Viewing page 4 of 5 pages.
Previous 1 2 3 4 5 Next
Sound pressure levels were measured in a small carpeted room (10'x10'x8') with a calibrated B&K sound level meter (model 2204) capable of reading down to below 0 dBA. The PC was set on a 20" tall stool near the middle of the room and the SLM positioned or held in various spots 1 meter away from the nearest surface of the PC case. The ambient noise during the acoustic measurements was ~18 dBA, well below the level of the PC being measured so that it did not affect the readings. Each measurement was performed after at least 15 minutes continuous use in the specified activity.
Listening Analysis
Subjectively, the ARM Systems StealthPC P4-3.2 is very quiet at idle. It compares favorably with any normal production PC, including quiet Dells. It is very unobtrusive, though not inaudible in a quiet room. Overall, it is not much different from the first Stealth XP2000+ system reviewed in late 2002. Given the massive increase in sheer wattage under the hood, this is no small feat.
The maximum noise is from the rear where both the PSU and case fans are mounted, the norm for standard box PCs. The simple expedient of attaching a large piece of soft foam on the wall behind the PC will lower this noise by several decibels. This is not obtrusive when the PC is under a desk; it is a noise-reduction technique practiced by many PC silencers. As mentioned previously, DVD or CD playback adds little or no noise overall.
At full load, the front and rear 120mm fans speed up by a couple hundred RPM; the details are recorded in a graph below. Noise rises by several decibels; most of it is the relatively benign whooshing of airflow turbulence. This is where the current system differs most from the previous ARM Stealth; when pushed hard, over time, there is a gradual increase in noise (to still a very modest level) as the case fans ramp up to move the added heat out of the case. Considering that the total CFM through the case is probably more than double compared to the 80mm Panaflo-L equipped 2002 model, the increase in noise at max load is modest, indeed.
The loudest noise comes during HDD seek/write. The clickety-click noise of the Seagate Barracuda SATA V drives is mostly muted by the EAR soft grommet mounting and by the AAM, but it is not completely eliminated. In this regard, it is Seagate who is responsible for the noise change: The Barracdua IV was a quieter drive, but it is no longer available, and the V is not as quiet. ARM Systems has experimented with decoupled suspensions as a way of achieving the lowest HDD noise. They are not yet convinced that such mounting systems can consistently survive the rigors of shipping. When they devise a HDD suspension system that reduces noise further without increasing the risk of shipping damage, ARM Systems will certainly inplement it.
TEMPERATURES & FAN SPEEDS
Both temperature and fan speeds were monitored with the Intel Active Monitor preloaded on the system. The Panaflo 92mm fan on CPU heatsink remained at a steady 1060 RPM at all loads. It is not thermally controlled. The speed of Zalman fan on the VGA cooler also remained unchanged by load. I measured the voltage across its terminals at ~6V.
HDD1 is the bottom drive; HDD2 is the top drive. The latter's higher temps are easily explained by its position above the rising heat of HDD1 and its greater distance from the airflow of the intake fan. Seagate's maximum safe temp for these drives is a toasty 60° C, so the operating temps here are perfectly good.
|
FAN RPM
|
|
Fan
|
Idle
|
Folding
|
CPUBurn
|
|
CPU
|
1060
|
1060
|
1060
|
|
Front
|
500
|
670
|
800
|
|
Back
|
670
|
900
|
1050
|
As you can see, the case fans are thermally controlled, through the "Fan Control Config" in the BIOS of the Intel motherboard. This is the same feature found in an earlier Intel board that I have, which uses PWM (pulse width modulation) to control the speed. I found the feature on my board to cause fans to make so much extraneous noise that I wrote Intel's feature off as a complete failure. It's not clear to me what ARM Systems has done, but it is clear they have made it work. Perhaps it is a matter of choosing the right fans, which might explain why Steve Collins will not reveal what 120mm fan they are using.
|
SYSTEM POWER DISSIPATION
|
|
Activity
|
Peak AC Power Draw
|
|
Idle
|
127W
|
|
Boot
|
191W
|
|
Folding
|
194W
|
|
PCMark04
|
236W
|
The above table shows the highest wattages seen using a Kill-a-Watt AC power meter during the activities listed. AC power draw tells us the total electrical power delivered into a PC. It is also the total heat that's being generated; it has to be dissipated effectively for the PC to work well. The higher the wattage, the more difficult it is to evacuate the heat quietly. The 236W peak was seen in PCMark04 during a combination of HDD access and complex video rendering.
| Help support this site, buy from one of our affiliate retailers! |
|
» 51591 reads
