OK, you went into the local PC mega-retail store because you have been reading about these heatsink thingies on the ‘net. You peruse the aisles of overpriced video cards, keyboards and floating, optical mice that know where you want to click before you do. Then you spot them. Those hunks of metal with the funny looking blades and some kind of fan on top. At least it looks like a fan; it’s kinda small though.
They are supposed to be good for your computer. Gets rid of all that nasty heat that pours out of it like a radiator in a cramped efficiency in lower Manhattan. You know that heat is bad. Makes things melt. That can’t be good for all those electronics inside the metal shroud called a case. But, which one do you get? There are 2 or 3 different ones to choose from and your knowledge on this subject is marginal, at best. Well, it’s just a bunch of aluminum with a fan. There probably isn’t much difference in them anyway. I will just go for the $15 one from Antec. It’s called the Solution Series and it says, right on the box, it works for Intel and AMD CPU’s up to the XP 2100+(not AMD approved though). Sounds all right to me. Let’s head to the checkout and…hold it right there.
Before you pluck down your hard-earned cash, let’s see if it is worth replacing the one that shipped with your PC or saving those greenbacks for another great title from Capcom. First things first, let’s have a look at this metallic savior.
From the picture you can probably guess that it is made of aluminum; therefore, it is very light, just 195g. It comes attached to a Top Power 60mm thin fan. It’s very quiet even though it runs @5500 rpm. The heatsink is a one-piece design with 80 very thin fins. It has an aluminum shroud covering two sides of the heatsink to funnel air from the fan to the fins. It also has something I dislike, a single lug attachment to the socket. I can’t say enough about the full clip for attaching the heatsink. Better to be safe than sorry on this item. The clip does use a better system of install than the method of using a screwdriver to apply pressure to the clip and slip it over the lug. The backside of the cardboard packaging has instructions on attaching the unit and applying the thermal compound, which comes with the unit. Throw the compound away and get some Arctic Silver or Alumina. The instructions are short and adequate although the accompanying picture suggests you place the sink on at an angle to engage the backside clip. Do NOT do it that way! You can damage the CPU core. The clip is loose enough to attach it in a horizontal position. You can then use the thumb latch to apply pressure down and out until the clip is over the lug. It doesn’t look that much different from the stock hsf (heatsink fan) that comes with the retail CPU. We’ll see if it performs the same as well.
In order to evaluate anything, you must choose something to use as a comparator. In this case we have chosen the Akasa-824. If you are not familiar with it, don’t worry. Although it is slightly larger and accommodates an 80mm fan, it is aimed at the same market, the low price range. Each one sells for ~$15, retail. Both are all aluminum fin designs with the Akasa having 26 larger fins compared to the 80 smaller fins for the Antec. CFM (cubic feet/minute) for both is in the low 20’s. This is a measure of how much air the fan can move. The higher the cfm, the more air is blown across the fins. Theoretically, this increases the efficiency of the unit thus decreasing the temperature of the CPU as opposed to using passive (no fan) cooling. One item on both units I should mention is the base, the part that contacts the CPU core. Neither of these hsf has a lapped base. A lapped base is one that is very smooth, providing better contact between the base and core allowing for better heat transfer. They each have obvious milling marks but the Antec has visible circling on the bottom. But for $15 you can’t expect to see a lapped base from the factory.
ASUS A7M266 w/AthlonXP 1700+ [email protected] and [email protected]
Crucial PC2100 512MB DDR RAM
GeForce 4 MX420
WD 20BG 7200RPM
Enlight 340w PSU
Arctic Alumina Thermal Compound
Coolermaster ATC-710 w/4 generic case fans
The system was run at normal settings. No overclocking. We’ll save that for the 200 level course. Let’s keep it simple and applicable to our needs. As far as the tests go, I used some standards.
The test programs utilized were my competent favorites, [email protected], [email protected], Sandra Burn-in and a new one I found recently (thanks, cjh) Toast. I must add a disclaimer here. Unless you know what you are doing, I recommend you do NOT run programs like Toast for extended periods of time. Without adequate cooling, you can fry your CPU. Neither I nor any webhost take responsibility for blah, blah, blah, you get the idea. The methodology for the tests consisted of taking measurements at specific intervals, listed here:
Before power on (baseline)
30 minutes after boot (idle temp)
After 30 minutes of testing w/each program (max temps)
10 minutes after test completion (heat dissipation)
These are not overly long tests since most people don’t stress their PC’s for extended periods of time. Generally, systems are on for a few hours a day and really stressed only when playing graphic intensive games or running one of the above programs.
Another data point I will gather is the thermal resistance of each unit. This is expressed as C/W, where C is the temperate in Celsius and W is the power in watts. This is used to measure the overall efficiency of the cooler. It basically tells you how many degrees your temperature rises per watt produced. The lower the better. In order to arrive at these figures, we take measurements from the system temps, CPU temps at max load and the power rating of the CPU, or wattage. For this test we use a power rating of 64w for my 1700+. Again, lower is better.
I will monitor the ambient (room) temps and case interior temps with a dual display digital thermometer. The system (motherboard) and CPU temps will be monitored by a program called Motherboard Monitor. This is available on the ‘net and is a free download. I suggest you get it. It is invaluable for monitoring your system temps, fan speeds and voltages. Info you might need later on if you decide to pursue this hobby.
Okay, on to the testing. The results will be grouped into table format for easier reading/comparison. All temperatures are recorded in Celsius and both system and CPU temps were assumed to be equal to case temps before powering the unit up.
Thermal resistance Antec – .52 C/W Akasa – .45 C/W
Boy, that’s a lot of numbers but what do they mean. Let’s see the difference change from idle temps to max load temps. For the Antec it rose from 50c idle to 60c max. That is a 17% change, quite large. Anytime you can get the change under 10%, you have a pretty decent cooler. As for the Akasa it realized an increase of 16%, from 48c idle to 57c max. Not so good either, but marginally better than the Antec. OK, so now I’ve discovered that the Antec is not as good a performer as the Akasa, but how does it stack up against the stock cooler? If you look at the average C/W for the stock hsf you will find it is around .51 C/W. Statistically speaking, it is a dead heat. Of course we all know the problem with statistics, there are lies, damn lies and statistics. The question you want asked is this: Should I exchange the stock cooler for the Antec. In a word, no. You will get no better results from this cooler than you would with the stock hsf. Save your money for some CD-R’s so you can burn all of your favorite reviews on them:. However, since I love testing and refuse to throw anything out, I performed one additional test. There are no tables for it since I ran only the Toast stress test, but I replaced the low power fan with the high speed Delta from my Swiftech and charted the results. They were somewhat expected but surprising at the same time. Just how important is it to have a really good fan? Running Toast for 30 minutes with the Delta I achieved a high temp of 51c. A far cry from the 60c with the thin fan. My PC also returned to an idle temp of 46c instead of the 50c noted before. The Delta is a loud fan. You can definitely hear it whine where the Top Power fan supplied with the Antec is nearly silent. But with that silence comes with a price. You will notice higher temps and in turn decrease the lifespan of the CPU. I’m hoping you gained some insight as to your decision on coolers and came away with a better understanding of the importance of proper cooling. Take my results for what they are worth. My suggestions are just those, suggestions. Thus finishes another Blue-Collar Review. However, that’s just my opinion, your mileage may vary.