by Sergey Lepilov
07/17/2007 | 10:08 AM
New coolers take different paths to the market. Well-known companies like Zalman or Scythe announce their new products by sending out press releases a few months before the official launch. Such coolers become the subject of forum discussions right after the publication of product announcements together with photos and specs.
Companies of lesser renown offer their coolers for tests somewhat gingerly, laying no claim to the winner’s laurels. Two coolers from the yet-obscure CoolJag brand are going to be examined and tested in this review.
The Taiwan-based CoolJag began its market activities in 1986. The company’s not-yet-extensive product range includes coolers for servers as well as for home PCs. CoolJag develops cooling systems for server platforms with both Intel and AMD processors. Also present among CoolJag products are chipset coolers, fans, heatsinks and thermal interfaces.
I found two new coolers with heat pipe technology to be the most interesting offer of the brand. They are called Falcon 92-Al and Falcon 92-Cu.
The coolers come in same-size boxes with plastic handles.
A tight plastic jacket envelops the cooler inside each box. The cutouts in the cardboard give you a glimpse of the cooler. The package with the all-copper model has orange captions while the aluminum model has a silvery color of the text, which makes it easy to distinguish between the two. Besides detailed specifications and other, less useful, information, there is a very ambitious claim printed on the boxes that these coolers are simply the best on the market. Well, I’ve learned to regard such claims skeptically, having tested a number of various coolers already. I’ll just check out this claim in due time.
There is a flat box with accessories at the bottom of each package:
It contains the following components:
Don’t be surprised at finding no LGA775 fastener here because it is already secured on the cooler by default. Shinetsu 7762 thermal grease has been tested in our lab earlier. I mean we tried to test it. It hasn’t changed since our unsuccessful experience with it: it is still a thick hard-to-apply gray-colored substance with next-to-zero efficiency. After a couple of attempts to use it and seeing the CPU temperature fluctuate wildly, I gave up the included thermal grease. Perhaps it has to be applied in some particular way? Well, I didn’t find special instructions about the thermal grease in the user manual, so I suppose it is meant to be used like any other grease.
But enough of sad news, let’s now take a look at the Falcon 92-Al and Falcon 92-Cu coolers from CoolJag.
As you might have guessed from the names and packaging, one model is almost entirely made from copper and the other has aluminum heatsink ribs. This is the only difference between the two coolers, so I’m going to discuss the all-copper Falcon 92-Cu as the most interesting one from an overclocker’s standpoint.
It is a classic example of a cooler with heat pipes. There are four copper pipes, 6mm in diameter, in the cooler’s base. The pipes carry a heatsink.
An attentive overclocker is sure to notice the unevenness of the ribs in the top middle part of the heatsink. This defect is indeed present on both models and on both sides of the heatsinks:
The coolers’ plastic jackets just opened up during transportation and bent some of the ribs with their edges. This was easily to correct, though.
The heatsink ribs form a kind of a bowl in which the fan is located and secured on an aluminum leg attached to two side ribs.
Considering that the CoolJag coolers have only the impeller blades of a fan, without a housing, the ribs to the sides of the fan are going to be cooled by the airflow as well. I touched the top ends of the ribs while the cooler was working and found them to be quite hot, which meant that the entire surface of the heatsink was being utilized. This is an original solution indeed, and the tests will show you how efficient it is.
A total of 66 ribs are set at a distance of 1.5 millimeters from each other.
The contact area between the heat pipes and the heatsink ribs is enlarged by means of bottlenecks that were shaped when the ribs were being put on the pipes.
Each CoolJag cooler comes with a 92x25mm fan with seven translucent blades:
A wire with a 4-pin connector goes from the fan to be plugged into your mainboard. This indicates support for a PWM-based fan speed management, but I found out that the fan speed of each cooler model was not controlled even if the mainboard supported the PWM feature correctly (I checked this out on two mainboards).
I don’t know why the manufacturer mentions PWM support in the specs, but the rotation speed is actually regulated by a thermal sensor located in a very inappropriate place:
Easy to see, the airflow from the fan will be always cooling this sensor and the rotation speed will never be increased. Regardless of the specific mainboard and the thermal conditions of the CPU, the fan is always rotating at 1400rpm although its speed should be varied from 1200 to 2500rpm, according to the specifications. By bending the thermal sensor down so that it touched the ribs, I managed to make the fan work within 1400-1700rpm depending on the CPU temperature.
Examining the cooler from top to bottom we’ve got down to its base:
The finish quality is high:
The polyethylene film protecting the base from accidental scratches had dropped off from both coolers during transportation, but I found no damage on their bottoms.
A small aluminum heatsink is installed in the bottom part of the cooler, right above the heat pipes:
It is not exactly a heatsink as its main purpose is to press the cooler’s fastening plates down. It has no contact with the heat pipes: you can see open spaces between the heatsink and the pipes and no thermal interface, either.
The pipes are soldered to the copper base.
CoolJag’s Falcon 92-Al and Falcon 92-Cu are designed for installation on LGA775 and Socket 754/939/940/AM2 mainboards. The procedure is simple. As I mentioned above, two fasteners for LGA775 platforms are already secured on the bottom of each cooler by default. If your platform supports K8 processors, you should replace the fastener and hitch it at the teeth of the standard plastic frame around the socket. You won’t even have to take the mainboard out of the system case to do that. The cooler can be oriented in two possible positions on Socket 754/939/940/AM2 – this depends on the position of the CPU socket.
It’s more difficult to mount the cooler on LGA775 and you have to take the mainboard out. You then fasten cap screws into the back-plate and use them to secure the cooler’s fastening clips with other screws. I advise you to find and insert thin spacers between the back-plate screws and the mainboard (the manufacturer doesn’t include such spacers into the kit).
The Falcon coolers can be oriented in any direction on LGA775 mainboards and the cooler’s pipes are unlikely to conflict even with tall elements around the CPU socket.
The coolers are compact at 106x122x111 millimeters:
The aluminum model weighs only 392g while the copper one is heavier at 736g.
The cooler fans are equipped with three green LEDs to delight each modder’s heart:
The following table summarizes the coolers’ specs:
Looking up on the Web, I found that the copper model is currently offered for $42-$53 and the aluminum model for $34-$40.
The coolers were tested on an open testbed as well as in a system case with the following configuration:
I’ve been asked by email what quad-core CPU we have and why it is so cool. Well, I don’t think it’s cool, I think it’s quite hot. And here is a photo of it:
So, I use an engineering sample of the Intel Core 2 Quad Q6600 processor with a default clock rate of 2400MHz. There is no traditional 5-character marking on the heat-spreading cap and Intel’s website mentions only one quad-core CPU with a frequency of 2400MHz. It is marked as SL9UM and has the following characteristics:
As I found earlier (for details see our article called Thermalright Ultra-120 eXtreme vs. Super Coolers), the maximum stable frequency of this CPU was 3392MHz at a voltage of 1.5V and with an air cooler (an open tested, an Enzotech Ultra-X cooler at 2500rpm). According to eXtreme Power Supply Calculator (v2.5 Lite), the processor’s TDP is increased from the default 105W to 201W under such conditions.
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 25-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 24-24.5°C during the tests. The fan rotation speed is shown in the diagram as reported by monitoring tools.
I didn’t look long for an opponent to the CoolJag coolers. The Big Typhoon from Thermaltake is one of the most optimal solutions in terms of performance, availability and price. (But you can write to our forum for suggestions on what coolers should be taken for reference in our future tests). I have also tested several coolers recently (for details see our article called Thermalright Ultra-120 eXtreme vs. Super Coolers), including the Big Typhoon, so it’s easy to compare the results of today’s test session with those of the cooler you’re interested in.
The maximum CPU frequency was limited by the performance of the weakest cooler in this test session. Easy to guess, it was the Falcon 92-Al. Well, “weak” is not the right word since I managed to overclock the quad-core processor to as high as 3268MHz with it (the core voltage was set at 1.4875V). It’s an excellent result for a light aluminum cooler! For example, this frequency was the limit for the well-known Scythe Ninja with a 120mm fan working at 1200rpm.
But let’s view the full picture now:
Both coolers from CoolJag are inferior to the Big Typhoon, yet it’s not surprising as I’ve seen a similar picture in many of my tests. And I say that the Falcons do show high performance, especially the copper version. The fact that the Falcon 92-Cu is only 2-3°C better than the Falcon 92-Al at cooling the CPU is rather misleading. Let’s better see what maximum CPU frequency I achieved with each cooler and what temperature the CPU had under load. This test was performed on an open testbed:
The struck-out result of the Falcon 92-Al means that it didn’t pass the test at a core voltage of 1.5V. The frequency of 3268MHz with a voltage of 1.4875V is the best the aluminum Falcon 92 can do. The copper version turns in a superb result in terms of maximum frequency. Allowing the quad-core CPU to overclock to over 3300MHz, the Falcon 92-Cu enters the league of super-coolers!
The Falcon 92-Cu is worse than the Thermaltake Big Typhoon by 6°C under peak load and by 18MHz in the highest CPU frequency achieved, yet there is one more aspect I haven’t yet mentioned. After the manipulations with the thermal sensor, the maximum speed of the fan was only 1700rpm. A 92mm fan is going to be quieter at that speed than a 120mm fan at 1400rpm. This was confirmed by noise level measurements according to our methodology . The noise level of the Falcon 92-Cu was 35.9dBA when measured from a distance of 1 meter (and 53.1dBA from a distance of 3 centimeters) whereas the Big Typhoon produced 37.8dBA and 54.7dBA of noise, respectively, in the same test.
Unfortunately, I couldn’t find what the CoolJag coolers were capable of at the maximum declared speed of the fan (2500rpm) due to the reasons explained above. I guess that a stronger airflow would increase their performance a lot considering the small distance between the heatsink ribs.
The CPU coolers from CoolJag have made a highly positive impression on me. These efficient, quiet and rather compact coolers are sure to find their place on the market and in PC enthusiasts’ system cases. The pricing is important too as there is a lot of worthy competitors in this market sector. It’s hard to point out any defects or drawbacks in these coolers. They don’t support Socket 478 and Socket A, but this is normal nowadays. There are problems with the PWM regulator and thermal sensor, yet they are not critical because a super-cooler must be quiet and the Falcon 92 coolers wouldn’t be such at the maximum speed of 2500rpm. The auxiliary heatsink above the pipes isn’t designed well, which is a single drawback I can find, but I think an overclocker can easily transform this heatsink from decorative into a functional element.
So, I’m looking forward to seeing more CoolJag solutions in the market soon and, hopefully, new and even better products. I’d be interested to see a Falcon 120-Cu, for example. And for all the folks at CoolJag – please, replace the thermal grease you put into your cooler kits! :)