problems. They basically accused me of being incompetent. I was like – “Look here, I’ve been building computers for years and I have plenty of experience with not only mounting CPU’s into all sorts of cooling setups but actually making them from scratch. Not only that I build all kinds of stuff, so I have plenty of mechanical ability.”
Their response – “We have heard many call themselves ‘Pros’ but we have found that 99% of the time the problem was due to incorrect assembly, and not a design or component failure.”
I’ll be glad to do a write up when I mount a new CPU. I’ll have to gripe about the cheap self tapping screws they supply to close the two halves of the clamshell. It takes about 200 pounds of pressure to keep your screwdriver from slipping off the head and stripping the screws, all while making sure you don’t pinch any foam or wires and praying your hand doesn’t slip inside the tight confines of the razor case – but I digress 🙂
That’s all the relevant info I can think of. Please don’t hesitate to drop me a line if you have any questions.
Dan,
As you may already know from some of my posts in the forum, I recently purchased a VapoChill too. I will have to say that I share some of your experiences with you:
I will have to agree with you that the case is a complete HORROR SHOW! It was by far, the most difficult case to take apart that I have ever seen, although I managed to do so without breaking anything.
I won’t even talk about blood loss.
True user-end torture at its best!
The other really bad part of the design that you may not have been through yet is you have to remove all of your CD, DVD, and floppy drives just to get the plastic front off again! Living hell.
Also, did you notice NO RESET BUTTON!
The “Chill Control” system would release the reset no lower than -1.5C no matter if the jumpers are set to -16C or not. Tyson at Minotaur tells me they are aware of this problem.
My fan also spins the wrong direction, in my opinion (the same as yours).
I agree that to take this unit from a truly wonderful PROJECT to a truly wonderful PRODUCT, it will definitely require some extensive re-engineering of the case for ease of use, and more drive bays.
On the other hand, I believe that, for the price, it’s a good value. I can run my PIII-800@ 1152mhz all day long, stable without worry of pumps, fried Peltiers or condensation.
VapoChill W/ChillControl & Slot-1 capsule – $780.00
PIII-800 ————————————————- $242.00 PriceWatch
Total——————————————————$982.00-1152mhz
Versus:
PIII-1000mhz, bare————————————$1080 – 1000mhz
No rocket science there, and it’s upgradeable.
Even when booting for the first time of the day at -1.5C, my PIII-800 on a P3V4X starts up at 135FSB, 1080MHZ, every time – no problem.
I have found the same thing you did, in that the P3V4X needs to be “stepped” to get higher speeds. I’m not sure if its the chill control releasing the suspend at -1.5C, rather than the -16C like I would like it to, or an inherent problem with the mobo. I tend to believe it’s the latter.
Since I had already read a few reviews on the Asetek and the limitations of the case, I bought it planning on putting it into my SuperMicro full tower case from the beginning. Now that I have one, I’m in the process of building my own version of it using a different refrigerant and compressor combination, and running dual
processors at -45C.
“Stay Tuned……. Film at Eleven.”
Over all, I am happy with my purchase, and being an overclocker, if I didn’t have to tinker with it I probably wouldn’t have found it so interesting.
Be the first to comment