- Norwegian version at Overklokking.no. Prometeia Phase Change Cooling Part II here.
I’ve been into overclocking for many, many years, and I made my first watercooler back in 2000 -- a fearsome beast that soon got nicknamed “The FrankenCooler” by our readers, featuring an all homemade waterblock, two 90W high performance peltier elements, a Ford Fiesta –84 heatercore, a 240W separate variable powersupply for the TECs, insulation to combat condensation and what not. All this to cool off a Celeron 466, as I’ve always been a “if you’re gonna do it, you’re gonna do it right” kinda guy.
The preparations, the planning and the making of this cooler took many months, and to me, that was part of the fun -- especially after finding out that this cooler made my trusty old Celeron 466 the fastest Celeron 466 on earth at that time, running at an impressive 735MHz on an Abit BE6 II at 2.5V core. On the other hand, it was big, bulky, and the whole thing turned into what one of my lecturers used to refer to as "optical noise". Surely I made quite a few refinements after that, but with limited access to tools, it still had a bit of a ghetto look.
The homemade waterblock with two 90W peltier elements
Things have changed drastically over the last three years, and today, you can pick up a complete watercooler even more powerful than my homemade for a few hundred bucks. Then there are of course the even more exotic and sophisticated coolers using phase change cooling, like the VapoChill from Astek and Prometeia from Chip Con, both Danish companies specializing in thermal acceleration products.
The Prometeia/LianLi case
Today, we’ll take a closer look at the Prometeia phasechange cooling unit from Chip Con, one of the most powerful, if not the most powerful cooling solution available on the market.
In part one of this review, we’ll take a close look at the cooler's construction, supercooling theory, and the LianLi kit that is an optional accessory for this cooler as well as discuss a few issues we ran into while adding the LianLi kit to the basecooler.
Phase change cooling is an interesting concept -- A phase change is a change of the state of matter, like water to steam (or ice), or amorphous (non-crystalline) to crystalline, or vice versa. But how can this change of state be used for cooling, you might ask.
Vaporization requires the input of heat energy. Our bodies use this as a mechanism to remove excess heat from ourselves. We sweat, and its evaporation requires heat input, i.e. the excess heat from ourselves.
Diagram showing the three states of a matter
The Prometeia use the evaporation of Freon, and more specifically Freon r134a, to remove heat from the processor, where the Freon is condensed in the condenser in a process, which releases heat energy.
The four-step cooling cycle
There are four steps in the cooling cycle of a Freon compressor based solution -- compressor, condenser, capillary tube, and finally the evaporator that sits on top of the processor.
The four-step cooling cycle
What actually happens is that the gaseous Freon is sucked from the evaporator, into the compressor. The compressor increase the pressure to the condensation pressure of Freon, and the phasechange takes place –- the Freon changes state from gas to liquid.
The condenser, an inline radiator, gets rid of the heat generated by the evaporator and during the compression of the Freon. By now, the Freon is completely liquefied.
The capillary tube is a small diameter copper tube acting as a valve to lower the pressure. This capillary tube is important for the refrigerant to evaporate at such low temperatures.
When the refrigerant has passed through the capillary tube, it reaches the evaporator. The pressurized liquid expands inside the evaporator and changes state from liquid to gas. Now, remember what I said about vaporization requiring input of heat. The vaporization process inside the evaporator “consumes” the energy from the processor, and the processor is cooled down.
The Prometeia comes in two flavours – the whole barebone system, and basecooler only. As we knew Chip-Con was about to release a LianLi kit for their basecooler, we opted for the latter solution. We all know the formidable quality of the LianLi cases, so that aspect will not be covered in this review.
Chip-Con's aluminium add-on kit comes with all necessary parts, detailed instructions for assembly, as well as a modification template for a LianLi. In addition to fitting the LianLi dimensions, the Prometeia add-on kit also fits the design of the following Lian-Li cases:
- PC-60, -61, -65, -65B, -82, -6085A and -6085B.
The Prometeia add-on kit also fits the dimensions, although not the design of the following models:
- PC-68, -6100, -7, -0703, -0700, -6083, -6087, -6089 and -6099.
We chose the LianLi PC-60 for this review.
The Prometeia cooler as it came from ChipCon
This is what the original Prometeia build look like -- black sidepanels and a black plastic front bezel. Maybe not the most appealing design in the world, but that is about to be changed. Drastically.
The cooler after removal of the original front and sidepanels
The first thing we did was to read the instructions thoroughly. To me, it was at times obvious that the manual was written by a person whose native tongue is not english. It's not badly written, but it needs a few adjustments. A few detailed pictures for instructional purposes wouldn't hurt either, even though this procedure is not to be performed by "un-authorized personnel" as expressly stated in the very beginning of the installation guide.
Adding the aluminum profiles for the LianLi kit
All aluminum profiles are fastened with doublesided carpet tape, and seems to be working perfectly fine. Precision is a must, and despite my perfectionist tendencies, I had to do a couple of the aluminum profiles twice. On the other hand, that might be the very reason I did it all over again.
LianLi kit attached to the cooler
In the picture above, you see the LianLi kit added to the Prometeia cooler. The assembly of the sidepanels and front bezel only took a few minutes -- basically a non event. But things are about to change, as we ran into a quite a few problems with the rest of "Operation Prometeia Conversion".
I am sure many of you are familiar with the importance of proper cooling, especially when overclocking your system. But let us spend a few minutes to look into exactly how and why supercooling your processor works.
One of the most obvious reasons for cooling your processor when overclocking is the increased generation of heat. There are two reasons for the increased heat production – the first is simply that the faster your processor runs, the more heat is produced, and the hotter and faster your processor run, the more problems regarding stability you will have. The second, and in fact the main cause, for increased heat production is increased supply voltage (V).
The relationship between maximum operating frequency and supply voltage.
This figure shows us the relationship between maximum operating frequency and supply voltage. The maximum operating frequency is proportional to (Vth-V) 1,25/V, where we assume Vth is 0,6V. Between 1V and 3V, the operating frequency is approximately proportional to the supply voltage, meaning that if your processor does 2GHz at 1,5V you will most likely make it run at about 2,5GHz to 3GHz when you increase its core voltage to 2,0V.
Increasing the core voltage automatically means higher wattage output of the chip -- doubling the voltage means doubling the frequency potential, but it also increase the total wattage output by about 800%. If a CPU that originally emit 90W it will at double voltage and speed now radiate 720W of heat!
Not only will higher temperatures give you problems with stability, but it will also accelerate Hot Carrier Injection(HCI) aging and electromigration, which in the end drastically reduce CPU life time.
The other reason why you should cool your processor is based on simple semiconductor physics – the efficiency of semiconductors increase as the temperature is decreased.
Simply put, processors are made of a silicon substrate: wafers on which layers of copper used as conductors, and metal oxides used as insulators, are doped to create the circuitry. This technique is known as Dual Damascene, and was in fact introduced by IBM as early as 1984 (one can only wonder what took them so long to implement the Dual Damascene for commercial use).
The process used to make today’s CPUs is known as CMOS, Complementary Metal Oxide Semiconductor CMOS technology. One of the characteristics of semiconductors made with this process is that they become more efficient as we lower their operating temperature. For example at –120C the efficiency of a CMOS CPU actually doubles -– at least in theory.
This also applies to devices cooled less aggressively, so cooling a CPU to –20C will give us an increase by about 20%. Now, 20% may not get you overly impressed, but supercooling the processor in addition to an increased supply voltage will get you a substantial increase in efficiency.
We have now come to the part where we need to modify the LianLi case itself to make it fit on top of the Prometeia cooler. This should be relatively easy -- drill four holes for the bolts to fasten the LianLi tower to the Prometeia unit and cut a hole where the "thermal bus", as Chip-Con call it, enters the LianLi case. Chip-Con have provided a cardboard template for this job, with the four holes and the cut-out, so that you don't have to worry about cutting the wrong places. Simply place the tower upside down, put the template on the bottom and mark with a nailpunch.
Hole in the bottom of the case with protective plasic strip.
The hole for the thermal bus and three holes for the bolts were drilled when I discovered that something was wrong. The template provided showed all four holes nicely alligned in a square. However, the holes in the top of the Prometeia showed that the hole near the thermal bus should be offset to the rear.
Misalignment of the holes. Template is incorrect, the hole must be further to the back.
Showing template misalignment problems
The picture below show you the basecooler with the rubber stand-offs to reduce vibration and noise, but it also clealy show you the correct placement of the holes that must be drilled in the bottom of the LianLi case.
Rubber stand-off show us yet again the correct placement of the holes
And as if the misalignment problems wasn't enough, the template is to short, so you have to measure wether you should align the end of the template to the front or back of the case. Chip-Con need to make this template long enough to cover the whole bottom of the case, so that there is no doubt where to cut and drill the holes. While this is not a disaster as I managed to pull it of rather nicely after all, I think mistakes like this should not occur.
I also believe the cut-out for the thermal bus should be bigger, as the cut-out suggested by Chip-Con is very narrow and will squeeze the Aeroflex tube they use for insulation around the thermal bus, and may result in condensation problems.
Bolts and rivets securely fix the LianLi case to the Prometeia cooler
Furthermore, the bolts to secure the LianLi case to the Prometeia are too short, and I spent almost an hour to fasten the four bolts -- The rubber stand-offs and the rivets made it extremely hard to even reach the hole in the Prometeia. Very annoying, as 0.5 cm extra would have made the job a whole lot easier.
With patience and dedication comes results, and dispite my negativity in the previous pages, I must admit it -- Chip-Con, with its LianLi kit, has the potential to make the coolest cooler available on the commercial market. They do, however, need to iron out the issues mentioned in this review before we can call it exactly that.
Somewhat rough edges on the frontpanel
At a distance, the finished product looks wicked cool. A closer look at the frontbezel show you some rough edges I think Chip-Con needs to take rectify. And speaking of the front bezel, whenever I lift the case, the bezel keep sliding down, so they also need to find a way to keep that from happening.
The final result
Copyright © 1999 -
2019, DWPG.Com. All rights reserved.
All things considered, I think this is a great kit. Most of the things I've mentioned are details, but when these details add up, it's obvious to me that Chip-Con needs to sit down and carefully examine their design and come up with solutions to some of the problems. But when these issues are dealt with, I wouldn't have any problems recommending this kit to anyone -- the idea and concept is great, it only needs a bit of work.
Shortly we will take a look at the performance of the cooler, and hopefully, we'll break the 4GHz barrier, maybe even 4.5GHz. The basic testbed will consist of a P4 3.06GHz processor, Corsairs excellent XMS3500 DDR-RAM and the Asus P4G8X based on the latest and greatest from Intel, the GraniteBay chipset, so rest assured, this will be an interesting read.
Update: Part II available here.