Viewing page 8 of 13 pages.
Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 Next
HARD DRIVE MOUNTING / COOLING
One of the major design features of the P180 is mounting system for the hard
drives, which are decoupled from the chassis with soft silicone
grommets. In combination with the multi-ply construction of the case, this
is intended to address on of the most difficult-to-eliminate sources of noise
in a typical PC: Hard drive resonance. To test the effectiveness of the grommets,
the
resonant aluminum box that we usually use to test hard drives was pressed
into service.

Drive vibration is assessed by ranking the resonance produced by the drive
on a ten point scale when it is placed on this box.
A number of reference drives were installed in the lower drive cage, and
the entire cage was placed on the vibration box. The resonance of the whole
cage was then assessed on a ten point scale according to our usual standard
for vibration. In most cases, a ranking of 7 or above indicates that vibration
is unlikely to be a major source of noise. Most 3.5" drives fall in the
range of 4-6, while most 2.5" notebook drives rank 8 or 9.
P180: Drive Mounting System
|
Drive
|
Vibration Rating (1-10, 10=No Vibration)
|
Bare Drive
|
in Drive Cage
|
Seagate Barracuda IV
|
6
|
9
|
Maxtor DiamondMax 10
(sample 1)
|
2
|
7
|
Maxtor DiamondMax 10
(sample 2)
|
3
|
6
|
2 x Maxtor DiamondMax 10
(mounted together)
|
1?
|
5
|
Western Digital Raptor
Seagate Barracuda IV
Hitachi Deskstar 7K250
Western Digital Caviar
|
1?
|
6
|
The reduction in vibration transfer when a drive is installed in the P180
drive cage is impressive across the board. The Barracuda IV fairly
good when it comes to vibration is damped almost completely.
The two DiamondMax 10 samples are the most vibration-prone drives in our
lab and represent a sizable challenge. Installed in the drive cage, vibration
was reduced to good-to-excellent levels. For some reason, the worse of the
two drives benefited considerably more from the soft mounting than its lower-vibration
twin.
The two DiamondMax 10 drives were then installed together. The "5"
rating that was measured is about average for a single drive and is better
than either of the drives individually.
PSU/HDD Cooling With Four Hard Drives
Of course, the real test of the drive mounting system is to actually install
some drives in the P180 and see how much vibration noise occurs. The results
of the vibration testing suggested that a single drive wouldn't be a challenge
to silence, and the system drive did not get unduly warm during the previous
bout of testing. All this suggested that a more challenging system was needed
to test the drive capabilities of the P180.
Three drives a Seagate Barracuda IV, a Hitachi Deskstar 7K250 and
a Western Digital Caviar were added to the base system for a total
of four drives. In all other respects, this build was the same as
Configuration 4, the reference system that was used for testing the VGA
duct. The cage with these drives was also tested on the vibration test box,
as shown in the photo below. The resulting hum box resonating under the four
drives cycled in volume, but was only about as loud as a single bare Seagate
Barracuda IV at its peak.

The heavily loaded drive cage being tested for vibration.
This 4-drive package was used in the P180 to test different cooling options
in the PSU / HDD chamber. It represents the most restrictive airflow impedance
and the greatest amount of drive heat (short of multi-platter, >10k RPM
SCSI drives) that can be added to the lower chamber. Hot Potato Configuration
4 was used, and the power supply was a stock Seasonic S12-430 (latest version
with Adda ball bearing fan). The system was left to idle with each cooling
setup until HDD temperatures stabilized, typically an hour and a half or more.
For simplicity, only the data from the Deskstar has been listed in the table
below, but all drives behaved similarly.
P180 PSU/ HDD Chamber Cooling
|
Setup
|
Deskstar Temp.
|
SPL
|
1
|
46°C
|
26 dBA/1m
|
2
|
43°C
|
26 dBA/1m
|
3
|
37°C
|
28 dBA/1m
|
PSU/HDD Setup 1
The S12 PSU's fan was the sole source of airflow in the PSU chamber.
All of the drive temperatures stabilized in the mid-to-high 40's a
bit high, especially in a system with heavy drive usage. However, noise was
impressively low, rising only a one decibel above the same configuration
with a single drive. No vibration or resonance could be heard, meaning that
no further drive vibration dampening is necessary.
PSU/HDD Setup 2
The rear panel exhaust vents that surround the power supply were taped
up. This prevented the PSU fan from drawing any air in from the back, forcing
air to be pulled from the front vent and through the spaces between the drives.
This configuration brought most of the drive temperatures into the low 40's.
One even dipped to 38°C. These temperatures are perfectly acceptable in
a quiet system, although someone running a home server might prefer a larger
margin for error.
The decrease in HDD temperature was achieved without increasing the total
system noise. This remained true even when the system was placed under load,
the S12-430 had no problem exhausting the ~25W of heat produced by the drives,
and its fan never ramped up audibly.

The air vents around the power supply were sealed with tape to ensure that
the power supply drew its air through the front intake.
Note that sealing the exhaust vents will have even greater effect when a
PSU with a faster fan is used. In fact, when the power supply was swapped
out for a Seasonic Super Silencer 400W, hard drive temperatures decreased
under heavy load with CPUBurn (power draw of ~240W AC). The Super Silencer
400W fan ramped up somewhat more than the S12 under the same load. As a result,
airflow across the HDDs also increased, thus dropping their temps by 3~4°C.
PSU/HDD Setup 3
If the system used for data-critical applications, the lowest drive temperatures
may be desired. Employing a 120mm fan is likely to provide the best cooling
for both PSU and HDDs.
The supplied 120 x 38mm Antec TriCool fan was duly installed in the center
position and set to low speed. The improvement in drive temperatures soon
became apparent; they dropped into the mid-thirties, and noise rose only slightly,
although the growl of the fan worsened the quality of the noise a little.
Replacing the chamberfan with a Nexus 120 or one of the 120 x 25mm Antec
TriCool fans set on low would probably drop HDD temperature similarly while
minimizing the impact on system noise.
Sealing the rear vents had no effect on drive temperatures when the lower
chamber fan was installed; the pressure of the thick TriCool fan was enough
to force the extra airflow through the power supply when the rear vents were
blocked. There was no apparent impact on PSU fan, which was not ramping up
audibly anyway, but we can safely assume its internal operating temperature
must also have dropped a bit.
Split HDD Setup
Perhaps the best method of keeping four hard drives (and the PSU) cool is
to separate the drives. There is no shortage of drive bays in the P180, and
it is easy to move a pair of drives to the upper drive cage to provide more
breathing space in the lower drive cage. To test the effectiveness of the
upper drive cage, the system drive (a Raptor) was tried in the upper drive
cage.
P180: Upper Drive Cage
|
Drive Condition
|
Raptor Temp.
|
Lower Cage
(no fan, rear vents unblocked)
|
41°C
|
Upper Cage
|
41°C
|
With no additional cooling, there is very little difference between the two
drive cages; the temperature stabilized at 41°C wherever the drive was
installed. Noise levels were the same regardless of which drive cage was used.
Neither CPU nor GPU temperature seemed to be affected by the additional airflow
impedance produced by the hard drive in the upper cage. If the fan in the
PSU chamber is not used, there appears to be very little difference, acoustic
or thermal, between the two drive cages. Installing the lower chamber fan
or taping up the vents around the PSU makes the lower drive cage a cooler
place to install drives.
Help support this site, buy the Antec P180 Black mid-tower case from one of our affiliate retailers! |
|