by Sergey Lepilov
08/24/2007 | 12:54 PM
Today you can choose from several types of cooling systems for PC components: air passive, air active, liquid, thermoelectric, liquid nitrogen, and phase-change systems. Air coolers are the simplest to install and use and, consequently, the most widespread type of cooling devices. Liquid cooling systems are less popular, yet off-the-shelf and custom-made coolers of this type can often be found in an overclocker’s system case. Liquid nitrogen and phase-change solutions are even rarer and only utilized by PC enthusiasts for setting new records.
<%BANNER[article]%>There also exist hybrid solutions when one device incorporates features of, for example, thermoelectric and air, or liquid and air, or thermoelectric and liquid cooling. We already tested systems that embodied the first two combinations: the thermoelectric cooler Titan Amanda TEC (or Ultra Chill-TEC) and the liquid cooling system Gigabyte 3D Galaxy, respectively. Systems of the third type exist as well, but we haven’t had a chance to test them. We are going to fill in this gap by means of this review in which we’ll examine and test the air-liquid-thermoelectric cooler Freezone from CoolIT Systems.
The developer’s official website advertises two CPU coolers: Freezone and Eliminator. The Eliminator, released later than the Freezone, is meant for CPUs with a total heat output of 125W or lower. The more expensive Freezone is intended to dissipate up to 175W. Of course, the Freezone looks the more interesting solution of the two in the overclocker’s eyes, so we took it for our tests. Let’s see how efficient and quiet this hybrid of three cooling technologies is and if the CoolIT Freezone can beat one of the best air super-coolers in these two parameters.
The cooling system from CoolIT comes in a large cardboard box painted blue and white colors:
The front side of the box shows a full-size photo of the cooling system and lists its three main features: extreme CPU cooling, low noise level, and simple installation. By the way, CoolIT refers to this product simply as a “CPU cooler”.
The reverse side of the box shows the same photo of the Freezone, its detailed specs, and a comparison of the system’s capabilities with three different liquid cooling systems.
Having examined this comparison, you can’t but come to the conclusion that the CoolIT Freezone leaves no chance to modern liquid cooling systems in every parameter. Well, I’ll check out this claim shortly.
Inside the cardboard box there is a plastic box with compartments for each system component.
The largest compartment contains the Freezone itself. Accessory items are in the other compartments.
Here they are, from left to right and from top to bottom:
Each component is wrapped into a separate pack labeled with a letter (B, C, D, etc) and these letters are referred to in the installation guide, which is very helpful.
Here it is, the CoolIT Freezone cooling system:
It doesn’t look like a CPU cooler at all. The system comes to you filled, assembled and ready to work. Looking under the decorative casing with a sticker, you can see the main components of the system as follows:
I’ll enumerate the components following the direction of heat flow (or from right to left): CPU water-block; pump; two heat-exchangers on both sides of the radiator that contact it via Peltier elements; radiator; 92mm fan.
The radiator is best visible in this view:
It is not a sophisticated thing. It is a square 122x65x90mm piece of anodized aluminum weighing 850g. It is heavy.
The radiator consists of aluminum plates with a thickness of about 1mm.
The plates are connected with each other by means of aluminum inserts, about 1.5mm thick, and contact them via a white-color thermal interface. You just can’t invent anything simpler than that.
There are heat-exchangers (121x41x12mm, 80g) on each longer side of the radiator. The coolant is driven through them by the pump. Each heat-exchanger is connected to the heatsink via three Peltier elements. Thus, the Freezone features a total of six Peltiers.
Each Peltier is 40x40x3.5mm and weighs 20g. The manufacturer does not declare their performance rating, but it must be low, considering that the total power consumption of the system, with the exception of the pump and fan, is no higher than 56W.
CoolIT doesn’t declare the pump performance rating, either, only mentioning its service life (50,000 hours), weight (360g) and noise level (15dBA).
The fan to cool the radiator is 92x92x38mm and rotates with a variable speed from 2400rpm to 3200rpm, emitting from 26 to 37dBA of noise.
This fan is marked as DFB923812H and manufactured by an obscure Chinese firm Yong Lin Xing Electronic.
The last system component is the CPU water-block.
Its size is 42x42x17mm at 195g. The water-block has a single-channel design and is all made from copper anodized on the exterior.
The bottom of the water-block is protected with a plastic cap. It protects the pre-applied thermal interface rather than the bottom surface.
The thermal grease is thick but not dried up. It resembles Arctic Silver 5 in its consistence, color and adhesion. I performed a day-long check of the native thermal interface and found it to be just as effective as Arctic Silver 5. I only think that the manufacturer should apply a thinner layer of it on the water-block.
The operating principle of the Freezone is simple. The pump is driving the coolant in the closed circuit of the system. So, the heat flows from the water-block to the two heat-exchangers on the sides of the radiator and then it is quickly pumped into the radiator by means of the six Peltier elements. The radiator is cooled by a high-speed fan that exhausts hot air outside the system case. That’s all.
Summing up the weight of each component, it is easy to calculate the total weight of the CoolIT Freezone at about 1715 grams! Don’t worry, though. Your mainboard and CPU won’t have to bear all this weight on themselves. Let’s discuss the installation procedure first.
You should insert two or four bushings into the mainboard depending on the socket type:
The bushings are screwed through the mainboard via plastic spacers, so you have to take the mainboard out of your system case in either case:
A back-plate is not included into the kit.
An obligatory requirement for a successful installation of the Freezone into a system case is a seat for a 92 or 120mm fan at the rear panel. If your system case doesn’t offer one, you have to replace it (after all, the new case is going to cost only one sixth of the cooling system J). If you’ve got a 120mm seat, you have to mount the adapter plate on the Freezone fan:
Then you remove your system fans from the rear panel of the case and prepare to install the cooling system. It is convenient to put it nearby on the 5.25” cage:
The pipes are short, so you just can’t find any other temporary place for the radiator.
The water-block is secured on the CPU by means of two stiff wire clips (different for each socket type). They are fastened by means of headed nuts.
The pressure is high and the water-block remains motionless on the CPU heat-spreader. As I wrote above, each couple of clips lies in an individual polyethylene pack with a letter mark that is referred to in the detailed installation guide (PDF, 5.15MB), so it’s hard to do anything wrong.
Then you secure the Freezone through the rear panel of the case by means of four screws with spacers.
That’s all, actually. The system is ready to be connected. The connection boils down to attaching four connectors from the Freezone to the controller card:
There is potentiometer in the middle of the top part of the card (marked as RV1) that can be used to vary the system performance and fan speed. I’ll tell you more about it in the Test Results section. The card offers four connectors for fans, but the cooling system uses only one of them. The remaining two can be used to power other fans, not belonging to the Freezone.
There are two strips of dual-sided scotch on the reverse side of the card for you to glue the controller somewhere in the system case.
Frankly speaking, I only found one convenient place for the controller in my ASUS Ascot 6AR2-B. It was on the rear panel of the case, right below the mainboard’s PCI slots.
All other places were no good due to too short cables and a high capacitor on the controller prevented me from gluing it to the side of the HDD cage. Anyway, I guess you’ll find a place for a small card somewhere. And finally you have to connect the cables to the appropriate headers (the headers are unique, so you can’t confuse them). Moreover, you have to connect another cable (from the pump) to the CPU fan connector on the mainboard.
Assembled and installed into a system case, the Freezone looks like follows:
It’s still quite roomy in there but I had to remove the system fan from the side panel which would not close otherwise.
The specifications of the CoolIT Freezone are listed in the following table.
The recommended price (quoted from the official website) is terribly high in comparison with other cooling systems, but you can find this cooler in retail shops for $290. The latter price is high, too, and comparable to the price of a junior quad-core CPU.
The CoolIT Freezone and its opponent were tested in a system case with the following configuration:
The CPU with a slightly polished-off heat-spreader was overclocked to its maximum frequency on the weakest of the tested coolers. You’ll learn shortly which cooler it was. The resulting frequency was 3294MHz at a core voltage of 1.4875V.
All tests are performed in Windows XP Professional Edition Service Pack 2. SpeedFan 4.32 is used to monitor the temperature of the CPU, reading it from the CPU sensor. The CPU is heated up by means of OverClock Checking Tool version 1.1.0 in a 60-minute test during which the system remains idle in the first and last 4 minutes.
The mainboard’s automatic fan speed management is disabled for the time of the tests. The thermal throttling of the Intel Core 2 Duo processor is controlled with RightMark CPU Clock Utility version 2.25. Our CPU begins to skip clock cycles on reaching a temperature of 82°C and higher.
I perform at least two cycles of tests and wait for 25-30 minutes for the temperature to stabilize during each test cycle. The maximum temperature of the hottest CPU core in the two test cycles is considered as the final result (if the difference is not bigger than 1°C – otherwise the test is performed once again). Despite the stabilization period, the result of the second cycle is usually 0.5-1°C higher.
The ambient temperature was monitored by means of an electric thermometer and remained at 26°C during the tests (marked with a vertical red line in the diagrams). The fan rotation speeds are shown in the diagrams as reported by monitoring tools.
Fortunately, I don’t have an air cooler with a recommended price of about $400. I have a Zalman Reserator 2, yet it costs considerably less, below $300. And it is much harder to assemble it than to install a good air cooler. So, I took an Enzotech Ultra-X cooler in quiet mode (1300rpm) as an opponent to the CoolIT Freezone.
First, let’s see how the CoolIT Freezone and its opponent cope with the overclocked quad-core processor.
Alas, there is no miracle. The system from CoolIT is only better than the good air cooler in Idle mode, which is indicative of inefficient use of the Peltier elements that might be shut down at low CPU loads. Under constant load the temperature is growing up steadily and finally exceeds that of the Enzotech Ultra-X. Here are the monitoring graphs for the hottest CPU core:
CoolIT Freezone | Enzotech Ultra-X |
The CoolIT Freezone maintained a lower temperature at first, but the CPU temperature was growing up steadily with it. I guess the reasons lie on the surface: the volume of coolant is too small, the radiator is too small and simple, and the fan probably has too low performance. All of this limits the efficiency of the Peltier elements and of the pump which are ready to go on cooling at the same rate as in the beginning of the test but the other system components just cannot dissipate that much of heat.
It can be argued that I took a wrong processor because CoolIT developed the Freezone to dissipate 175W of heat rather than 200W (the estimated heat dissipation of the overclocked quad-core CPU in my tests), but then I can’t understand the purpose of this system and its positioning on the market as such a high price.
And here is the noise level diagram (the noise is measured using our traditional method described before). The subjectively comfortable level of 36dBA is marked with a dash line. The ambient noise level was about 34dBA.
I can’t say anything good about the Freezone again. The pump starts up first and is not too loud at work, yet its noise is far louder than that of quiet (and even not very quiet) air coolers. And then it becomes even worse as the fan joins in. Experienced overclockers should know that a 92x92x25mm fan won’t be quiet at a speed of 2400rpm, which is the minimum for the Freezone. Here, we’ve got a 92fan with a thickness of 38mm working at a speed up to 3200rpm! It is very loud.
I don’t forget about the system performance and fan speed control installed on the controller card but I just can’t say anything about it. It is to be turned with a cross-headed screwdriver counterclockwise to increase the performance of the Freezone and clockwise to reduce the performance and noise. However, my turning it around had no effect on the results I could track in the monitoring graph and on the CPU temperature. Perhaps the controller was defective, but whatever mode the Freezone defaults to, it is still too noisy and inefficient for its price.
I had expected much more from the CoolIT Freezone in terms of noise as well as cooling performance than it actually showed in my tests. Yes, this cooling solution has its strong points. It is universal and can be mounted on each modern platform as well as on the outdated Socket 478. It is ready to be installed right out of the box and does not require that you assemble it, fill it up, etc, as you have to with most other liquid cooling systems. As opposed to them too, the Freezone resides inside the system case without interfering with any components (except for the system fan on the side panel). Unfortunately, this is all the good I can tell you about this cooler. The rest are much heavier drawbacks.
I acknowledge that CoolIT has created a high-performance cooler that can cope with an overclocked quad-core processor that generates almost 200W of heat. But does its performance match its price? The answer is a definite No. I can recall the recent article about the XIGMATEK HDT-S963 cooler in which the Cooler Master Hyper TX with a recommended price of $22 (and available in retail for about $30) cooled the same CPU just as well. I don’t even mention such coolers as Enzotech Ultra-X. Added to this are the high noise level and high power consumption of the Freezone.
The CoolIT Freezone can be improved. For example, they can install a larger radiator, for a 120mm fan at least (there is enough room in the system case) and with better design (like wavy ribs to enlarge the total area, corrugated butt-ends of the ribs to reduce the resistance to airflow). They can also increase the amount of coolant, perhaps even include an expansion tank into the circuit. After these minor improvements it will be clear if the performance of the Peltier elements and of the pump should be increased as well. But will the developer improve the system further? I’m not sure at all.
