Installing pretty much every cooler on the market nowadays is pretty easy. Some can be a little more difficult than others, but that can be expected. The ThermalRight Ultra 120 would seem to be one of the easier ones, or so you would think. Not to say that it's impossible, but some parts were a little frustrating. We'll start by going over some of the aspects of installing the ThermalRight Ultra 120.
As you can see in the above pictures, the ThermalRight 120 really fills out a case. Now, the rear exhaust fan had been removed from this Thermaltake Eureka Aluminum case for some other testing. But having it removed helps to illustrate how the Ultra 120 will fit, and it fits well. The Ultra 120 should fit practically any case, though may be cramped living in some Mid-Tower housing. There is also the standard back mounting plate for the Ultra when it's attached to a LGA775 board.
And here we have the Ultra mounted to our 939 test rig. As you can see, the Ultra fits just fine in this 939 box, and looks good doing it. But there are a couple of little things with this cooler that really had me steaming.
First of all, installation is simple enough in both cases, but an extra pair of hands would have helped. This really came to light when trying to install the cooler into the LGA775 box. Because of the spring loaded screws, I would be able to start a couple of threads on one or two of the fasteners, but encountered problems with the rest. This was where an extra pair of hands would have come in handy, because trying to get those screws started by yourself is absolute murder. In fact, it was so difficult with the 775 rig, that I ended up breaking a small portion of the back mounting plate. Here, take a look . . .
This was one of the lower points of this cooler. The other low involved the wire retention clip for the fan. Like I mentioned earlier, with room for two 120mm fans, why did ThermalRight include only one wire retention clip? That reason for which is beyond me, but I would want that extra fan. Wait a minute, I can have it!
Rubber bands to the rescue! Hey, it works at least, right?
Another reservation of mine involved the orientation of the cooler itself. With the cooler mounted in an Intel 775 based board, the orientation of the cooler allow for all of the hot air to directed right out the back of the case. This is where hot air should go. But with the cooler mounted in an AMD 939 system, all of that hot air drifting up from the graphics cards below is being pushed into the Power Supply, which is a very hot commodity to begin with. This difference in position may come to have an adverse effect upon case cooling, especially if your running dual graphics cards in a very small enclosure.
That's enough of my complaints for now. I think it's about time to start testing the Ultra 120. And for that we'll be using the following parts. . .
This system is pretty typical of what we use to test a cooler. And of course, the first thing we start with is installation of the cooler. But before we do that, we have to clean things up and apply some Thermal Paste.
To do that we use Arctic Clean Thermal Paste Remover, dissolving the bulk of the paste away. Then we use Arctic Silver Thermal Surface Purifier. This liquid will remove any oils or residues from the heat spreader on the CPU and or the base of the cooler. And Finally, a small dab of Arctic Silver Lumière Thermal Testing Compound. This specially designed white paste doesn't require the long burn-in times of regular old run of the mill Thermal Paste. It's intended for cooler testing and that's it.
Now for the fun part. We fired up the machine and started to cook the CPU. To do this we used an old testing favorite by the name of CPUBurn-in. This little app pushes untold bits through the chip and gets it toasty warm. In fact, it can get a chip so hot that decided to try something. Just out simple curiosity, I decided to let the processor cook, with no fans. Using Asus Probe 2.23.06, and trusted tool for measuring CPU temps. In this one daring case, I had gotten the CPU probe to read 79° Celsius. The top of the Ultra 120 was really hot. When the CPU temp clicked over to 80°C, I decided to call it quits and shut it down. But this just went to show that this cooler kept the chip at a somewhat stable temperature. Any lesser cooler, and we would have one dead processor. But I digress . . . on to the tests.
Our first test is simple, sound levels. To accomplish that, we use our trusted Omega HHSL1 sound meter to measure the dBA levels quite accurately to help give an interpretation of the sound put out by a fan and or fans.
We measured the cooler noise on an open bench system from two distances to help give an accurate sound range. The first is 10cm, with the microphone of the sound meter out of the direction of the fans air flow. This helps to help reduce the introduction of false noise created by wind turbulence striking the microphone pickup. The second is at 1m(one meter) to provide a measurement of what the average home user might hear when using this particular hardware configuration. Of course, when you install fans in a case, the noise they produce is somewhat insulated and absorbed. But then there are also other components you will hear. Simply put, don't interpret these measurements as something you'll get in the real world, this is meant more for comparison's sake between various coolers we've reviewed.
The two fans we chose to use were the Coolink SWiF 120mm case fans. These fans move tons of air and do it with barely a whisper to be heard. For controlling the fan speeds, a pair of Zalman Fan Mate 2 speed controllers were used. We ran both a single and dual fan configuration and tested both the low and high speed settings. Keeping in mind that lower is better, here's what we got.
It seems that the ThermalRight Ultra 120 is incredibly quiet. The only reason for this which I can surmise is the angled fins. ThermalRight has engineered these fins to help reduce wind drag and increase overall air flow. It would be practical to assume that this improved fin design would also help with cooling. First of all, though the Ambient Temperature may be high, but the figures are accurate. Secondly, we thought that the Ultra 120 was in a class of its own, so we picked its school mates accordingly. The Noctua NH-U12 is near identical and should provide an excellent comparison. And the Corsair Nautilus water cooler is a machine to be respected. Let's see what happens, shall we?
Wow! Take a look at that. The ThermalRight Ultra 120 beat out our king of air overclocking, Capitan Cooler himself, the mighty Noctua. The figures were really close, but the Ultra 120 did see some advantages. Though this is nice to see, there are a couple considerations for this. A one Degree Celsius difference here and the doesn't amount to much, therefor, I would place the Ultra 120 and the Noctua in the same performance class. And even better, the Ultra 120 drowned the Corsair Nautilus, showing again that superb air cooling can offer better value and better performance than beginner water cooling kits.
Addendum: Aug. 26, 2006: After our initial review was published we had long discussions with Thermalright over the results. Thermalright's own internal R&D labs have done their own inhouse comparisons and found the Ultra-120 to perform at a better than 1°C improvement over the Noctua NH-U 12. After a few retests we were unable to confirm the results that Thermalright had achieved inhouse, but during the course of our review of OCZ's Tempest cooler we got newer results that showed the Ultra-120 outperforming the Noctua NH-U 12 by 3°C. These results were obtained using a set of higher speed, higher CFM 120mm fans by Nikao, and also using Artic Lumiere instead of Artic Silver as the thermal paste. The results in that review can be seen here.
forums
send article 
hardware newsletter 
article comments (1) 
Lowest prices check
Digg this!
del.icio.us