In this third and final article of our system-building guide, we look at noise reduction by choosing a sound-dampening case. We also take our new machine out for a test drive, with performance and energy benchmarks. We also give a complete parts list, along with total price. Was it worth the expense and effort?
Choosing a Quiet Case
In our quest for a silent PC, the case selection becomes critical. Happily, there is one case that currently stands out in terms of thoughtful design, price, and soundproofing. We are talking about the Antec Performance One P180 Silver cold-rolled steel ATX Mid Tower Computer Case, our hands-down choice this year.
The P180 is the culmination of a year-long collaboration between Antec and Mike Chin, the founder and editor of Silent PC Review (http://www.endpcnoise.com). Mike has been an evangelist for the quiet computing movement for a number of years. Nowhere are the results of his efforts more palpable than in the P180 case.
The list of features and attention to detail is quite long. You will discover some of these on your own as you build into this case. For example, the case has a unique upper and lower chamber structure. The power supply has been thermally isolated by placing it into a separate, lower chamber (see Figure 1).
Figure 1 The P180 isolates the power supply into a separate chamber, thus helping to isolate extra heat from the system.
The goal of this is to reduce overall system heat, thus requiring less fan noise.
The case is also constructed of a unique three-layer side panel. Like the description of Achilles' shield in Homer's The Iliad, our case panels use sandwiched layers of different materials (aluminum, plastic, aluminum) in order to achieve a unique effect. In this instance, our desired effect is sound dampening—and it works. Just tap on this case and you can feel and hear how muffled the vibration is.
Another stellar feature is the integrated cooling system. The case comes standard with three separate 120mm Tricool fans. Each of these system fans has its own toggle switch on the outside of the case, allowing any or all to be toggled between three speeds (low, medium, and high). That allows you to have granular control in the balance between soundlessness and cooling. Unlike other cases in which the fan controls may hang freely, these controls are external, nearly invisible, and neatly embedded into discreet control plates on the case.
The case is also roomy, with slots for 11 Drive Bays (4 × 5.25-inch external drive bays, 1 × 3.5-inch external drive bay, and 6 × 3.5-inch internal drive bays). The hard drive cage is made of extra-thick steel in order to minimize vibration while the drives spin. Note that the hard drive cages are modular, with convenient pull-rings to help you slide them out. They are attached with thumbscrews; in fact, you'll rarely need any tools when working on this case.
The case's front bezel has a convenient, double-hinge door design. This allows the door to open up to 270 degrees, which keeps it out of the way when installing drives, changing air filters, and so on. These front-mounted air filters keep much of the dust out of your system, and they are easily removable for washing.
There are a few other things to note about this case. This is a big, heavy case. Don't hurt your back when moving it. In addition to all the dampening material, the chassis is high-gauge, cold-rolled steel. This is not the ideal portable case for LAN parties. Rather, it is durably constructed, which is what you really want in a quiet case.
Another superior feature is the elastic rubber "isolation" grommets for mounting the hard drives. This isolates the hard drive cage and chassis from drive vibration while the drive plates are spinning. Note that these grommets can be frustrating to use if you have not installed them before. There is no documentation whatsoever on these. However, you will notice that the case ships with some strange, longer case screws with exaggerated flat heads. These actually mount the hard drive. To install them, lay the drive flat in the cage, sitting on top of the grommets; then, pass the screw through case and through the grommets, thus securing the bottom (not the sides) of the hard drive to the cage.
As good as this case is, there are some things that need to be improved. One of the main drawbacks of the case is the tight fit for cables in the separate PSU area. Guiding your PSU and other cables on this case is one of the most labor-intensive parts of building this system. One tip not in the user's manual is that you can route cables under the mainboard. When you remove the other side panel, you will see that Antec has included some rubber cable straps to remind you of this fact (see Figure 2).
Figure 2 Remember to remove the P180's back panel so you can wire most of the cables under the mainboard.
This secret cabling trick will make your life a lot easier, as well as improve the airflow in your case by keeping it free of too many cables.
Still, with this case, cabling is going to be tight—especially the PSU cabling. All of the PSU cables run through a small slot. And if you do install a hard drive to the lower chamber, its cables will come within a few millimeters of hitting the whirling lower-chamber fan blades. It gets a little frustrating and annoying. But with patience and perseverance, you will be rewarded.
Another drawback is that you will have to get a PSU with a long mainboard connector, although ours had no problem reaching. Another disadvantage is that the front door bezel is a little flimsy, which can crack or snap off altogether. Note that the door is double-hinged, so to open in the full 270 degrees you will need to be careful.
In addition, Antec does not make the best fans. The Tricool fans we tested are quiet enough for us, but the ball bearings tend to wear out. When they do, consider replacing the Tricool fans with super-quiet Papst case fans. Overall, however, these complaints are minor compared to the massive advantages. Chances are you will like this case so much, you will use it again for your next two to three system builds.
To help benchmark this machine, we compared it to the performance of our 2006 build that we described one year ago. There are a few benchmark tests that are useful. One is Passmark. This does a complete system benchmark.
After running Passmark, we found that our old 2006 system at default levels scored a 507. Note that this score is the baseline, without any overclocking whatsoever. Our new machine, with no overclocking, scored a respectable 839. Note that these were only partial tests ("incomplete passmark tests"), because we did not perform the CD "burn-in" feature of Passmark.
Another benchmark focused on graphics and processing is 3dMark06. Our old system from 2006 scored a 2313. In contrast, our new system scored 3937. Again, this is at baseline with no overclocking. 3dMark06 allows online benchmarking comparison. Based on our current system, the fastest anyone has been able to overclock a similar system is to a score of 8945.
What about overclocking? We stated that mild overclocking was one of our goals. Unfortunately, at the time of writing this, the Gigabyte mainboard overclocking tools are still somewhat temperamental. For example, Gigabyte provides a windows-based overclocking program called nTune 5. Unfortunately, nTune 5 seems to lock our system from booting at various random settings. Also, it seems poorly written; it is buggy and often consumes CPU as time goes on, thus tending to negate any beneficial effect it has on overclocking. You should therefore overclock the mainboard directly via the BIOS. However, we still had some trouble with rebooting loops when doing this. We are waiting for a BIOS upgrade from Gigabyte so that we can push this system to the max.
For the video card, you can download a third-party program called CoolBits. This will unlock NVIDIA's built-in overclocking features. Note that our video card is already factory-overclocked about 10 percent. However, we tried pushing it a bit higher, with good results. For example, overclocking the video card a mere 5 percent beyond factory settings caused our 3dMark06 scores to jump about 12 percent. This card definitely offers one of the highest price-to-performance ratios in the industry right now.
Lastly, we want to test the efficiency of the system. To compare, last year's system at idle drew 0.86 amps (measured at the wall). Surprisingly, this year's system came in exactly the same at 0.86 amps idle. Even more surprisingly, our new system saved power even while dramatically outperforming last year's model under stress. For example, while running 3DMark06, last year's system drew an average of 1.40 amps. In contrast, our new system only pulled 1.22 amps. That means we have a built a much more efficient and powerful machine, considering the lower power draw in the face of significant improvement in performance scores. Much of this increased power and efficiency comes from the optimized Core 2 Duo processor, the Energy+ power supply, and the optimized video card that we chose.