Viewing page 6 of 13 pages.
Previous 1 2 3 4 5 6 7 8 9 10 11 12 13 Next
THERMAL & ACOUSTIC TESTING
Two systems were assembled in P180 samples for testing:
1) A very hot, powerful system based around an Intel 660 overclocked
to 3.8 GHz, an nVidia 6800GT video card, and a Western Digital Raptor 74G drive.
This is definitely a high-end, high power gaming system that can only be significantly
bettered by a dual VGA card setup. With a lack of care and planning, such a
system could easily measure 40 dBA/1m. Our task here was find the most efficient,
quietest method of cooling this hot system in the P180 case.
2) A mid-thermal but quite powerful system based around an AMD
Athlon 64 3500+ (Winchester core), nVidia 6600GT video card, and a dual Maxtor DiamondMax 10
RAID storage setup. This is probably as powerful a system as most silence-oriented
readers will want to build. The rule of thumb is that the hotter the components,
the harder it is to keep them running well quietly.
Together with Ralf Hutter's work in Part One, our review covers the P180 with three different classes of systems:
- a low-thermal as-quiet-as-possible system,
- a mid-thermal powerful system,
- a hot gaming system.
This should provide users with a comprehensive appreciation of the P180's performance and potential.
Testing Tools
SYSTEM ONE: THE HOT POTATO
System Components
Ambient temperature during testing was usually 21°C, rising occasionally
to 22°C. Ambient noise level was 18 dBA/1m. EIST (SpeedStep) was enabled,
as was the PowerMaster feature of the AOpen motherboard. It provided a 5% overclock at full load and underclocked the processor down to 2.56 GHz at idle, which is reflected in the idle temperatures and AC power.
Configuration 1
The starting configuration was as follows:
- The CPU was overclocked by 5% under load. The FSB was run at 210
MHz, bringing the CPU frequency up to 3.8 GHz, on par with the top of the
line Intel 670 processor. The system remained completely stable under this
modest overclock.
- The built-in fan on the VGA card was undervolted to 8.25V, the level
at which the fan noise dropped to the level of the other fans.
- The 120 x 38mm fan in the PSU chamber was removed on the assumption
that the airflow provided by the Seasonic S12 would be enough to cool the
Raptor installed in the lower drive cage.
- The upper drive cage and front fan holder were removed entirely to
minimize the airflow impedance near the intake.
- The VGA duct was also removed in hopes that the airflow around the
VGA card would already be sufficient.
- Airflow in the upper chamber was provided by the the rear Antec TriCool
fan, set on Low.
- The top fan was left installed but unplugged, and the top vent was covered
by placing a thick notebook over the hole.
- No fan was used on the Ninja heatsink in hopes that the airflow of
the twin TriCool fans would be enough to allow fanless operation.
Firing up CPUBurn for the first time proved a bit disappointing. CPU temperature
shot up to 74°C, where throttling kicked in to prevent the temperature
from rising any higher. To reduce the temperature, as much impedance as possible
near the intake was removed: The front door was opened and both the filter
and the filter door removed. Even the VGA duct was sealed up to address the
possibility that air was uselessly being pulled though and exhausted without
doing anything.
Unfortunately, none of these adjustments reduced temperatures enough to prevent the CPU
from throttling, although the changes in the amount of throttling showed that
these slight differences caused equally slight changes in airflow. The most
significant factor proved to be whether or not the front door was open, which
reduced temperatures by 2°C but audibly worsened the system noise somewhat more than the measured 2 dBA/1m would indicate.
Configuration 2
The next step was to unblock the top vent and turn on the second exhaust
fan, set to Low. Both the 120mm fans were blowing out, as per Antec's
default setup. All the other details were left the same as in Configuration
1. Unfortunately, the additional fan seemed to produce no significant benefit;
the CPU continued to throttle under full load, even after the opening the
door and removing the filter.
In spite of the additional fan and the unblocking of a potential noise path,
the level of noise was very similar to the first configuration. The SPL measurement
bore out this conclusion: They measured the same.
Configuration 3
The ineffectiveness of the second fan suggested that there might be a bottleneck
in the airflow, so a Nexus 120mm fan was installed in the front fan bay.
The rest of the system was the same as in Configuration 1. Like the previous
two experiments, this one also proved ineffective. In fact, the 2 dBA/1m increase
in system noise made this a particularly poor attempt at preventing the CPU
from throttling.
Configuration 4
At this point, we began to give up on being able to run our processor heatsink fanlessly.
However, one final option, inspired by Intel's TAC duct, remained to be tried: Go back to Configuration 1, and unblock the top vent
without turning on the fan. This made the top vent the primary intake for the back panel fan, and this air would be forced to flow through the huge Ninja HS, effectively
lowering the ambient temperature around it. The disadvantage of this configuration
would be that less air would be drawn through the front intake. System and VGA temperatures
were watched closely to ensure they didn't rise to unhealthy levels.
This configuration takes advantage of the large size of the Ninja heatsink.
Because the Ninja fills almost the entire width of the case, air drawn from
the top vent by the rear fan is forced to travel over the fins of the
heatsink. A lower profile heatsink would benefit less from this configuration,
and a great deal of air would be needlessly sucked through the case without
ever cooling anything.
This experiments proved to be a success. With the CPU
under load, the temperature stabilized at 65°C — warm, but not high
enough to start throttling. Surprisingly, the system and GPU temperatures
did not seem to be affected much. The GPU even dropped a degree or two in
this configuration.
The level of noise in this configuration was almost identical to the noise
in Configuration 1. Surprisingly, blocking or unblocking the top vent made
very little audible and no measurable difference. In fact, I personally preferred
the sound of the case with the top vent open, as it sounded slightly less
"choked" and less resonant. (Editor's Note: I and
another innocent bystander called in to witness also preferred the sound with
the top blow hole open for the same reasons.)
Configuration 5
The final configuration was the obvious one: Add a Nexus 120 fan at 12V
to the Ninja heatsink in Configuration 4. It was set up to blow
through the fins towards the rear case fan, creating a push-pull airflow effect
across the heatsink. The CPU temperature at full load dropped dramatically
to a mere 55°C, while the noise was similar to Configuration 3. The ten
degree drop in temperature is substantial enough that most gamers would probably be
willing to pay the acoustic price.
Acoustic Recordings and Summary of Configurations
MP3:
P180 "Hot Potato" Configuration 4: 25 dBA/1m
MP3:
P180 "Hot Potato" Configuration 4, door open: 26 dBA/1m
MP3:
P180 "Hot Potato" Configuration 3 (Nexus 120mm installed in front bay): 27 dBA/1m
|
Antec P180 / Intel High End System
Test Summary
(Stressed by 2 x CPUBurn)
|
|
Configuration
|
CPU Temp.
|
SPL
|
|
1
|
74°C*
|
25 dBA/1m
|
|
2
|
72°C*
|
26 dBA/1m
|
|
3
|
72°C*
|
27 dBA/1m
|
|
4
|
65°C
|
25 dBA/1m
|
|
5
|
55°C
|
27 dBA/1m
|
*CPU Throttling occurred in these configurations.
>> AC Power Draw at full load: 236W <<
|
| Help support this site, buy the Antec P180 Black mid-tower case from one of our affiliate retailers! |
|
