Viewing page 3 of 6 pages. Previous 1 2 3 4 5 6 Next
The cross-flow fan is highly unusual, and deserves a mini-review of its
own. Unlike most typical axial fans used in computers, it is made completely of metal, and it weighs considerably more.
It comes with its own bracket, so the additional weight does not have
to hang from the graphics card. Noise seems to have been
taken into account, as all of the moving parts (motor
included) are isolated from the frame with soft rubber.
Air is sucked in through the opening at the top and forced out through the
rectangular vent at the bottom.
A good view of the intake and the fan blades.
Unfortunately, the fan vibrates at lot, even at lower speeds and the rubber decoupling
is not that effective against this vibration. Perhaps the relatively high mass of the metal blades makes
it more difficult to balance than a plastic axial fan.
The motor is decoupled with soft rubber...
...and so is the axle at the other end.
The bracket is quite substantial, and should not have any problem
supporting the weight of the fan. The bracket comes in two parts: The bracket
itself, and a mounting plate for the fan. The two parts and the fan are screwed
together with machine screws, and can be assembled easily outside of the
The bracket also comes with a simple fan controller that allows the user to
switch between three preset speeds. The controller is powered by a standard
Molex plug and accepts 12V as its input voltage.
A weighty bracket for a weighty fan.
Three speed fan control.
The fan was placed on soft cushioning foam and driven by a variable DC voltage power supply. We listened and measured the noise with our B&K 2206 sound level meter (SLM) from one meter away. The fan's airflow was pointed away from the SLM. It was measured both in free air (unmounted SPL) and installed on the heatsink (mounted column). The supplied fan controller was used except for the 5V test, where the fan was connected directly to our power supply.
The three preset fan speeds delivered 6.5V, 9.5V, and 11.8V to the fan. Only
the lowest speed could be considered quiet, and even that had an unpleasant
noise character that would be out of place in a quiet system. The fan started
reliably at 5V, though, and was considerably quieter than the lowest preset
All of the preset fan speeds had very rough, "dirty" noise signatures. The dominant
noise was always the hum of the motor, but a sizable amount of clicking and
rattling could be heard underneath. At the medium and high settings, turbulence
noise was also a factor. However, the sheer volume of noise at these levels
makes the subjective description less relevant; both high and medium are too
loud, whatever the quality of noise. The lowest setting was marked by a distinct
pure tone that was decidedly unpleasant to listen to. The tone had a very slight
warble that called attention to itself.
At 5 volts, the hum of the motor noise disappeared almost entirely, replaced
by a chattery chugging that made us wonder how well balanced the fan was. A
faint secondary tone could also be heard, and this one did not warble or change
in any way. At this level, the fan was quiet enough for most systems, and would
probably blend in to the other noise sources in the system. However, it is still
clearly audible, and would not be suitable for use in a system that is meant
to be inaudible.
In addition to the changes with fan speed, the noise also changed significantly
when the heatsink was placed in front of it. The resulting impedance (to airflow) caused
the noise level to increase noticeably and measurably, as the table above shows.
The most noticeable changes were:
- motor hum rose to a higher pitch
- increased broadband turbulence noise
|Help support this site, buy from one of our affiliate retailers!|