Viewing page 6 of 6 pages.
Previous 1 2 3 4 5 6
BIOS
One of the BIOS features mentioned in the manual for the Kloss is the "Silent
Mode" that is apparently used to underclock the CPU instead of straightforwardly
setting the FSB. Along with "Turbo Mode" (overclocking), they are
the the only CPU clocking options available. According to the manual, "Silent
Mode" can be used to permanently underclock the processor to 50%, 37.5%,
25% or 12.5% of full speed. No voltage or fan adjustment options are available.
Unfortunately, "Silent Mode" and "Turbo Mode" were mysteriously
missing from our sample system, even with the latest BIOS installed. Whether
this is because these modes are not supported with our choice of CPU or whether
its simply an error in the documentation is unknown. It's worth pointing out
that Silent Mode is mentioned in the release notes for the latest BIOS ("Silent
Mode modified [CPU Throttling --> SouthBridge Throttling]").
Without Silent Mode or Turbo Mode, no CPU frequency adjustment is possible,
so our test system was limited to the stock settings for our processor: 200
MHz FSB and 2.8 GHz CPU.
The limited options available in the BIOS are probably indicative of the target
market for the Kloss: HTPC users who are not experienced in building computers
and would rather treat their computer as an appliance than spend time tweaking
its options. Whether this approach is satisfactory for a quiet system will depend
on the implementation of the fan control and the effectiveness of the CPU cooling.
TESTING
The following tools were used during testing:
Testing consisted of an examination of how the system behaved acoustically
in response to changes in thermal load. No performance benchmarks were run;
performance in a SFF system depends mainly on the components it is tested with,
with the motherboard accounting for at most a 3% difference in performance.
Ambient temperature during testing was 23°C. Ambient noise level was 17
dBA/1m.
Fan Behavior
|
Kloss KL-I915B Noise Characteristics
|
|
System State
|
CPU Fan Speed
|
System Fan Speed
|
System Noise
|
|
Start-up
|
–
|
–
|
33 dBA/1m
|
|
Idle
|
1150 RPM
|
1600 RPM
|
27 dBA/1m
|
|
Load
|
1150 RPM
|
1600 RPM
|
27 dBA/1m
|
|
System Fan Stopped
|
1150 RPM
|
0 RPM
|
23 dBA/1m
|
The Kloss starts up with a quick burst with the fans at full speed, then reduces
the fan speed to a lower level once Windows starts to load.
At all speeds, the major source of noise is the thin 70mm exhaust fan that
produces a distinct midrange drone. Surprisingly, the tiny 40 mm fan in the
power supply did not seem to ramp up much under load. The noise it did produce
was in the form of a low frequency buzz. The slow speed of the power supply
fan is a testament to the effectiveness of removing the major sources of heat
from its airflow path. The one remaining fan — the 92 mm CPU fan —
contributed a small amount of low frequency hum and the mechanical buzz that
is typical of ball-bearing fans.
After the initial boot sequence, the fan speed remained more or less constant
regardless of CPU load. Only after half an hour of constantly
running CPUBurn did the CPU fan increase to approximately 1550 RPM. After about
a minute at this level, it dropped back down to its original level at 1150 RPM.
These short spikes in CPU fan speed were repeated thereafter at roughly ten
minute intervals. They did not noticeably increase the noise level above the
baseline level.
Power and Thermal Characteristics
|
System
|
Different Components
(est. wattage change)
|
AC Power:
Idle
|
AC Power:
Load
|
|
Kloss KL-I915B
|
reference point
|
67W
|
131W
|
|
Shuttle SB86i
|
* Samsung MP0402H (no change)
* 2 sticks DDR400 RAM (+5W)
|
72W
|
150W
|
|
AOpen XC
Cube EX915
|
* Samsung SM-352B Combo Drive (+3W)
* Samsung SP0802N (+5W)
* 2 Sticks DDR400 RAM (+5W)
|
81W
|
158W
|
The power consumption was impressively low for this processor. Other SFF systems
we've tested with this processor have drawn as much as 14W more at idle, although
part of this difference may be attributed to the more conservative test build
that we used in the Kloss. However, the gap widens to almost 30W under load
— too much to put down to the difference in components. This suggests that
the power supply is either more efficient than usual, or perhaps it benefits
considerably from not having to exhaust the heat produced by the CPU.
|
Activity
|
AC Power Max
|
CPU
|
HDD
|
|
Idle
|
67W
|
47°C
|
39°C
|
|
Full Load: CPUBurn
|
131W
|
68°C
|
44°C
|
Temperatures were hardly impressive — SFF systems rarely are — but
were adequate for our test components. Although the CPU temperature hit 68°C
— around the point where CPU throttling typically kicks in — no throttling
was detected at any point during our testing. As the occasional spikes in fan
speed under load attest, our low-end processor was enough to push the thermal
limits of the system at this noise level. Even a slight increase in the ambient
temperature would likely be enough to push the fan speed up permanently when
used under load. A more powerful processor in this system would almost certainly
cause an increase in noise or trigger CPU throttling.
A rough idea of the airflow through the lower chamber can be drawn from the
change in hard drive temperature under load. This change is significant, as
the hard drive was not in use during the CPU stress test. The 5°C rise in
temperature that our drive recorded can be attributed to the additional heat
generated by the power supply under load.
Audio Recordings of the Kloss KL-I915B as tested:
MP3:
Kloss KL-I915B w/ P4 520 & Fujitsu notebook drive, Idle and Load: 27 dBA/1m
Comparatives
MP3:
Arctic Cooling Silentium T2 with test system (3.5" HDD suspended), Idle: 23
dBA/1m
MP3:
AOpen EY855-II w/ P-M 1.6 GHz & suspended Samsung notebook drive:
10 seconds normal (25 dBA/1m), 10 seconds w/PSU fan stopped (20 dBA/1m)
MP3:
Shuttle XPC SN95G5 with Samsung Notebook Drive, Idle: 27 dBA/1m
MP3:
Shuttle XPC SN95G5 with Samsung Notebook Drive, Load: 30 dBA/1m
|
SPCR MP3s: HOW TO LISTEN & COMPARE
The recordings above were made with a high resolution studio quality digital recording system. The microphone is 3" from the edge of the fan frame at a 45° angle, facing the intake side of the fan to avoid direct wind noise. The ambient noise during all recordings is 20 dBA or lower.
A quick and simple way to use these recordings for valid listening comparisons is to play the quietest recording on only one speaker (or a pair of headphones) and set the volume so it is just barely audible a meter away. You must turn off any special sound effects, and set equalizer / tone controls to neutral or flat. Don't touch the volume setting afterwards, and use the same one speaker when you listen to any of the other files. The end result will be reasonably close to the actual recorded sound levels.
Here is a recording of a very quiet sound that is barely audible from 1 meter away even in a super quiet room.
For full details on how to calibrate the playback level of your sound system to get the most valid listening comparison, please see the yellow text box entitled Listen to the Fans on page 3 of the article SPCR's Test / Sound Lab: A Short Tour.
|
CONCLUSIONS
Trigem has done a good job designing the Kloss as a livingroom-worthy barebones
system. The unusual double-decker configuration has merits in both ease of installation
and thermal performance. Although the system is far from perfect, it is one
of the better performing SFF systems we've tested.
The key to the Kloss' performance is well-designed airflow. The CPU duct allows
the CPU to be cooled more effectively (with slower fans) than a more conventional
system, although the "blowhole" that feeds it is probably not in the
safest position and actually forces the air to go against the natural direction of rising heat. The lower chamber is also effectively a duct for the power
supply, as the drives only add a minimal amount of heat to the chamber. The
benefits of the double-decker design are effective cooling, and lower
ambient temperature that lets the power supply operate more efficiently.
The Kloss' 27 dBA/1m noise level at idle is comparable with some of the better
systems we've tested, and the fact that it stays at this level even with our test
processor fully loaded is impressive. Nevertheless, this system
could have been even better if a quieter fan had been used to exhaust
the system heat. Manually stopping the fan caused the noise level to drop by
4 dBA/1m to 23 dBA/1m — quiet by any standard, and approaching the ambient
noise level in many environments.
For the most part, installation was better than the competition. Here again
the two-chamber design comes into play, making it easier to access all relevant
parts of the system without first needing to move anything out of the way. The
biggest downside may be the mounting method for the drives, which is liable
to cause vibration noise if full-size drives are used. The video card and HSF clearance issue is probably serious only for those who wish to install a large (read: High end gaming) VGA card. But it is still worth noting as a deficiency that could have been anticipated by Trigem. A different HSF or a slight repositioning of board components (such as the CPU socket) could have avoided the problem altogether.
It is important to keep in mind that the system was tested with low noise as
the primary goal, not performance. Hotter components may well eliminate the sonic advantage
that the Kloss has over its competition. However, if the highest performance is not a requirement
— and an HTPC should work perfectly well with our choice of processor —
the Kloss is an good choice for a system that is pretty quiet "out of the
box".
|
PRO
Innovative double-decker design
Simple installation
Good cooling performance
Power supply isolated from system heat
Quiet with low-heat components
|
CON
Blowhole position is potentially dangerous
Possibly vibration-prone drive installation
Noisy system fan
Large bezel size
|
Much thanks to Trigem
Computer Inc. for providing the Kloss KL-I915B sample and to
Newegg.com for
the Intel 520 loaner.
* * *
Discuss this article in the Silent PC Review Forums.
| Help support this site, buy from one of our affiliate retailers! |
|
