George Maling, Noise Control Center and David M. Yeager, Motorola Corp
Introduction
Air-moving devices (AMDs), such as fans and blowers used for cooling in
electronic systems invariably generate acoustical noise which must be considered
if people are exposed to the emitted noise. In this article, we discuss
quantitative measures for the description of noise, design guidelines for the
selection of low noise air movers, upper limits for the noise emissions of
computer and business equipment and measurement procedures which may be used to
evaluate the noise radiated by electronic equipment.
Important Terms
Sound pressure and sound power are two important terms in acoustics and
noise control engineering. A complete glossary of terms in noise control
engineering is available [1].
Sound pressure is the rms value of the small variations above and below
atmospheric pressure that constitute a sound wave. The human ear is sensitive
to sound pressure, and it is the quantity measured by a sound level meter. Sound
power is the energy per unit time emitted by a sound source and is a good
measure of noise source emission. It is usually calculated from measurements of
the sound pressure taken around a source. The sound pressure level and sound
power level are corresponding logarithmic quantities. The unit of sound pressure
level is the decibel, and the unit of sound power level is either the decibel or
the bel [1].
A hemi-anechoic room is a special test room which has a hard floor, while
other room surfaces are highly absorbent of sound. Such a test room is useful
for measurement of sound pressure levels and determination of sound power
levels.
All of the surfaces in a reverberation room are highly reflective of sound,
thus making it useful for determining of sound power levels.
Measures of Noise Emission
The sound power level [1,2] is widely used to characterize the noise
emissions of electronic equipment such as computers and business equipment.
While the sound pressure level at a given position is easily measured with a
standard sound level meter, the sound power level is highly dependent on the
position of the microphone and the acoustical characteristics of the room in
which the equipment operates. For these reasons, standard methods have been
developed [3-5] for determination of the sound power level of both air movers
and complete systems.
From an acoustical viewpoint, the most important consideration in air-cooled
systems is the selection of one or more AMDs to provide cooling air. The type of
AMD used for cooling depends on the volumetric air flow required and the static
pressure rise across the AMD required to force air through the system. General
design guidelines can, however, be given for the selection of AMDs.
Design Guidelines
The following guidelines should prove to be useful in the design of cooling
systems for electronic equipment.
1. Choose the thermal design point appropriately, taking proper
account of thermal and acoustical effects. Small changes in electronic case
temperature requirements can result in dramatic changes in AMD noise emission
levels. A system designed to withstand environmental extremes (room temperature
and air density) can usually benefit considerably from an adaptive cooling
design in which AMD rotational speed is controlled by an on-board controller and
thermal sensor. In many cases, an adaptively-cooled system can provide better
thermal protection under extreme conditions while having reduced noise emissions
under more "typical" conditions.
2. Design the system to be cooled to have the lowest possible static
pressure rise for the required air flow. A low static pressure rise indicates
that the AMD can operate at a low tip speed, resulting in a low noise level. The
static pressure rise across a system is caused by several sources of resistance,
such as the devices being ventilated and finger guards which may be required for
safety. If unnecessary sources of resistance can be eliminated, the air flow
will increase. It should then be possible to reduce the tip speed of the device
to obtain the desired air flow at a lower noise level.
3. Select the operating point for a centrifugal blower so that it
operates near its point of maximum static efficiency, considering the required
air flow rate and the pressure drop through the system. Operation away from the
point of maximum static efficiency should be in the direction of lower static
pressure rise and higher air flow.
4. Select a point of operation of a fan that is away from the best
efficiency point in the direction of higher air flow and lower static pressure
rise. Small fans are often unstable when operated at air flow rates less than
the air flow rate at the best efficiency point. They are often very noisy under
conditions of high static pressure rise and low air flow rate.
5. Select a fan or blower with a low sound power level and avoid
AMDs that have high level peaks in their one third-octave-band sound power
spectrum. Such peaks usually indicate the presence of discrete frequency tones
in the spectrum. Such tones can be difficult to eliminate and generally are a
source of annoyance.
6. Select a fan or blower having the lowest speed and largest
diameter consistent with the other requirements.
7. Minimize system noise levels by designing the system so that
obstructions are not present within one fan diameter of the inlet to axial-flow
fans [6], so that the airflow into the inlet of axial-flow fans is as spatially
uniform as possible. Avoid the direct attachment of the AMD to lightweight sheet
metal parts.
8. Mount axial-flow fans so that the air-flow direction is towards
the equipment being cooled. Pulling air over equipment being cooled usually
causes undesirable turbulence at the fan inlet and produces an increase in noise
level.
Limit Values for Sound Power Levels
The wide variety of electronic equipment cooled by AMDs makes it impossible
to specify a set of limit values for sound power levels which are applicable to
all equipment. However, there are two sets of requirements which have been
adopted in Europe for data processing equipment. In Germany, the German
Institute for Quality Assurance (RAL) has been authorized to grant the
environmental logo Blue Angel to certain products, including workplace computers
and workstations consisting of a system unit, keyboard, and monitor. These
environmental requirements are mostly concerned with recycling; there are,
however, acoustical requirements specified. A "non-official"
translation of the acoustical requirements into the English language reads:
"In accordance with section 3.2.5 of ISO 9296 [7], the "declared
sound power level, LWAd, of the components, measured when idling and
multiplied by 10, must not be more than 48 dB(A). In other conditions of
operation (access to diskette or hard disk) the maximum value must not exceed 55
dB(A). The measurements are to be performed in accordance with DIN EN 2 7779."
The multiplication by 10 mentioned in the above paragraph is required for
the conversion from bels (as specified in ISO 9296) to decibels. The
measurements are to be made by an independent "test house", or by an
applicant whose test facilities and procedures have been certified by an
independent certification body in accordance with a European standard EN 2 9000
or ISO 9000.
In Sweden, Statskontoret Technical Standard 26:3 specifies recommended upper
limits for the noise emissions of computers and business equipment. These
requirements are summarized in Table 1.
| Product Category |
Product Description |
Recommended upper Limit Sound Power Level in
bels |
| LWAd Operating |
LWAd Idling |
Category I Equipment for use in dedicated rooms |
A. All products |
7.0 + K |
7.0 + K |
Category II Equipment for use in general business
areas |
A. Fully-formed character typewriters and printers |
7.2 |
5.5 |
| B. printers and copiers (more than 4mm distance from workstations) |
7.0 |
6.5 |
| C. Tabletop printers and tabletop copiers |
7.0 |
5.5 |
| D. Processors, controllers, disk & tape drives, etc. (more than 4m
distant from workstations) |
7.0 |
7.0 |
| E. Processors, controllers, disk & tape drives, etc. (more than 4m
distant from workstations) |
6.8 |
6.6 |
Category III Equipment for use in quiet office
areas |
A. Printers, typewriters, and plotters |
6.5 |
5.0 |
| B. Keyboards |
6.2 |
N/A |
| C. Floor-standing processors |
6.0 |
5.5 |
| D. Tabletop processors, controllers, system units including built-in disk
drives and/or tapes, display units with fans. |
5.8 |
5.0 |
| E. Display units (no moving parts) |
4.5 |
4.5 |
| Note: K=lg (S/So) where So
is equal to one meter, and S is the footprint in square meters, i.e., the
projection in square meters of the machine on the floor. If S < 3 square
meters, use S = 3. The calculated value of the recommended upper limit may be
rounded to the nearest upper 0.1 bel |
Table 1: Swedish recommended upper limits
for declared sound power level values. Technical standard 26:3. Statskontoret,
Swedish Agency for Administrative Development, Stockholm, Sweden. First day of
validity: 1993-05-01. In case of conflict, the Swedish text prevails over the
English text in the table.
Noise Measurement for Personal Computers
National and international standards have been developed for determination
of the sound power level emitted by machinery and equipment [2]. The computer
and business equipment industry has developed procedures which can be adapted
for the measurement of the noise emissions of a wide variety of electronic
equipment. While it is possible to determine sound power in ordinary rooms
through the use of sound intensity [2], it is more common to install the
equipment in a special test room, a hemi-anechoic room or a reverberation room.
When one of these facilities is available, the following seven-step procedure is
useful for determination of sound power level.
1. Install the personal computer system in test chamber. The test
chamber may be a hemi-anechoic room or a reverberation room. Install the
personal computer in accordance with ISO 7779 on the floor of a hemi-anechoic
chamber or a reverberation room for determination of sound power levels.
2. Calibrate the measurement system and check the frequency
response. Calibration and frequency response should be checked on a weekly
basis, or just prior to and after a set of measurements is completed.
3. Measure the background sound pressure and determine the
background sound power level. These measurements are made with the personal
computer system turned off.
4. Power up and warm up the equipment so that all initialization is
complete and it is operating in the steady state before acquiring any data.
5. Determine the personal computer sound power levels in idle and
operating modes. See ISO 7779, Annex C, for further details of modes of
operation.
6. Measure the personal computer sound pressure levels in idle and
operating modes. Desktop units should be placed on a standard test table (see
ISO 7779, Annex A). With the personal computer in the same modes of operation as
in step 4 above, the by-stander and operator position sound pressure levels
should be measured. Although sound power can be determined in either a
reverberation room or a hemi-anechoic chamber, sound pressure determinations
must be done in the latter (see ISO 7779, section 7).
7. Perform a discrete tone analysis at the operator position. The
existence of discrete frequency tones in the spectrum often requires that an
additional measurement be performed to assess the prominence of the tone(s).
Tonal prominence is related to annoyance and the psychoacoustics of how discrete
tones are perceived in the presence of noise. Annex D of ISO 7779 specifies the
detailed procedure.
Conclusion
Designers of electronic systems cooled by air must always be conscious of
the acoustical noise emitted by the system if people are exposed to the noise.
The three elements required for a low noise design have been described in this
article, selection of air moving devices and design of systems for low noise,
quantitative limits on the noise emissions of equipment, and measurement
procedures which can be used to determine the sound power level of the
equipment.
Footnote: The material in this article is taken from a
chapter in the Handbook of Thermal Measurements in Electronics Cooling, by K.
Azar, CRC Press, 1996.
| Dr. George C. Maling, Jr. |
Dr. David M. Yeager |
| INCE/USA |
Motorola Corp., Room 2319 |
| P.O. Box 3206 Arlington Branch |
8000 W. Sunrise Blvd. |
| Poughkeepsie, NY 12603, USA. |
Fort Lauderdale, FL 33322, USA |
| Tel: +1 914 462 4006 |
Tel: +1 954 723 4997 |
| Fax: +1 914 463 0201 |
Fax: +1 954 723 4334 |
| Email:INCEUSA@aol.com |
Email:edy003@email.mot.com |
References
1.Glossary of terms used in noise control engineering,
Noise/News International, 3, 161-168, 1995.
2.Beranek, L.L. and Ver, I., Editors, "Noise and vibration control
engineering", John Wiley & Sons, Inc., New York, 1992, Chapter 4 ,
Determination of sound power levels and directivity of noise sources.
3.ISO 10302, Acoustics - Method for the measurement of noise emitted by
small air-moving devices, International Organization for Standardization,
Geneva, Switzerland, 1995.
4.ISO 3744: Acoustics - Determination of sound power levels of noise
sources - Free field conditions over a reflecting plane, International
Organization for Standardization, Geneva, Switzerland.
5.ISO 7779, Acoustics - Measurement of airborne noise emitted by computers
and business equipment, International Organization for Standardization, Geneva,
Switzerland, 1988.
6.Washburn, K.B. and Lauchle, G.C. 1988. Inlet flow conditions and tonal
sound radiation from a subsonic fan, Noise Control Eng. J., 31, 101-110.
7.ISO 9296, Acoustics - Declared noise emission values of computer and
business equipment, International Organization for Standardization, Geneva,
Switzerland, 1988
Web page conversion/development by
InfoVision Technologies, Inc. sales@ivtinc.com
|