Anechoic Chamber Design and Construction
Web: http://www.pearl-hifi.com 86008, 2106 33 Ave. SW, Calgary, AB; CAN T2T 1Z6 E-mail: [email protected] Ph: +.1.403.244.4434 Fx: +.1.403.245.4456 Inc. Perkins Electro-Acoustic Research Lab, Inc. ❦ Engineering and Intuition Serving the Soul of Music
Please note that the links in the PEARL logotype above are “live” and can be used to direct your web browser to our site or to open an e-mail message window addressed to ourselves. To view our item listings on eBay, click here. To see the feedback we have left for our customers, click here.
This document has been prepared as a public service . Any and all trademarks and logotypes used herein are the property of their owners. It is our intent to provide this document in accordance with the stipulations with respect to “fair use” as delineated in Copyrights - Chapter 1: Subject Matter and Scope of Copyright; Sec. 107. Limitations on exclusive rights: Fair Use. Public access to copy of this document is provided on the website of Cornell Law School at http://www4.law.cornell.edu/uscode/17/107.html and is here reproduced below:
Sec. 107. - Limitations on exclusive rights: Fair Use
Notwithstanding the provisions of sections 106 and 106A, the fair use of a copyrighted work, includ- ing such use by reproduction in copies or phono records or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for class- room use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include:
1 - the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
2 - the nature of the copyrighted work;
3 - the amount and substantiality of the portion used in relation to the copy- righted work as a whole; and
4 - the effect of the use upon the potential market for or value of the copy- righted work.
The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors
♦ PDF Cover Page ♦ ♦ Verso Filler Page ♦ RESEARCH DEPARTMENT
The design. of a new free-field sound measurement room: The selection of sound absorbent material
RESEARCH REPORT No. L-055 1964/42
THE B R I TIS H B R 0 A 0 CAS TIN G COR P 0 R A TI 0 N
ENG I NEE R I N G 0 I VI S ION This Report is the propertf or the British Broadcastin& Corporation and aa1 not be reproduced in anf tOrm without the written pera188ion ot the Corporatlon.
li
Research Report No. L-055
(1904/42 )
THE DESIGN OF A NEW FREE- FIELD SOUND MEASUREMENT ROOM: THE SELECTION OF SOUND ABSORBENT MATERIAL
Section Title Page
SUMMARY ..' ...... 1
1. I NTRODUCTI ON...... 1
2. GENERAL . . . · ...... 1
3. FIBROUS MATERIALS 2
4. FOAMED PLASTIC MATERIALS...... 4
5. CONCLUSIONS . . . · ...... 6
6. ACKNOWLEDGEMENTS. · ...... 8
7. REFERENCE . . . . · ...... 8
iv July 1964 Research Report No. L-055
(1964/42)
THE DE SIGN OFANEW FREE-FIELD SOUND MEASUREMENT ROOM: THE SELECTION OF SOUND ABSORBENT MATERIAL
SUMMARY
In designing wedge-type sound absorbers for use in a new free-field room, various kinds of plastic foam were considered as possible alternatives to glass wool or glass fibre. The reflexion coefficient of the plastic materials was found to be influenced by mechanical resonances in the material, which in turn were markedly affected by the method of �ounting the wedge. The material finally adopted was a polyurethane ether foam having sufficient mechanical hysteresis to damp the internal reso nances. The introduction of additional mass in the form of a steel rod at the centre of each wedge was found to extend the low-frequency range.
1. INTRODUCTION
In the new sound measurement room built in 1963 at Kingswood Warren, it was desired to ensure substantially free-field conditions at frequencies down to 50 cis. The acoustic treatment in the existing sound measurement room consists of glass fibre wedges 8 in. x 8 in. (20 cm x 20 cm) at the base, and occupies a depth of 3 ft 6 in. (101 m); from experience with this room, it was estimated that the required performance in the new room could be obtained by similar means by allowing a total of 5 ft (105 m) for the length of the wedge plus the air space, if any, between it and the wall. However, as an increase in length of wedge would in theory alter the optimum value of flow resistance, it was necessary to experiment with materials other than the original grade of glass fibre.
2. GENERAL
Measurements of reflexion coefficient for normal incidence were made by means of a travelling-wave duct having an internal cross-section 16 in. x 16 in. (40 cm x 40 cm) thus accommodating four wedges. Although the base cross-section of the wedge was kept constant throughout the experiments, the remaining dimensions were varied; for converiience, the length of the tapering p�rtion of the wedge, the length of the parallel portion, the air gap between the base of the wedge and the wall and the overall depth of the acoustic treatment are designated, as shown . in Fig. 1, as A, B, C and D respectively.
1 The lower limit of the useful D frequency range of an absorbent material c.,;-.------for a free-field room is usually taken as the point at which the reflexion co efficient rises to 10%. To facilitate c omparison between different types of s ound absorber occupying different 8 A ------I l- -...... i.----- amounts of space, some workers have taken as a figure of merit the value of DIA Dimensions of wedge and position o Fig. 1 - where is the wavelength, in air, of relative to wall. Ao sound at that frequency for which the reflexion coeffici ent is 10%. This pro cedure, however, produces a figure which decreases when the working frequency range of the absorber increases; to avoid this incongruity, the inverse quantity AolD will be used in the present report as a measure of the efficacy of various types of acoustic treatment.
For the sake of brevity and to preserve anonymity , commercial brands of absorbent material have been designated throughout by Roman numerals.
3. FIBROUS MATERIALS
The first group of materials to be investigated were of the fibrous type, in which the absorption of sound depends principally on the flow resistance. Accord ing to theory, the curve of reflexion coefficient as a function of frequency should exhibit signs of interference between the wave reflected from the front of the wedges and that reflected after passing through the wedges and reaching the rigid wall behind them. Fig. 2(a) shows the theoretical form of reflexion characteristic given in a paper by Walther l in 1960 and Fig. 2(b) the reflexion characteristic of a sample of low-density glass fibre, in which the effects of interference are likewise apparent.
50 50
a} Theoretical curve from Walther 1960 b) Experimental data - glass fibre 40 40 -! \V -! ..:0 ..:0 / c , c � � 30 30 V .....� .....� ...... \ 8 If" 8 u I u c: 20 r\ c: 20 \ �r-. Q Q )( )( \ V ...... !! \ I � ...... !! � � � L L \ ,; 10 \ II 10 \ \ / \/\ �I\. V '" o o 1"\ 30 50 100 500 30 50 100 500 freauenc V. cis frequency, cIs
Fig. 2 - Reflexion coefficient of low density porous wedges. 2 Although the reflexion characteristics of absorbent wedges do not always follow the theoretical fonn as closely as in the examples just given, the same general principles appear to hold. The experiments on fibrous materials yielded a number of conclusions, applicable to a wide range of materials, regarding the effect of varylng the dimensions of the absorbent wedges and their spacing from the wall.
Fig. 3, for example, shows the effect of varying length of taper A on the reflexion characteristic of mineral wool wedges and Fig. 4 the effect of varying the air gap C for low-density glass-fibre wedges. Fig. 5 shows, for mineral wool wedges, the effect of varying the ratio C/B (air gap to parallel portion), keeping the rate of taper and overall depth of treatment constant.
Reflexion Coefficients
5 0 I taper length A (25cm) air C o�� r...... JOin gap .... . " �V 16in (41cm) Oin (Oem) � \100.. \ \ \ \ � '. -, 1';b)4in(61cm) 1--4in (1Qcm) ,.-...... 1-- ... - '\ 1\ 1\ \ .....8 in ��2jn (B1cm) ..: (20cm) '\ \ .( � r 1-12in (31cm) r... 40in (102cm) \ \ \- ...,... /1I1/\� ,...�4in(61cm) \ \ " l� 36ln(92em) u � � 1\1 , II � \' '\. \ \ � Ij �,j ,� r'�, " i 0 �.� \ 1\ '" 1'0.' '\ , �::"".. '\ " ).�� � , �-r\. � \." l...!!::: �-lIII �...... '''1'-.' � ...... \. r� 0 o 30 100 300 1000 30 100 300 1000 • frequ�ncy, cIs frequency, cIs Fig. 3 - Variation with length of taper 'A' for wedges Fig. 4 - Variation with depth of airgap 'C' behind of mineral wool. wedges of low density glass fibre.
,.., 5 5 0
0 0 a) - Glass fibre = 60" (150)cm t-. I, '0'
1\' parallel 01\ '1\ air gop C portion B £. ..0 \ ./ Oln Fig. 5 - Variation of ratio of air gap 'C' to parallel Fig. 6 - Wedges having optimum dimensions, low portion 'B '. Length of taper 'A' and total density glass fibre_ depth constant. Wedges of mineral wool. 3 Of all the materials of this group, the best was found to be a particular grade of resin-bonded glass fibre, I, for which the reflexion characteristic with optimum wedge dimensions is given in Fig. 6 curve (a); Ao/D is 4°3. Unfortunately, at this stage in the investigation the commercial production of this material ceased and the nearest substitute type glas s fibre available, II, gave the characteristic, also shown in Fig. 6, curve (b), for which Ao/D is only 3'3. Attempts were made'to reduce the reflexion coefficient at low frequencies by constructing behind each wedge a Helmholtz resonator, the inlet of which was accommodated in a gap between adjacent wedges. No worthwhile improvement in performance was, however, achieved by this artifice. 4. FOAMED PLASTICS MATERIALS Attention was then turned to foamed plastic materials, 'which have the advantage of being free from dust and not easily damaged in handling. The general principles , illustrated above, which govern the choice of wedge dimensions appeared to apply als o to this class of absorbent material. A wide variety of plastic foams, both rigid and flexible, was tested, but in every case in which the performance looked promising, the reflexion characteristic was found to be profoundly affected by mechanical resonance in the body of the material. Fig. 7 shows, by way of example, the reflexion characteris tic of a polyurethane foam, III, together with the output of an accelerometer applied to the wedge; for this material, if it were possible to ignore the ris e in reflexion coefficient in the region of resonance, Ao/D would be 3°8. For comparis on, the reflexion characteristic of mineral wool wedges for which Ao/D is 3'9 is also shown. Mechanical resonance phenomena in the abs orbent material are naturally affected by mechanical constraints, and the influence of these resonances on the reflexion characteris tic thus depends a good deal on the method of mounting the wedges. As an example of this , Fig. 8 shows the effect of partially attaching polyurethane ester wedges, IV, to the wall by adhesive. The effect of the additional constraint is to reduce the value of Ao /D from 4'3 to 2'4. Reflexion Coefficients 50 0 .-J Reflexion coefficients a) Mineral wool 0 --y({ a) Wedges are free to move. -IX�)Polyurethane III ,..... �{ '0' = 48" (120cm) '\'"' 1 b) Bases of wedges glued � c) Acceleration of 1\ "'\/ to wall over part of 1\ \ / body of wedge III i.., their area. 0 \ \ , \ \ , I " /� 0 1\ � � � � \ '0' = 48" (120cm) !'W, .' lV '"'" o 0 �I 30 �=3'9 *=3.8100 300 1000 30 100 300 1000 frequency. cIs 8=4·3 *=2'4frequency. cIs Fig. 7 - Wedges having optimum dimensions. Fig. 8 - Polyurethane ester wedges IV for two boundary conditions. 4 Most foam plastics have very little mechanical hysteresis; their internal resonances are therefore lightly damped and the effects of these are subject to wide variations in manufacture. As an example, curves (a) and (b) in Fig. 9 show the reflexion characteristics of wedges from two different batches of polyurethane ether foam, V, for which the best value of A is 4-5. olD Fortunately, in the course of the investigation a new type of polyurethane ether foam, VI, having an appreciable amount of mechanical hysteresis, became available. Fig. 10(a) shows the reflexion coefficient obtainable with this material, for which 4-3; VI AoID is the effect of the increased damping is apparent. Foam was found to be sufficiently reproducible in quantity production and was therefore adopted for use in the new free-field room. It was at first intended to mount the wedges by impaling each one on a � in. (6 mm) diameter steel rod projecting from the wall, floor or ceiling. This form of support had, however, to be abandoned as the mechanical constraint so intro duced adversely affected the performance of the wedges. Nevertheless, it was found that by leaving the steel rod in place but not attached to the wall, the value of A lD could be increased, as shown in Fig. 10(b), to 4-S. Experiments made with rods oof other materials showed that the mass of the rod rather than the stiffness was responsible for this effect. Reflexion Coefficients 50 5 0 ,.. ..., ';40 = ' 0 ' 60" (150cm) ... \\ '0' = 60" (150cm) . 1 0 �3 0 l� �a) no rod � \ V u I I I III \;,(b)with 38in (g7cm) rod 0 0 • , \Cb) �. 1; I " , , ,� '" 0 ,- V, "" 1\ \ 1. " .. ... -- ..--- " Fig. 9 - Polyurethane ether type V from different Fig. 10 - Polyurethane ether wedges VI as used batches. for free-field room, with and without steel rod. Measurement of the reflexion coefficient of wedges of foam VI was also made at frequencies above SOO cis, the limit imposed by the onset of transverse resonance modes in the 16 in. x 16 in. (40 em x 40 em ) test duct. To this end, a smaller duct, of internal cross section 4 in. x 2 in. (10 em x 5 em), was built, which allowed measurements to be made up to 4 kc/s. It was not possible to accommo date a complete wedge in this duct, but a sample was taken from the centre of the wedge, including the tip. The results thus obtained have been combined with the data from Fig. 10(b) to give the overall curve in Fig. 11. S 50 o 40 .. ... \ c , = ' 0 ' 60" (150cm) �30 � \ � o u c 20 o j \ -QI , L. 10 � �, '- ...... o :---... 2 30 10 1 o 3 l..4.e frta.Quency, cIs o Fig. 11 - Typical reflexion coefficient of polyurethane ether wedges VI as used for free-field room. To give some idea of the degree of reflexion to be expected when the sound approaches a bank of wedges at nearly grazing incidence, the tests above 800 cis were repeated using sections of a wedge arranged with their side surface normal to the direction of sound incidence. The reflexion coefficient at frequencies up to 205 kcls was less than 8% and fell at higher frequencies, as shown in Fig. 12. As plastic foam is rather soft, it was necessary to provide some support to prevent excessive sagging. The method of achieving this is shown in Fig. 13. The position of the point of support was found to be important; any additional constraint in the neighbourhood of the wedge tip increased the reflexion coefficient in the region of 90 cis; a similar effect appeared, as shown in Fig. 14, if the length of the steel rod was made too great. To keep a check on the production variations of the material, sample wedges made from about 2% of each batch were tested before the remainder of the material was cut. 5. CONCLUSIONS Not all types of plastic foam are suitable for sound absorbent purposes since their mechanical damping is fre· quently too low and the acoustic properties of the material may be very variable in large scale production; in fairness to the manufacturers, however, it should be remembered that most of these materials were never intended for sound absorption. o�----�--��------� Certain types of poly urethane �3 �4 ether foam are, however, equivalent, as far frequency, cis as can be discovered from small scale Fig. 12 - Reflexion coefficient from side of poly tests, to the best of the fibrous absorbent urethane ether foam wedge VI at high freq uencies. materials. Final judgment on this matter 6 must await full-scale tests on the new free-field room which IS now approaching completion. Fig. 13 - Method of supporting free-floating wedges to prevent sagging. 50 0 •,e4 ' 0 ' = 60" (150cm) ..... \ , 0 'aJ with 38in (�7cm) rod r- wIth 431n (1Q9cm) rOd ,.. lb) �- l� V '" "'-: 0 30 100. 300 1000 to: 4·8 frequency. cis Fig. 14 - Reflexion coefficient of polyurethane ether VI showing effect of increasing length of steel rod. 7 6. ACKNOWLEIXiEMENTS Thanks are due to Mr. W.I. Manson and Dr. D.J. Neale, tqgether with Messrs. R.L. Deane and K.H. Hewitt, for their assistance with the measurements. 7. REFERENCE 1. Walther, K., 'Reflection factor of Gradual Transition Absorbers for Electromagnetic and Acoustic Waves', I.R.E. Transactions on Ant ennae and Propagation, November 1960, page 608. Printed by BBC Research Department, Kingswood Warren, Tadworth, Surrey 8 PREVIOUSLY ISSUED NUMBERS OF BRUEL & KJ/ER TECHNICAL REVIEW 1-1964 Statistical Analysis of Sound Levels. 2—1964 Design and Use of a small Noise Test Chamber. Sweep Random Vibration. 3—1964 Random Vibration of some Non-Linear Systems. 4—1964 The Accuracy of Condenser Microphone Calibration Methods. Part I. 1—1965 The Accuracy of Condenser Microphone Calibration Methods. Part II. 2-1965 Direct Digital Computing of Acoustical Data. The Use of Comparison Bridges in Coil Testing. 3—1965 Analog Experiments Compare Improved Sweep Random Tests with Wide Band Random and Sweep Sine Tests The Frequency Response Tracer Type 4709. 4—1965 Aircraft Noise Measurement, Evaluation and Control. 1—1966 Windscreening of Outdoor Microphones. A New Artificial Mouth. 2—1966 Some Experimental Tests with Sweep Random Vibration 3—1966 Measurement and Description of Shock. 4—1966 Measurement of Reverberation. 1-1967 FM Tape Recording. Vibration Measurements at the Technical University of Denmark. 2—1967 Mechanical Failure Forecast by Vibration Analysis. Tapping Machines for Measuring Impact Sound Transmission. 3—1967 Vibration Testing — The Reasons and the Means. 4—1967 Changing the Noise Spectrum of Pulse Jet Engines. On the Averaging Time of Level Recorders. 1—1968 Peak Distribution Effects in Random Load Fatigue. TECHNICAL REVIEW No. 2 - 1968 ♦ Verso Filler Page ♦ The Anechoic Chambers at the Technical University of Denmark By F. Ingerslev, Prof. Dr., O. Juhl Pedersen, M. Sc, and P. K. Mailer, M. Sc, The Acoustics Laboratory, The Danish Technical University, Lyngby, Denmark J. Kristensen, M. Sc, The Danish Academy of Technical Sciences, Lyngby, Denmark ABSTRACT In connection with the extension of the Technical University of Denmark at Lyngby the Acoustics Laboratory has aquired new and larger measurement facilities. Two of the new rooms are built as anechoic chambers using wedge-type construction. The acoustical design of these rooms as well as a number of investigations carried out on the wedge-construction itself are described and discussed. Mention is also made of the electrical installation, the ventilation system and the measurement cable installation. Finally some measurements on the completed rooms, both with regard to their free-field performances, and with regard to sound and vibration insulation are described. These measurements show that the rooms can be used even for very exacting measurements, and under extreme outside conditions. SO M MAI RE En correlation avec I'extension de I'Universite Technique du Danemark a Lyngby, le Laboratoire d'Acoustique a acquis de nouvelles et plus amples installations de mesure. Deux des nouvelles salles sont construites en chambres anechoiques par le recours a une structure du type a coins absorbants. La conception acoustique de ces chambres est decrite et discutee, comme un certain nombre d'investigations menees sur I'emploi de coins absorbants. II est aussi fait mention de Installation electrique, du systeme de ventilation et de I'installation du cablage de mesure. Pour terminer sont decrites quelques mesures effectuees sur les chambres parachevees, tant sous I'aspect du fonctionnement en champ libre que sous celui de ['isolation aux sons et aux vibrations. Ces mesures montrent que les chambres peuvent etre utilisees meme pour les mesures les plus ardues et dans des conditions exterieures extremes. ZUSAMMENFASSUNG Es werden zwei schalltote Raume beschrieben, welche bei der Erweiterung der Danischen Technischen Universitat fur das Akustische Laboratorium im neuen Hochschulzentrum bei Lyngby eingerichtet wurden. Beide Raume sind mit keilfdrmigen Absorbern ausgestattet, deren Eigenschaften besprochen werden. Erwahnt wird auch die elektrische Installation, die Beliiftung, sowie die Verlegung der MeGkabel. Endlich sind Messungen in den Raumen aus- gefuhrt, die sich auf die Freifeld-Eigenschaften sowie auf die Luft- und Korperschalldam- mung beziehen. Die Ergebnisse zeigen, daB die Raume selbst bei ungewohnlichen Urn- gebungsbedingungen fur genaue Messungen verwendbar sind. Introduction In connection with the extension of the Technical University of Denmark at Lyngby just north of Copenhagen, the Acoustics Laboratory has acquired three new buildings with a number of special purpose rooms (two anechoic cham- bers and six reverberation rooms). The new buildings were taken into use in 1965. The two anechoic chambers, which are situated in a separate building, were constructed as one large room for precision measurements and research 3 Fig. 1. The farge anechoic room. purposes (Fig. 1), and one smaller room for routine measurements, for example in connection with the undergraduat� teaching. The a.nechoic chambers are designed j o intly by the architects and engineers constructing the new technical university and the staff of the Acoustics Laboratory. Fig. 2 shows a plan of the building with the anechoic chambers and the associated rooms. Holl Store ,,1-0'1'1 Fig. 2. Plan of the building with the anechoic rooms. 4 Design Considerations Dimensions. The dimensions of the large anechoic room were chosen mainly with regard to the largest measuring distance to be used. As it was desirable to be able to measure in the far field of large sound sources such as loudspeaker columns of 1-3 m length, measuring distances of up to 15 m were considered suitable. Another factor which also influences the room dimensions is the lower limiting frequency, i.e. the lowest frequency for which a free field is obtained in the room. This frequency is not only determined by the nature of the material covering the walls, but also by the ratio between room dimensions and wave- length. From the theory of sound transmission in heavily damped ducts we know that the condition for undisturbed sound transmission parallel to heavily damped surfaces is that X ^ 1.2e-c where h is the width of the duct, I is the wavelength and z is the acoustic impedance of the walls. With a view to the electro-acoustical measurements, which are to be carried out in the room, the lower limiting frequency was chosen to be 50 Hz, which gives h > 6 m. As a result of all the considerations to be taken the free part of the room was fixed at 12.1 m X 9.7 m X 8.5 m. The free part of the smaller anechoic room was fixed at 4.8 m X 4.1 m X 2.9 m, which should give a lower limiting fre- quency of about 100 Hz. Wall lining. The absorption lining of anechoic rooms can have widely different thicknesses, depending upon the size of the room and the frequency range of the measure- ments to be carried out, and widely different absorption qualities depending upon the type of material used and its physical structure. When investigating the lining of anechoic rooms the concept of the reflection coefficient is usually employed. This is the ratio between the sound pressure of the reflected sound and the incident sound. For a suitable absorption lining the reflection coefficient is generally a decreasing function of frequency and the lowest frequency for which the reflection coefficient is 0.1 is usually taken to be the limiting frequency. This term is used in the following. The lining of anechoic rooms can be constructed in different ways. In recently built rooms the lining is usually made up of wedges of mineral wool or similar absorbing materials. Other types of construction are of course possible. How- ever, on the basis of literature studies, economical considerations etc., it was decided to concentrate further studies upon wedge constructions only. The investigation of the reflection qualities of absorbing materials should be carried out for various angles of incidence. Measurements at normal incidence 5 may be done with the tube method, which is well known. The measurements at other angles of incidence require extensive instrumentation and suitable measuring rooms, as well as a reliable measuring method and a clear defini- tion of the quantities to be measured. A theoretical consideration of the possibility of measuring the reflection in a free field shows that the necessary test wall should have dimensions several times the wavelength at the lowest measuring frequency. This resulted in a decision to carry out model tests in scale 1 :8 with a wedge shape which was determined on the basis of tube measurements. In order to investigate the influence of the wedge shape in addition to the influence of the absorption coefficient of the wedge material, measurements were taken with wedges made of Sillan (mineral wool) as well as of lacquered wood. The main results of these model experiments were that the reflections seemed to follow Snell's law, that variations in the wedge pattern had negligible influence upon the reflected energy, and that maximum reflection was obtained for sound incidence parallel to the wedge axis. On the basis of these results it seems reasonable to give more importance to measurements on wedges by the tube method than would have been given without the know- ledge gained from the model tests. This conclusion is undoubtedly only valid for wedge linings, and is probably quite wrong for linings such as layers of woven material. The tube measurements were carried out in a brick tube of dimensions 60 cm X 60 cm, especially built for the purpose, of 23 cm bricks. The tube is constructed vertically in an empty lift duct. A sketch of the tube is shown in Fig. 3. Access to the tube is made at the bottom through a steel door with a Fig. 3. Duct for reflection measurements on wedges. 6 4-5 cm concrete lining. The tube is not acoustically entirely loss free, but measurements and calculations show that the inherent damping has negligible influence upon the results obtained. The initial measurements were carried out with wedges made in the laboratory, of glued 2 cm thick layers of Sillan of specific weight 100-110 kg/m3. Some of the tests were repeated later with Sillan wedges made in a factory. Fig. 5. The lower limiting frequency of the wedge lining as a function of tan (v). The first part of the investigation was carried out in order to see if the wedge angle v, see Fig. 4, had any influence on the absorption, and if so, to find the angle which would give the lowest limiting frequency. The results from these measurements are shown in Fig. 5. The total length of the wedges was kept constant and, except at the limiting wedge angles, also the height of the base as well as the basis length to width ratio, while the wedge angle and base sides were varied. The figure shows the limiting frequency as a function of the tangent to the wedge angle, tan (v). It can be seen that there is a relatively wide range of angles for which the limiting frequency is lower than for the remaining angles. In order to determine the influence of the length of the wedges, measurements were taken with varying wedge length and constant wedge angle. The results 7 Fig. 6. The reflection coefficient of the wedge as a function of frequency with wedge length as parameter. are shown in Fig. 6, where the reflection coefficient is plotted as a function of frequency. From considerations of the relationship between thickness and absorption of porous materials, one would expect the limiting frequency to be inversely proportional to the thickness of the lining, which is also demonstrated by these results. Fig. 7. The reflection coefficient of the wedge as a function of frequency with base length as parameter. 8 Small variations in base width 6b seem to have negligible influence on the reflection coefficient, neither did measurements on wedges of parabolic shape give any significant improvement, in fact parabolic wedges would probably have an economic disadvantage because of difficult cutting and larger wastage of material. The base length 6L may have some influence on the limiting frequency, as can be seen from Fig. 7. All dimensions except the total length of the wedge are here kept constant. It is seen that the lowest limiting frequency is obtained with a base length of about 10 cm, for wedges of length U = 40 cm. One might perhaps expect that a built-in resonator below the base of the wedge would improve the absorption of a wedge lining. Fig. 8 shows the re- sults from measurements on wedges placed against a hard termination and in front of a Helmholz slit resonator of resonance frequency 80 Hz, which is just below the lower limiting frequency for the wedges placed against a hard termination. Fig. 8. The reflection coefficient as a function of frequency for wedges placed in front of a Helmholtz resonator. Masses corresponding to the wedge weight were placed on the front plate of the resonator during tuning. As can be seen from the figure the combination of wedge and resonator gave rather discouraging results. Other experiments gave similar results. At this point in the investigations it was necessary to fix the wedge module and length, and the dimensions 145 cm X 30 cm X 30 cm with a base length of 25-30 cm and 100 cm X 24 cm X 24 cm with a base length of 15-20 cm were chosen for the large and the small anechoic room respectively. This 9 of course influenced the following part of the investigation. The chosen dimensions give wedge angles of 13°-17°. In order to gain experimental evidence of the suitability of different materials, measurements were carried out on a variety of glass and mineral wool samples of different specific weight, and on foamed poly-urethane-ether AOP 35. In addition to measurements of absorption coefficient load tests were carried out on the mineral wool wedges in order to compare their mechanical strength. The tests were carried out with the wedges placed in trays of perforated plates and fixed with rods going through the perforation and the base of the wedges. A load of 100 grammes placed at the tips of horizontally mounted wedges for a week gave less than 1 cm deflection for all the mineral wool wedges except for lightweight Rockwool wedges. Corresponding tests were made with vertically mounted wedges and a load pulling the wedges out. All the wedges except the light Rockwool wedges could stand an extra weight corresponding to their own weight without being pulled out. Loads of twice the wedge weight could be carried by the glass wool types and the heavy Rockwool wedges. Without any proper testing of the ability of the different wedges to withstand mechanical handling, it can be said that except for the light glass wool wedges which could take several large deformations without damage, they were all very brittle and would easily break. The wedges made of foamed poly-urethane-ether could withstand large de- formations without any damage. They would show a considerable extension before breaking, but a very low stiffness made additional support of the wedge tips necessary for horizontal mounting. Fig. 9. The reflection coefficient as a function of frequency for wedges of different materials. 10 Fig. 10. The reflection coefficient as a function of frequency for wedges of different materials. Figs. 9 and 10 show the reflection coefficient as a function of frequency for the different materials. On the basis of these investigations four different manufacturers were invited to give in tenders for the supply and mounting of wedges in the two planned anechoic chambers, as the opinion was that the price should also be taken into account in the total considerations. After a thorough testing of wedges from the two lowest quotations, Sillan wedges of specific density 120 kg/m3, manufactured by Grunzweig und Hart- mann, Ludwigshafen (Germany) were chosen as the best alternative. Fig. 11 shows the reflection coefficient as a function of frequency for the Fig. 11. The reflection coefficient as a function of frequency for wedges used in the anechoic rooms. 11 wedges chosen. The measurements at frequencies above 250 Hz are carried out in a smaller tube than that shown in Fig. 3. Both whole wedges and parts of wedges were investigated. In order to obtain an estimate of the maximum departure from a free field to be expected in an anechoic room lined with these wedges, a digital computer was used to calculate the sound field for a point source in rooms of different size and with different values for the reflection coefficients of the walls. The calculations were made with the assumption that the sound waves were reflected in accordance with Snell's law at the outer walls of the room and that phase coincidence of the reflected sound existed at the measuring point. Calculations were carried out for different positions of the sound source. With these assumptions and a reflection coefficient of 0.1 one should have a departure from free field of maximum ± 1 dB at 2 m and ± 3 dB at 11 m in a room of free dimensions of 12 m X 10 m X 9 m. Sound insulation As the anechoic rooms were also intended for a number of psycho-acoustical measurements, it was desirable to have a background noise level of about 10 dB below the threshold of hearing at frequencies above 200 Hz, whereas for frequencies below 200 Hz a level lower than 10 dB re 2 X 10"5 N/m2 was required. 4 For extraordinary outdoor sound levels, such as those caused by jet aircraft at low altitudes, slightly higher background levels would be acceptable. It would, however, be an advantage if normal acoustical measurements were not appreciably disturbed even in these cases. These requirements necessitate a wall construction, which has a transmission loss of about 80 dB at 200 Hz, as a background level of 60-70 dB re 2 X 10"5 N/m2 must be expected outdoors during working hours. In order to achieve this, a double wall construction must be used where the inner part, the anechoic room itself, has no stiff mechanical connection to the outer building. The outer walls of the building (Fig. 2) consist of brickwork and reinforced concrete with a layer of heat insulation between, total thickness 30 cm. The roof is made of 20 cm reinforced concrete with heat insulation. The inner boxes, with the anechoic rooms, have walls, floor and roof made af 40 cm reinforced concrete. The space between the inner and outer walls is 1.1 m and acoustically damped by covering both the inner side of the outer shell as well as the outer side of the inner shell with 5 cm wood wool cement slabs. Also the inside of the outer roof is covered with this material. The large and the small room are placed on 24 and 4 rubber vibration isolators respectively, see Fig. 12. Each vibration isolator is loaded by about 50 tons, and a resonance frequency of about 7 Hz was intended for the rooms placed on the rubber pads. The vibration isolators, made from a mixture of natural and artificial rubber (hardness 50° shore) are protected by a layer of neoprene and placed so that they can be inspected and replaced if necessary. 12 F/g. 72. Vibration isolators for the large anechoic room. Doors The doors are important parts of the construction with regard to lining and sound insulation. For the large anechoic room the entrance, which is 2.5 m X 2.7 m, is closed by two door sections travelling on rails and made up from three layers of steel (10 + 5 + 5 mm) with concrete in between and covered with wedges. When the doors are closed the wedge lining of the room is thus continuous. For practical reasons there is also a hinged door covered with mineral wool, which can replace one of the door sections for less critical measurements. The opening in the outer wall is closed by an air tight steel sliding door and the corridor outside the door has no windows and is heavily damped. The 2.2 m X 1.6 m entrance to the small anechoic chamber is closed by two hinged doors made from steel plate and concrete. One of the doors is lined with wedges. In front of the other, which is lined with a 5 cm layer of Sillan, there is a light, wedge lined, sliding door which can be slided into the corner of the anechoic room where there are no wedges, but only a 5 cm Sillan lining. The wedge lining of the room is thus continuous when the doors are closed, and at the same time the construction is simple and saves space. Floors The floors of the anechoic rooms should not influence their acoustical pro- perties appreciably. Wire netting was decided upon, which has been used for many anechoic rooms with satisfactory results. The mesh size was chosen to be 50 mm using 2 mm diameter steel string for the large room and 3.5 mm 13 for the small room. The smaller wire thickness in the large room makes it necessary for people to use large overshoes with wide rubber soles when walking around in the room, in order to protect the strings and to distribute the weight over a larger area. In the smaller anechoic room, which is used partly by the students, the in- fluence of a thicker wire has been accepted, in order to avoid the use of extra shoes. The wires are fixed to adjustable tension grips situated on the outside of the inner concrete box. Each wire is pre-stressed by 200 kg, so that the deflection with a person (75 kg) situated at the center of the floor is about 1 cm. The maximum load for each netting is 1,000 kg, however it must be distributed in such a way that the load per meter of the periphery of the load does not exceed 100 kg. The wire nettings are electrically isolated from the building. This is done in order to reduce the risk of electric shocks from measuring instruments and to make it possible to fix the potential of the floor at any desired level independent of that of the surroundings. The nettings are placed at the same level as the floors of the outside rooms and the central planes of the anechoic rooms, where most of the measurements will be taken, are situated about 1.5 m above the nettings. Parts of the rooms which are below the nettings can be reached through removeable gratings in front of the doors. A fine mesh perlon netting is placed below the steel nettings, in order to catch small items accidentally dropped. Lighting As the linings of the rooms are highly insulating with respect to heat, the lighting must give a minimum of heat radiation. At the same time it is neces- sary to have quite a high light intensity in order to be able to work with small microphones, hearing aids etc. As the wall lining is also very light absorbing and as no reflectors can be tolerated around the lamps, which must also be placed as far away as possible from the measuring area, the lighting installation must have the highest possible conversion efficiency. The main illumination comes from 200 W mercury vapour lamps, 12 in the large room and 2 in the small room. As these lamps ignite slowly and are probably not completely noiseless, there are also a set of ordinary 200 W incandescent lamps which may be used separately, 8 in the large room and 2 in the small room. It is a relatively simple matter to remove the lamps if this should be necessary for acoustical reasons. Ventilation The ventilation system is dimensioned with a view to obtaining a suitable rate of change of air in the rooms, and at the same time to avoid any appreci- able increase of background noise level with the system operating. In both 14 rooms the ventilation system consists of a row of air inlets along the bottom of the walls of the room and outlets along the top of the opposite wall. The fans are placed in the basement outside the anechoic rooms, and on its way into the rooms the air passes through two long, heavily damped concrete ducts, which are connected by flexible tubes. The exhaust air is similarly taken through two concrete ducts connected by flexible tubes to the fans with outlets into open air. For the large anechoic room the ducts are 16 m and 10 m long with internal cross-section 70 cm X 80 cm. The thickness of the concrete wall is 10 cm. The acoustical attenuation is obtained by lining one of the inner walls with 50 cm mineral wool, which is divided into sections by thin metal plates to avoid sound transmission through the material along the ducts. The ducts for the small anechoic room are 7 m and 5 m long, and their concrete walls are 10 cm thick. The internal dimensions are 45 cm X 60 cm and the sound absorbing lining consists of a 20 cm layer of mineral wool along one wall, divided into sections as that for the large room. The inlet and outlet nozzles in the anechoic rooms are made of sheet metal and to avoid sound radiation from possible vibrations in the nozzles, these are covered externally with a 5 cm layer of mineral wool. The inlets and outlets in the large room have the dimensions 6 cm X 100 cm and in the small room 6 cm X 40 cm. Also the ducts leading to the open air inlets and outlets are heavily damped so that noise from the fans will not disturb measurements in rooms situated on the same side of the buildings as these openings. The air can be changed 4 to 5 times an hour in the large room and 12 to 17 times per hour in the small room. The air speed at the inlets and outlets into the rooms is about 2.9 m/sec. Starting and stopping of the ventilation machinery and control of the air temperature is conducted by control knobs outside the doors for each room. It is not, however, possible to obtain a lower air temperature than that of the outside air as there is no refrigeration system. A refrigeration system was discussed during the planning stages, but due to the particular insulating properties of the wedges, which made calcula- tions difficult it was decided to leave space for a refrigeration unit but to delay the procurement and installation until the room had been used for some time. Experience from 1966-67 does not indicate any need for refrigeration of the air. Apparatus and Store Rooms During the planning of the acoustical building with the anechoic rooms it was considered imperative that rational working conditions were obtained in the finished construction. By having several apparatus rooms available for setting up instrumentation in connection with measurements in the anechoic rooms, it is possible to have instrumentation systems for the most common measure- ments permanently set up in one or two of the apparatus rooms, while the 15 others are used for special set ups, which can be built up or taken down without disturbing routine measurements. Three larger apparatus rooms of about 10 m2 floor area and one smaller room of about 7 m2 are used in connection with the large anechoic room while the smaller anechoic room has two 10 m2 rooms. The apparatus rooms are situated along the corridors outside the anechoic rooms as shown in Fig. 2. At the end of each corridor there is a large store room, which may also be used for instrumentation in large measurement systems. See Fig. 25. Between the anechoic rooms and the apparatus rooms there are cable con- nections terminating on connection boards 0.5 m above the wire netting in the anechoic rooms. All the connections boards are of the same size as the cross-section of the wedges, and they are covered by removeable wedge tips made from foam plastic, when not in use. In addition there is a connection board in the ceiling of each of the anechoic rooms. All the cables are connected to switch-boards in two of the apparatus rooms, from where connections can be made to connection boards in the other apparatus rooms and in the other laboratory buildings. There are three types of cable connections: 1. Microphone cables 2. Loudspeaker cables 3. Cables for remote control etc. The most critical cables are the microphone connections, where the trans- mission of small signals must be uninfluenced by electrical noise. The con- nection is made with a special cable containing two pairs of cores with one screen each and surrounded by a common screen. All the screens have a double braid whereby a very efficient screening is obtained even for electrical fields of high frequency. This is considered necessary because of the com- paratively strong fields in the university area resulting from radio and television transmitters. There is also a possibility of disturbance from a nearby radar station and from measurements in the electromagnetically anechoic room, which is adjacent to the acoustical laboratory. Because of this all plug connections are specially designed for screening against high frequencies, with screening properties about 50 dB better than normal screened plugs. These measures were found less expensive than total screening of the buildings. It is assumed, however, that the metal shelves for the wedges, which are electrically con- nected, and the reinforced concrete walls and ceiling will provide an effective screening against stray fields. It is intended to use one pair of cores in the microphone cable for trans- mission of the microphone signal, while the other pair may be used for calibration of the microphone amplifier. The loudspeaker cables also contain two pairs of cores, one pair is used for transmission of the signal from the generator in the apparatus room while the other pair is used for measuring the voltage at the loudspeaker. The third 16 cable system uses a ten-core cable, which may be used for remote control purposes (for example of a turntable in the anechoic room), for temperature measurements, power supply etc. Each connection board in the anechoic rooms has possibilities for connecting three microphone cables, three loudspeaker cables and two remote control cables. In addition there is an earth connection which is connected via a 16 mm2 cable to an earthing netting buried in front of the building. The re- sistance to earth is about 1.5 Q. All the measuring instruments can be supplied by 220 V AC, which is balanced with respect to earth and supplied by a balancing transformer, which supplies this building only. This type of power supply, which reduces the hum problems in some of the measuring set-ups was chosen on the basis of experience from earlier work in the laboratories. Furthermore, a maximum distance between power lines and measuring cables has been aimed at throughout, and all the cables are lead through steel tubes whenever possible. At the inlets to the anechoic rooms the steel tubes are substituted by flexible metal hoses in order to avoid sound transmission. Measurements in the Completed Rooms Measurement of deviation from 1/R-law The supplier of the wedge lining for the two anechoic rooms gave certain guarantees with regard to the acoustical quality of the completed rooms. These were in the form of a maximum deviation of the sound pressure from 1/R-law (which indicates inverse proportionality between sound pressure and distance Fig. 13. The maximum values of the deviation of sound pressure from the 1/R-law as a function of the distance from a point source. 17 from a point source in a free field) for given distances from a point source radiating sinusoidal waves within a specified frequency range. These maximum values and the specified frequency range for each room are given in Fig. 13. The guarantee included any measuring point situated at least 1.5 m and 1.2 m from the wedge tips in the large and the small room respectively. Checking that the limits were not exceeded was left to the laboratory, which had some 2-3 months for the measurements. This time limit and a wish to have an effective and practical method of control led to the development of a measuring system, which would automatically and continuously record the deviations from the 1/R-!aw in a certain direction in the room. The principles involved in the measuring system are shown in Fig. 14. The micro- phone, M, is carried away from the sound source, S, on a small electrically driven carriage, and travelling with constant velocity on a rail suspended immediately below the wedges in the ceiling. The microphone is placed at the end of a rod of maximum length 5 m. This suspension of the microphone Fig. 14. Arrangement for the measurement of deviations from the 1/R-law. 18 gives the possibility of measurement in several horizontal planes using different rod lengths. For measurements along the diagonal of the room the carriage and the microphone are each connected to a pulley, so that it is possible to give the microphone a vertical movement and a horisontal movement at the same time. The amplified microphone voltage is applied to the potentiometer Pi (linear characteristic, 10 turns) which is mechanically connected to the carriage, so that the attenuation of the signal by the potentiometer is inversely proportional to the distance between the microphone and the source. Thus there will be a constant voltage on the slider of the potentiometer under the following two conditions: I. The 1/R-law is fulfilled II. The attenuation of the microphone signal at the starting position of the carriage is adjusted to correspond to the distance between the acoustical centres of the microphone and the source with a second potentiometer P2. If condition II is not complied with, there will be either an increase or a decrease of the voltage on the slider of Pi near the sound source. This fact may be used for correct adjustment of P2 since deviations from the 1/R-law are especially small near the sound source. Thereby it is not necessary to determine the position of the acoustical centres. After correct adjustment of the potentiometer P2 the voltage changes on the slider of P} will be a direct measure of any deviations from the 1/R-law. After amplification and filtering the potentiometer voltage is recorded and a curve obtained as shown in Fig. 15 where deviations from the 1/R-law are read as deviations from the centre line of the recording paper. By carefully covering the rail with about 5 cm glass wool, and by consistently checking that the cables for carriage and microphone do not enter into the Fig. 15. An example of the recording of deviations from the 1/R-law as measured with the set-up shown in Fig. 14. 19 measuring field it is possible to obtain a measuring accuracy of about 0.2 dB. The accuracy of the electrical system alone, is about 0.1 dB. The sound sources must have a good omnidirectivity in the frequency range for which they are used. Small deviations will, however, have only secondary influence on the strength of the reflected sound field. If this influence is neglected so that only the effect of deviations in the strength of the direct sound field is considered, it is possible to calculate, quite simply, a set of curves showing the maximum allowable deviation from a spherical charac- teristic as a function of the maximum deviation from the 1/R-law in the room and with the maximum error of the deviation measurement as parameter. Such curves are shown in Fig. 16. A sound source with up to 1 dB deviation from the spherical characteristic will thus give an error in the measurements smaller than 0.1 dB if the maximum deviation from the 1/R-law is 0.8 dB. By about 1.9 dB deviation from the 1/R-law the error caused by the source would be smaller than 0.2 dB. Fig. 16. The maximum allowable deviation of the sound source from the spherical characteristic as a function of deviations from the 1/R-law with measuring error for the deviation as parameter. In order to cover the frequency range guarenteed by the wedge manufacturer three sound sources were used. The range 60-250 Hz was covered by an electrodynamic loudspeaker, Lorenz LPT 245 (24.5 cm diameter), in a closed, damped, 18 litres cabinet. The maxi- mum deviation from a spherical characteristics was 0.6 dB. 20 250-2,000 Hz was covered by two loudspeakers screwed together front to front, type Philips AD 2400 (10 cm diameter). The loudspeakers were connected in such a way that the unit was operating approximately as a sound source of zero order. The maximum deviation from a spherical characteristic was less than 0.6 dB. 2,000-10,000 Hz was covered by a dismounted microphone cartridge, type Sennheiser MD211, which was closed at the back and furnished with a Bruel & Kjagr nose cone at the front. Discs of glass wool were furthermore placed in front of and behind the wire mesh of the nose cone, as shown in Fig. 17. When the cartridge was supplied with a voltage of 2.5 V RMS a sound pressure level of 70 dB re 2 X 10"5 N/m2 was measured at a distance of 1 m at 10 kHz. The maximum deviation from a spherical characteristic was less than 0.7 dB with careful adjustment of the glass wool discs. Fig. 17. Sound source with spherical characteristic in the frequency ranqe 2,000 to 10,000 Hz. In the range 10,000-20,000 Hz, which is outside the guarantee, a hollow polarized barium titanate sphere of diameter 4 cm and about 4 mm wall thickness was used. With 80 V RMS across the silvered inner and outer surfaces the sound pressure measured was 80 dB re 2 X 10-5 N/m2 at a distance of 1 m at 16 kHz. The maximum deviation from a spherical charac- teristic was 0.5 dB at 13 kHz and 4 dB at 16 kHz. This rather large variation in deviation was probably due to the rather irregular thickness of the walls of this primitive test sphere. The same considerations as above are valid for the microphones with regard 21 to directional characteristics. The range 60-10,000 Hz was covered by a V2" Bruel & Kja9r condenser microphone with nose cone. The maximum deviation from a spherical characteristic was less than 1 dB. The range 10,000-20,000 Hz was covered by a VA" Bruel & Kjeer microphone, with a maximum deviation from a spherical characteristic of less than 2 dB. The measurements were taken in steps of Vs octave within the guaranteed frequency range except for a single measuring direction in the large room where measurements were taken in the range 25-20,000 Hz. About 550 measure- ments were taken for each room. For frequencies above 5,000 Hz the measurements were corrected for ab- sorption in the air. Fig. 18. Maximum deviation from the 1/R-law as a function of frequency for the small anechoic room. Fig. 19. Maximum deviation from the 1/R-law as a function of frequency in the best measuring direction for the small anechoic room. 22 Fig. 18 shows the maximum deviations from the 1/R-law for the small room as a function of frequency and with the distance from the sound source as parameter. The curves shown in Fig. 19 were found for the best measuring direction in the room. It is seen that the deviations are smaller than 0.4 dB at a distance of 1 m in the range 100-10,000 Hz. Fig. 20 and 21 show the corresponding curves for the large room, the latter for the frequency range 25-20,000 Hz. It is seen that in the best measuring direction of the room the deviations are smaller than 0.2 dB from 30 to 16,000 Hz at a distance of 1 m from the sound source. A comparison with the guaranteed limits shows that the limits have been exceeded to a small extent at certain points. This was nearly always in the Fig. 20. Maximum deviation from the 1/R-law as a function of frequency for the large room. Fig. 21. Maximum deviation from the 1/R-law as a function of frequency in the best measuring direction for the large anechoic room. 23 direction towards the doors where the gratings mentioned before had to be covered with 5 cm polyurethane foam in order to bring the deviations down to the level shown by Fig. 18 and Fig. 20. It should be noted that 9 9 % of the total number of measurements were inside the guaranteed limits. Measurements of background noise Background noise measurements have been carried out in the anechoic rooms with a noise level in the surroundings of the building which must be considered normal (60-70 dB re 2 X 10"5 N/m2). An electrodynamic loud- Fig. 22. Background noise level in the large anechoic room. Fig. 23. Background noise level in the small anechoic room. 24 speaker with a diaphragm diameter of 24.5 cm was used as "microphone". Its output voltage was analysed in the frequency range 20-200 Hz with a band- with of 3 Hz, the upper frequency limit being set by the signal to noise ratio of the measuring arrangement. This analysis was used together with the diffuse field sensitivity of the "microphone" for a calculation of the sound pressure level in Va octave bands. The results obtained from measurements in the two rooms are shown in Fig. 22 and 23 with and without the ventilation system working. For frequencies above 200 Hz the sound pressure level is bound to be lower than the value obtained at 200 Hz. The intended values for the background noise are seen to have been reached with the ventilation system not operating. With the ventilation on, the back- ground noise level is still below the normal threshold of hearing. Measurements of sound insulation An investigation into the sound insulating properties of the double wall con- struction meets with certain practical difficulties, especially at the higher frequencies, since as mentioned before it was intended to achieve a trans- -- -- rriTTiT rmfywi Fig. 24. "Sound source" used for measurement of the sound insulation properties of the double wall construction. 25 mission loss of some 80 dB at 200 Hz for random incident sound. It is therefore necessary to employ extremely powerful sound sources at this and higher frequencies in order to measure the transmission loss. In the present case a jet helicopter type S 61 was used, which was positioned about 15 m above the roof of the large anechoic room, as shown in Fig. 24. The S 61 radiates sufficient sound power in all directions up to about 2 kHz. The sound pressure level at a distance of 15 m is about 110 dB re 2 X 10"5 N/m2, and a Vz octave analysis gives values within 6 dB up to about 2 kHz. During the test the sound pressure level was checked at 5 points outside the outer structure for the large anechoic room, at 5 points between the outer and the inner wall, at 2 positions inside the large anechoic room and finally at several places in the building, such as for example inside the small anechoic room. All the sound pressure levels were recorded both on a level recorder and on magnetic tape, making a later frequency analysis possible. The measuring system is shown in Fig. 25. Fig. 25. Arrangement for measuring the sound insulation of the double wall construction. The results obtained are shown in Fig. 26, where the attenuation of the outer wall and through the two walls is shown as a function of frequency. It is demonstrated that the intended transmission loss has been achieved, as at 200 Hz for example the attenuation is 84 dB for the double wall construction. In order to investigate the efficiency of the vibration isolators supporting the rooms, a series of vibration measurements were conducted for the large anechoic room. The measurements were taken using a pile driver working at a site nearby as signal source. The vertical component of the vibrations 26 Fig. 26. The transmission loss through the outer walls and the double wall construction for random incidence no'se on the large anechoic room. caused by the pile driver was measured on either side of a vibration isolator at each corner of the large anechoic room using high sensitivity ceramic accelerometers (about 3,000 mV/g). The accelerometer voltages were recorded on a general purpose tape recorder via an AM system. The modulator was working with a carrier signal of 500 Hz, which was suppressed before recording on the tape. The following frequency Fig. 27. Attenuation of vertical vibrations through the vibration isolators supporting the large anechoic room. 27 analysis was carried out with a bandwidth of 3 Hz and with a tape speed 8 times higher than the speed of recording. The lower limiting frequency of the measuring system was around 0.5 Hz. An average of the 4 measurements is shown in Fig. 27. The curve indicates that the resonance frequency of the vibrating system consisting of concrete box and vibration isolators is about 8 Hz, or fairly close to the intended value of 7 Hz. Conclusion The measurements carried out show that both anechoic rooms can be used even for very exacting measurements, the data for the large room probably representing something near the limit of what one can achieve today. From the background noise measurements it is seen that it has been possible to attenuate the noise from the ventilation system to such a degree that the ventilators may be used even during psycho-acoustical tests. The insulation against noise in the vicinity of the buildings is extremely good, in as much as even with sound pressure levels of the order of 110 dB re 2 X 10 5 N/m2 the sound pressure inside the room just reaches the threshold of hearing. It is therefore possible to carry out a large number of normal acoustical measurements in these rooms even under extreme outside condi- tions. Selected Bibliography 1. Meyer, E., Buchmann, G., und Schock, A.: Eine neue Schluckanordnung hoher Wirksamkeit und der Bau eines schallgedampften Raumes. Ak. Zeifschr., 5, (1940), 352. 2. Beranek, Leo L. and Sleeper, Harvey P.: The Design and Construction of Anechoic Sound Chambers. Journ. Ac. Soc. Am., 18, (1946), 140. 3. Kurtze, G.: Untersuchungen zur Verbesserung der Auskleidung schall- gedampfter Raume. Acustica, 2, (1952), AB 104. 4. Meyer, Erw., Kurtze, G., Severin, H. und Tamm, K: Ein neuer grosser reflexionsfreier Raum fur Schallwellen und kurze elektromagnetischen Wellen. Acustica, 3, (1953), AB 409. 5. Epprecht, G. W., Kurtze, G. und Lauber, A.: Bau eines reflexionsfreien Raumes fur Schallwellen und elektrische Dezimeterwellen. Acustica, 4, (1954), AB 567. 6. Kraak, W., Jahn, G. und Fasold, W.: Ein neuer grosser reflexionsfreier Raum fur Schallwellen. Hochfrequenztechnik u. Elektroakustik, 69, (1960), Heft 1. 7. Rivin, A. N.: An Anechoic Chamber for Acoustical Measurements. Soviet Physics-Acoustics, 7, (1962), No. 3. 8. Schroder, F.-K.: Der Schallschluckgrad als Funktion des Schalleinfalls- winkels. Acustica, 3, (1953), 54. 9. Krogh, H. J. C: Eksamensprojekt 1964 (Master Thesis), DTH, Copenhagen, 1965 (in Danish). 28 10. Wiuff, A. E.: Eksamensprojekt 1960 (Master Thesis), DTH, Copenhagen, 1961 (in Danish). 11. Diestel, H. G.: Zur Schallausbreitung in reflexionsarmen Raumen, Acustica, 12, (1962), 113. 12. Thielemann, G.: Ermittlung der akustischen Eigenschaften des grossen schalltoten Raumes. Diplomarbeit, Institut fur Elektro- u. Bauakustik d.T.H. Dresden, 1957. 13. Wichert, O.: Erganzende Messungen zur Ermittlung der akustischen Eigen- schaften des grossen reflexionsfreien Raumes. Belegarbeit, Institut fur Elektro- u. Bauakustik d.T.H. Dresden, 1958. 14. Nissen, Preben: Bending Moments in Members of Tension and Stressed Wires Subject to Concentrated Transversal Loads. Bygningstekniske Med- delelser, 38 (1967), nr. 1 (in Danish with an English Summary). 15. Ingerslev, F.3 Pedersen, O. J., Moller, P. K. and Kristensen, J.: New Rooms for Acoustic Measurements at the Danish Technical University. Acustica, 19, (1967/68), Heft 4. 29 Anechoic Chambers Anechoic room according to L. Cremer ♦ Verso Filler Page ♦ Brüel Acoustics Technical Review 96-02 Front page Anechoic Chambers By Dr. techn. Per V. Brüel Noise standards for many products ranking from household equipment, tools and ventilators to cars and buses require anechoic rooms for accurate testing. Also in development work anechoic chambers are a must for serious work with abatement of noise. The most easy way to make frequency and directional characteristics of loudspeaker systems is in an anechoic chamber. As noise standards will have an increasing importance for manufacturing and selling of industrial products containing moving parts, a large number of anechoic chambers must be produced. In the future we will see an increasing number of noise standards which require three types of rooms: 1) Free field where no reflections from any items and inner surface of the measuring place can be tolerated, and 2) Simulated open field where the bottom plane in the measuring room is sound reflecting. That means the floor in the room must be made of concrete, very hard wood or tiles. In Fig.1 different rooms are shown. 3) Simulated open field with a standard table for testing household equipment, computers, tools and other items used on a table. Free Field Room This room is used for testing microphones, telephones, hearing aids, sound level meters, and other small products. In most cases a relative small room will be adequate as the objects are small. Free field rooms will in some institutions be used for psycho- acoustic research, but here a bigger room is necessary and the sound insulation is also essential which makes this type of testroom very expensive. It is often seen that an anechoic chamber intended for measuring on industrial products is built on floating concrete and with heavy sound insulation. This is an expensive solution of little use, as practical all products can be tested at high levels as most elements in the measuring chain are linear. An exception is testing of very low noise level for example a disc drive to a computer or a watch. But such low noise level objects should be tested and controlled in a little hard room with very high sound insulation for both room and door, see Fig.2. A small box like that is much cheaper than an anechoic test room and much faster to work with. The test box can be used on a production line checking the noise of the product. Floor in a Free Field Room Most anechoic rooms made hitherto have a wirenet floor, see Fig.3, which is expensive. A very heavy metal frame is necessary to resist the high tension from the many wires, 150 kg/3cm which is 5 ton/m. The net is from an acoustical point of view excellent, but not practical in daily use, because you can- not walk on the net without shaking the object and the measuring instrument. If you happen to drop a microphone cartridge, pencil or watch, it is complicated to get it back. Therefore it is a good idea in a free field room to have some hooks in the ceiling to support the objects and a rotating beam for the microphone. Simulated Open Field As more and more regulations and standards require a reflecting floor it should be worthwhile to consider to build only a semi-anechoic room. For the time being we estimate that only 25% of all measurements require a free field room and future trend is going to use more rooms with reflecting floor. Because it is more in line with daily use of the object, it is cheaper to build and require less room in the building, see Fig.1. Large products like cars, buses, motor cycles, tool machinery, washing machines, refrigerators, and all other equipment intended for staying on a floor or on the ground should be measured in a semi-anechoic room. Smaller household equipment e.g. coffee machines, mixers, sewing ma- chines etc., and all other products intended to stand on a table should be placed on a table in a semi-anechoic chamber, see Fig.4. What do we measure in an Anechoic Chamber For many years we measured the sound pressure around the object and from this we could deduct the radiated energy. We can integrate over the total outer surface and get a single number for the total radiated energy in every 1/3 octave. The small reflection from the walls of the anechoic room or from some other object may give a small inaccuracy in the measured value. The reflections arrive to the microphone in different phases with the consequence that the reflected sound pressure sometimes add to the pressure and at other times the reflected sound diminish the pressure. So at different points at the surface the variation of the pressure can be rather big, but with integration over the whole surface the total error is small. In recent years we have got instruments which can measure intensity direct. As shown in Fig.5 the instrument multiply the product of sound pressure and particle velocity using two microphones. The pressure p is taken from the sum of the two microphones and the velocity of the particles v is calculated from the difference. In this way any influence from the reflections or from outside noise are eliminated. From this one might think that it is not necessary to have an anechoic room at all when using an intensity analyser. But that is wrong as the class 1 intensity analyser according to IEC standard is required only to handle ± 12dB pressure variation and naturally also 24dB velocity variation. This is a very small dynamic range which often is not sufficient in a normal acoustic hard room. Therefore anechoic test rooms are still necessary. But we do not have to build them as perfect and expensive as it is done when we only measure sound pressure. The conclusion is that we shall use different anechoic chambers. Besides free field and semi-anechoic we shall have small rooms and large rooms. We need test rooms with good quality for lower frequencies and others for higher sound frequencies. We have to decide if we mainly will use pressure or intensity measurements. Lastly we have to look at the economy. A) Several layers of tissue (cloth, woven material, fabric) with increasing flow resistance from inner to outer, see Fig.6. The inner layer could be a mosquito net and the following ones should be some heavier woven material. The other layer can be made of non-woven material with increasing thickness and flow resistance. Just inside the fixed solid wall there should be an air space of minimum 10 cm for low frequency absorption. It is very difficult to calculate the optimum configuration as there is so many parameters to choose among such as different spacing between layers and different dynamic flow resistance. But by measuring in a SWT hundreds of combinations and different materials can fast be tested and good results can be obtained. A positive thing is that it is possible to avoid mineral fibres which many people do not like. The negative side is that it is nearly impossible to find useable fire resistant material. The fire risk can, however, be reduced by chemical treatment of the woven material. The chemicals normally smells and by fire it may produce toxic gasses. A higher price is also a hinderness for these acoustical excellent rooms. b) The wedge system is the most common principle used for anechoic chambers. It is from an acoustical point of view an extremely good solution and easy to make. Fig.7 shows the principle. The wedges are cut out of a thick mat of mineral wool which makes the whole production economic with no waste. At higher frequencies the wedges have the drawback, that when the sound comes nearly parallel to the wall, it is possible to get a reflex from the large flat surface of the wedge. This may be avoided by having the density of the glass wool less heavy in the tip of the wedge than in the bottom. A distributed density is shown in Fig.8. This changing density would also give a better impedance matching. No one have until now made wedges with distributed density. The orientation of the fibres is also very important. Another disadvantage with the wedge solution is that the wedges often are placed direct or close to the solid wall, where the particle velocity is minimal. The consequence is a very reduced absorption. The result is that most of the mineral wool is wasted as the air cannot go through the solid wall. The mineral wool should be placed where the velocity of the particles is considerable, and it should have the correct impedance around 400 Rayl. L. Cremer In 1952 professor L. Cremer at Berlin Technical university suggested to build a room as shown in Fig.9. The absorbing wall consists of a solid glass fibre layer close to the solid room wall, and hereafter cubes of glass wool of different density. The cubes should not be mounted too regularly, but rather at random. Professor Cremer called this construction an "acoustic jungle". The construction works beautifully up to some MHz. So the Cremer principle should be uses when accurate high frequency measurement is required. The author has contributed to several of these rooms etc. Test rooms at B. & K., Nærum, Lund and Gothenburg Universities in Sweden, and in Rio de Janeiro, Brazil. All rooms work perfectly. A thumb rule says that a serious absorption only can be obtained if the absorbing construction is at least l /8 or better l/4 thick. So by 50 Hz where the wave length is l = 6,8m the thickness of the wall has to be 80 cm to give some absorption and 1,5 m thick to be good. If this large volume has to be filled with mineral wool the price for the last octave at lower frequency will be extremely high. But as mentioned earlier the mineral wool close to a solid wall is of little use as the velocity of the particles with low frequencies is only fractional. Consequently it can just as well be air. Economise with Mineral Fibres As can be seen from Fig.10 the velocity of the particles is low close to the partly reflecting wall, but maximum around l /4 from the wall. Consequently the absorbing material should be placed in some distance from the wall. 100% absorption can be obtained with a very thin layer of flow resistant material (R), placed l /4 from the wall. But for frequency one octave higher the absorption will be very low, as l /2 will have moved to R. Here the velocity is low and consequently also the absorption. If we want several octaves to be absorbed we use a flow resistor with a certain thickness. Fig.11 shows a simple absorber built of a perforated plate, a flow resistor, and an airfilled volume between plate and to the right the electrical analogy. The inductance L corresponds to the moving mass of air in and around the hole in the plate. The capacitor C corresponds to the air volume behind each hole. The resistor R is the flow resistance at the hole. From the electrical analogy can be seen that if we want maximum electrical energy wasted in R the generator impedance should also be R and the impedance of both C and L should be as small as possible. From this can be deduced: low frequency will be hampered by C, so therefore low frequency absorption can only be obtained by having a large air volume behind the perforated plate. The high frequency will be stopped by L. Therefore the perforated plate have to be thin and the perforation should be many small holes placed close together. The flow resistor should have the same impedance as the acoustic impedance of air. In his case you can get an absorption as shown in Fig.11. A large frequency range where the absorption is nearly 100%. Both at the low and high end we will see a drop in absorption due to high impedance in L to C. We can see that a lot of mineral fibres with low frequency absorption can be saved just by having the right flow resistance in a relative thin layer (5-10 cm) and a large volume behind the flow resistor. In this way a considerable absorption can be obtained down to 30-50 Hz with a small amount of fibres, but the air volume must be large 1-1,5 m. An acceptable Solution By combining all our knowledge and experience about anechoic chambers we can make the following suggestions and describe a modular system, which not only is modular in linear dimensions (40 or 60 cm), but also in a way so that clients can start building only a part of the construction and later on get further units for a more complete system. We can make test rooms specially for cars, buses or large machinery. In these cases the lower frequency is very important, but the higher frequencies (over 1kHz) have less importance. The construction is very simple and can be built for a reasonable price. To protect the flow resistor an open metal net can be installed. The net allows all frequencies up to 4 KHz to pass. Besides the lack of absorption of higher frequencies the large flat surfaces are also a drawback. It can be considerably better by installing some curved metal net containing a thin layer of 2-3 cm of mineral fibres with a very low density, see Fig.13. This sound distributors should not cover the whole surface, but leave around 1/2 of the total surface. Rooms of this simple construction are normally for measuring big objects and will be semi-anechoic with a hard floor. By using an intensity analyser instead of a SLM a very high precision can be obtained. Smaller rooms for testing objects are household equipment, motor cycles, gears, and other mechanical products should also be semi-anechoic rooms made as indicated in Fig.12 and Fig.13. but do not make them too small. If the object generate frequency of any importance over 1,5 KHz Cremer units should be added, see Fig.14 and then there is no limits for higher frequencies. Movable Floor Units and Doors Another way to avoid the large flat surfaces in walls and ceilings is to mount the plates which carry the flow resistor in a zigzag way, see Fig.15. Careful investigations of constructions of radio studios have shown that an angle of ±5° is absolutely sufficient. Objects like sound level meters, loud speakers, hearing aids and rooms for psycho-acoustic work should all be free-field rooms with absorbing floors. As a conventional floor net is both expensive and impractical, it is suggested that the floor is built up as separate units, Fig.15. The units should be light, so it is easy to stack them and by this method make a corridor for installing the equipment. This is a very flexible system, fast and accurate to work with. The absorbing floor units are very light (3,5 kg/for 40x40 cm) see Fig.16 so they can be lifted with hands. The units are placed on the floor during measurements. During preparation for the measurements some units are lifted up and placed on top of the neighbour unit. In this way the testroom is very flexible as it can be used both as a free field and a semi-anechoic chamber. The only disadvantage is that the floor does not have the low frequency absorber. The door to the room is like the floor units a very simple and light construction made of 10mm Al- tubes. The door can also easily be lifted out and put in position as it only weights a few kilograms. To make hinges and handles for a door 0.5-1.2 m thick is a difficult and costly procedure. The result is always a clumsy solution which takes time to operate. The disadvantage with a light door is that the insulation from outside noise is low. But as mentioned earlier this is normally only required for psycho-acoustic measurements. Industrial noise is always measured at a higher level, where a limited sound insulation can be tolerated. It may be practical to roll the door on gliding tracks both out and then sideways. Testing Anechoic Chambers The classic method to test an anechoic chamber is to use a point sound source and apply the distance law: 6dB drop for doubling the distance. By using a log scale for the distance the pressure drop should be a straight line, see Fi.g17. It is very difficult to make a monopole sound source. A loud- speaker in a baffle or even in a cabinet will always radiate some sound from the reverse side. A solution is to install a mini loudspeaker in a solid sphere (Fig.17) filled with absorbing material. Using pure tones for testing is unrealistic as practical all the noise sources which should be tested in the room are broad band noise. Therefore if you from an inverse distance curve will estimate the accuracy the test signal should be a narrow band. Fig. 17:Monopole source for checking the inverse square law in anechoic chambers. Right some results. Standing Wave Tube Another solution for testing the quality of an anechoic wall is to evaluate the absorbers on the wall in a Standing Wave Tube (SWT). We have made and used two tubes with different cross-section. One is a quadrangle 40x40 cm and the other is circular with D=10 cm and 3 cm, Fig.18. With the shown set up it is possible to measure from 50 Hz to 6500 Hz. In another paper we will describe the SWT used here in more details. Fig. 18: Standing Wave Tube (STW) for testing absorbers intended for anechoic rooms in frequency range 50 Hz to 6000 Hz. Divided over 3 tubes 40 x 40 cm, D=10 cm and D=3,2 cm. The square section tube has sand filled wall to eliminate vibrations in the walls. Fig.19: Example of "absorption drop" at half wavelength l/2 when the resistor is very thin. With a thicker resistor the absorption curve is much flatter The SWT gives a high accuracy of the absorption of the particular sample. The problem is to make a sample which is representative. The psychological disadvantage is: 1) it is the lowest absorption coefficient which exists. 2) With a thin flow resistor just l/2 from the background a drop in absorption occurs. This effect is shown in Fig.19 for a very thin plate and an interspace of 7 cm. With a somewhat thicker resistor this phenomenon is gone. Fig.20 illustrates how the absorption increases with the angle of incident. The solid curves are valid for most absorbents consisting of textiles or mineral fibres. The dot and dash curves indicate when the soft material are covered with some perforated plates which introduce an imaginary impedance. Fig. 20: The absorption of porous material with different angle of incidence Fig. 21: Absorption for the same flow resistor mounted with different airspace. The origin of these theoretical curve dates back to P. M. Morse, R. Bolt and L. Beranek in the years 1938-39. The points are experimental verification of the theory. It is shown that the absorption always is found grater than the theory predicts. We have not been able to verify the theory beyond 75° incidence. The curve marked Cremer is the theoretical prediction for absorbers made from glass wool cubes. It appears that if the absorption measured in a SWT is 98% the effective absorption already with 20° incidence is close to total absorption. We will later write a paper about this interesting problem. We have together with Florence University under the MONICA project made some measurements converting SWT results to effective absorption in rooms. Practical Configurations Fig.21 illustrates how the absorption changes with the distance between the solid wall and the flow resistor. It is seen that low frequency absorption only can be obtained by using a large airspace or thickness of the absorber. This construction is often sufficient for large projects like buses, cars, large tool machinery, etc. as it is economical for large rooms. A room of this type will work excellent together with some layers of Cremer cubes. On Fig.22 is seen how the low frequency room shown in Fig.21 can be transferred into a room with a perfect high frequency absorption. The drop at 500 Hz and 1500 Hz is due to perpendicular incidence. It is averaged out for random incidence for frequencies over 300 Hz. As shown here the Cremer configuration is excellent for middle and higher frequencies, and the lower frequencies is be taken care of with a resistor type absorber. Fig.22: A good construction for a wall for an anechoic chamber with Cremer absorbers. ® Tutti i diritti riservati 1995-2003, SI AUTORIZZA LA COPIA E LA DIFFUSIONE CON CORTESE RICHIESTA DI INDICARE LA FONTE ( www.bruel-ac.com ) e l'E-Mail: [email protected] Prof. Mario Mattia. Brüel Acoustics Anecho Rooms Anecho Test Facilities: 62BA and 62BB Brüel Acoustics have developed two modular Anecho Test Rooms for industrial research, acoustical measurements and calibration. The test chamber is a light construction named type 62BA (modul 40 cm) and the system for larger rooms type 62BB (modul 60 cm). The modular system makes use of a 1imited number of standard elements which can be connected together and form more than 100 rooms with different lenght, weight and height. On a given space all dimensions can be made by choosing the nearest number of moduls. The price for the rooms are proportional to the element units. Cremer Contra Wedges: Brüel Acoustics are using the Cremer principle which has proved very useful and economic. The absorbing walls are built up with cubes which near the inside of the room are small and made of special glass fibres with a very low density. The cubes increase in both size and density as you go into the wall. In this way you obtain an extremely good impedance matching with the heavy absorbing material in the internal part of the room. This principle is analogous to the exponential horn in front of a loudspeaker driver unit. Wedges are normally build of the same material both at the bottom and the top. In the loudspeaker analogy this correspond to a linear cone in front of the driver. Consequently the Cremer room is much better at high frequencies than a Wedge room of the same size. At mid frequencies the two types are equel, and at low frequencies the Wedge type is normally slightly better. From an economical point of view it can be said that a Wedge room contains more absorbing material than a Cremer room, but a Cremer room requires more sophisticated work which means higher labour cost. At B&K we used only Cremer type rooms because it is very important that the room is good at higher frequencies up to 15 kHz. The lower frequencies below 300 Hz are not so important as the wavelength is large so that other methods can be used. With frequencies over 3 kHz a Wedge room will always give some uncontrollable fase shifts because the Wedges have large plane surfaces whereas the Cremer rooms have an acoustical jungle. Construction: Both Cremer types are built of tubes where no plane surfaces exist. Consequently all reflections are minimized in all directions. Most of the supporting structure is imbedded in the absorbing material. Type 62BA have a modul length of 40 cm and also a wall thickness of 40 cm. Supporting tubes have a diameter of 10 mm. This room works very accurately from 350 Hz to 25 kHz depending of the size. Type 62BB has a modul length and a wall thickness of 60 cm and is intented for bigger rooms and working frequencies from 200 Hz to 20 kHz. Both rooms should be erected on a plane smooth surface e.g. reinforced concrete or very hard wood. A concrete slab are ideal and can conveniently be supported by steel springs for vibration isolation. Floors: Both types can work with and without an absorbing floor and therefore satisfy all standards in force today. The absorbing floor consists of easy removable units. It takes only a few minutes to remove the floor. It is therefore important that the base reflects the sound and is plane. Doors: The door can consist of one or two units in width and two, three, or four units in height. The door itself consists of units which like the floor fits in the normal wall units. The door is rolling out perpendicular to the wall. When it is out, it is rolled parallel sidewards. This movement is done on rails with roller bearings. How much space is needed?: The rooms can be erected from the "inside" and therefore the space required is only 5 to 10 cm space between testroom and fixed building. But remember the test rooms can only be delivered in units of 60 or 40 cm. The space from the wall where the door is installed has to be the wall tickness (60 or 40 cm) + 10 cm. This small space is minimum. The erection is faster and easier if more space is available. Ventilation: Air should not be blown into the test room. Heavy draught will break some of the glass fibres and produce dust. But the air around the room can be airconditioned without problems. Lisht and Electric Cables are easily installed in a Cremer room as there is a lot of space in the absorbing walls. It is very easy to change the cables. ♦ Verso Filler Page ♦ EDITOR ACOUSTICAL SOCIETY Dick Stern ([email protected]) OF AMERICA CONTRIBUTORS Michael A. Bahtiarian, Jack Freytag The Acoustical Society of America was founded in 1929 to Marshall Long, Elaine Moran, increase and diffuse the knowledge of acoustics and to pro- Paul Schomer, Dick Stern, mote its practical application. Any person or corporation Nancy S. Timmerman, Douglas F. Winker interested in acoustics is eligible for membership in the Society. Further information concerning membership, as EDITORIAL BOARD well as application forms, may be obtained by addressing Dick Stern, Chair; Elliott H. Berger; Elaine Moran, ASA Office Manager, Acoustical Society of Carol Y. Espy-Wilson; K. Anthony America, Suite 1NO1, 2 Huntington Quadrangle, Melville, Hoover; James F. Lynch; Allan D. NY 11747-4502, Telephone (516) 576-2360; Fax: 516-576- Pierce; Thomas D. Rossing; 2377; E-mail: [email protected]; Web: http://asa.aip.org Brigitte Schulte-Fortkamp ADDRESS Acoustical Society of America Suite 1NO1 2 Huntington Quadrangle Melville, NY 11747-4502 ASA EXECUTIVE DIRECTOR Charles E. Schmid ASA EDITOR-IN-CHIEF Allan D. Pierce EDITORIAL PRODUCTION FUTURE ARTICLES Maya Flikop—Director, Special Publications and Proceedings, AIP Acoustical Scattering—Part 2 ADVERTISING Feedback Design Robert Finnegan, Director, Journal Advertising & Exhibits Pitch Circularity Bow Bells Cavitation and Cetacean Marine Mammals SONAR ACOUSTICS TODAY (ISSN 1557-0215, coden ATCODK) October 2009, volume 5, issue 4 is published quarterly by the Acoustical Society of America, Suite 1NO1, 2 Huntington Quadrangle, Melville, NY 11747-4502. Periodicals Postage rates are paid at Huntington Station, NY, and additional mailing offices. POSTMASTER: Send address changes to Acoustics Today, Acoustical Society of America, Suite 1NO1, 2 Huntington Quadrangle, Melville, NY 11747-4502. Copyright ©2009, Acoustical Society of America. All rights reserved. Single copies of individual articles may be made for private use or research. Authorization is given to copy arti- cles beyond the use permitted by Sections 107 and 108 of the U.S. Copyright Law. To reproduce content from this publication, please obtain permission from Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA via their website: www.copyright.com/, or contact them at (978)-750-8400. Persons desiring to photocopy materials for classroom use should contact the CCC Academic Permissions Service. Authorization does not extend to systematic or multiple reproduction, to copying for promotional purposes, to electronic storage or distribution, or to republication in any form. In all such cases, specific written permission from the American Institute of Physics must be obtained. Permission is granted to quote from Acoustics Today with the customary acknowledgment of the source. To reprint a figure, table, or other excerpt requires the consent of one of the authors and notification to AIP. Address requests to AIP Office of Rights and Permissions, Suite 1NO1, 2 Huntington Quadrangle, Melville NY 11747-4502; Fax (516) 576-2450; Telephone (516) 576-2268; E-mail: [email protected]. Acoustics Today is also repro- duced in the Acoustical Society of America’s Digital Library shortly after a print copy becomes available. Members and non-member subscribers may download arti- cles or issues for their personal or professional use. The articles may not be altered from their original printing and pages that include advertising may not be modi- fied. Articles may not be reprinted or translated into another language and reprinted without prior approval from the Acoustical Society of America as indicated above. Non-member subscribers may purchase articles at a nominal cost under the same conditions that are specified for members and non-member subscribers. 2 Acoustics Today, October 2009 DESIGNING AND BUILDING A LOW-NOISE HEMI-ANECHOIC CHAMBER Douglas F. Winker ETS-Lindgren Cedar Park, Texas 78613 oise emission and noise control “The chamber exceeded the air conditioning (HVAC) system must are common topics in the field not affect the chamber’s noise floor and Nof acoustics. Noise emission specification with a noise must maintain controlled temperature from devices needs to be measured to levels of ± 2° C with a high heat load. An comply with noise standards and to floor that was > 0 dBA for addition to the HVAC requirement was develop quieter devices. To make these that the lighting system should not con- measurements, a quiet noise-controlled frequencies ≥100 Hz and only tribute additional heat to the chamber environment is required. A large con- and the system must be low noise. The vertible hemi-anechoic chamber was 5 dBA at 80 Hz.” fifth design requirement has not been constructed at ETS-Lindgren’s head- fully implemented—the chamber must quarters in Cedar Park, Texas for that be able to be converted into a fully ane- purpose. The project was a noise control project unto itself choic chamber. As part of this requirement, the chamber’s with the goal to measure the noise produced by a wide vari- wedge basket doors were designed to accommodate remov- ety of test specimens. able floor wedge carts, but these carts have not been con- ETS-Lindgren Acoustic Systems operated a laboratory in structed. south Austin, Texas beginning in 1985. The south Austin lab- The inner chamber rests on an isolated concrete slab that oratory had a suite of reverberation chambers and a hemi- was designed with the future in mind. The chamber’s loca- anechoic chamber with a 1-meter, precision-grade free field tion in Cedar Park, Texas is approximately 40 km from above 125 Hz. Over the years, tested product noise became downtown Austin and there is considerably less traffic in the lower and lower, especially for information technology (IT) area. While the area is somewhat remote now, continuing equipment. As product noise emissions decreased, their expansion will eventually engulf the area. With this expan- acoustic performance began to encroach on the chamber’s sion comes increased traffic and road noise. Highway expan- noise floor and the quietest of products were tested late at sion is underway and a light rail system will go online in the night when the ambient levels were at their lowest (approxi- spring of 2010. The isolation system was designed with this mately 17 dBA). Highway expansion was the main contribu- growth in mind. tor to the noise floor increase. The laboratory was located less The concrete work began with excavating the host site and than 2 miles away from the Interstate 35 and US Highway 290 pouring a concrete pit so that the chamber’s entry would be at interchange. When the chamber was constructed, this inter- floor level, which makes handling large test specimens much change did not exist and was not part of the design parame- easier. The existing concrete was cut out and the underlying ters. After the construction of this interchange, low-frequen- soil was excavated leaving a 7.92 x 7.92 x 0.56 m (Lx Wx H) cy traffic noise became measurable inside the chamber. depression. After the excavation, the pit was framed and con- In 2006, the decision was made to move the ETS- crete was poured. An expansion joint was left between the Lindgren Acoustic Systems’ laboratory and production facil- outer edge of the pit and the host slab to isolate the pit from ities to ETS-Lindgren’s headquarters in Cedar Park, Texas. host site noise. Figure 1 shows the pit while it was curing. The team decided to decommission the old chambers and build an entirely new lab with a double-wall hemi-anechoic chamber (i.e., a chamber within a chamber) that had a larger free field and a much lower noise floor. Low-noise testing had to be possible at all times of the day regardless of the noise sources outside the chamber. The chamber’s design was dictated by the project’s goal— a mandate to test a wide variety of devices. First, the ability to test low-noise products was required and a noise floor of NC- 10 at any microphone on the measurement surface was spec- ified. Additionally, low-noise testing had to be possible at any time regardless of the activities outside the chamber includ- ing future highway expansion. Second, a 2-meter radius pre- cision-grade (ISO 3745:2003) hemispherical free field was required for frequencies ≥80 Hz. Third, using a paral- lelepiped array, two full height equipment racks must be able to be tested side by side. Fourth, the heating, ventilating, and Fig. 1. Concrete pit. 16 Acoustics Today, October 2009 process can be seen in Fig. 3. Upon the completion of this operation, the bottom of the isolated slab was 7.62 cm above the floor of the pit. The inner and outer chambers were constructed simulta- neously after the concrete work was complete. The outer chamber walls are 10.16 cm thick and rest on the edge of the concrete pit. An 11-gauge steel perimeter channel was installed around the edge of the pit to receive the modular wall panels. A closed-cell foam gasket was installed beneath the perimeter channel to fill any voids that might exist between them and the concrete surface. Each of the wall pan- els was constructed in ETS-Lindgren’s factory and transport- ed to the job site. These panels were constructed to achieve high noise reduction. The outer surface of each panel is 16- gauge cold-rolled steel. Inside the panel, a 2.54 cm thick layer of gypsum board was laminated to the outer skin to increase the outer surface’s overall mass. The remaining cavity was filled with cotton fiber fill that is 7.62 cm thick. The inner surface is another layer of 16-gauge cold-rolled steel. The assembly is held together by 16-gauge cold-rolled steel chan- nels, which are welded in place. The wall panels fit into a labyrinth joiner system constructed using 11-gauge cold- rolled steel. Every joint and seam was filled with a bead of latex caulk. All of the steel in the chamber was powder coat- Fig. 2. Spring isolators during installation. ed for a lasting rust-free finish. The external dimensions of Once the pit was poured, the isolated concrete slab could the outer chamber are 8.69 by 8.69 by 7.32 m (Lx Wx H) be framed and poured. The floating slab resides entirely in above the finished floor. Figure 4 shows an outer chamber the pit and is supported by 54 spring isolators (Fig. 2). The wall installed in one of the perimeter channels. spring isolators were placed in the pit and strapped together The inner chamber walls are 30.5 cm thick and rest with reinforcing bar. Next, 45.46 metric tons of concrete were entirely in another 11-gauge steel perimeter channel that is poured resulting in a 30.5 cm thick concrete slab with a foot- attached to the isolated concrete slab. They are separated print of 6.1 by 6.1 m. The floating slab acts as the chamber’s from the outer chamber walls by 30.5 cm. This space is filled reflecting plane. The slab also features a 30.5 by 30.5 by 10.2 cm (Lx Wx H) recess at its center to hold floor- mounted sound sources or to flush- mount a motorized turntable capable of rotating test specimens up to 225 kg. A 1.25 cm thick steel plate covers the opening when the turntable or sound source is not in use. A 30 cm wide cable recess allows cables to run into the pit without disturbing the reflect- ing plane. A series of 1.25 cm thick steel plates cover this recess. Several holes were bored into the slab and act as mounting points for the various microphone arrays used during test- ing. Finally, an epoxy finish was installed to create a smooth and durable finish. The inner chamber was construct- ed on the isolated concrete slab and raised into position after construction. This was a tedious process that involved gradually raising each spring isolator in succession by a small amount and repeating the process until the isolated slab was flush with the host slab. This Fig. 3. Raising the inner chamber. Low-Noise Hemi-Anechoic Chamber 17 Fig. 6. Transmission loss for 10.16 cm and 30.5 cm thick panels (5.95 m2 specimen area). with Johns Mansville 30.5 cm thick R-30 fiberglass insula- tion. The installation of the chamber walls can be seen in Fig. 5. The inner chamber walls were constructed in a similar fashion to the outer chamber walls. They have the same 16- gauge outer skin with a 2.54 cm thick gypsum board lamina- tion. Since the panels are thicker, the cotton fiber fill is 27.94 cm thick. The main difference between the two panel designs is the inner skin, which is made of 22-gauge perforated steel with 23% open area consisting of 1.59 mm holes spaced on 3.18 mm staggered centers. By perforating the inner skin, the absorptive properties of the fill material can be used in con- junction with the wedge system. With this system, greater low-frequency performance can be achieved with shallower Fig. 4. Outer chamber wall install in a perimeter channel. wedges. The dimensions of the working area of the chamber (area inside the wedge tips) are 6.1 by 6.1 by 5.4 m (Lx Wx H). To verify the performance of the design, the outer wall panels were tested in accordance with ASTM E90 (Standard Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building Partitions and Elements) in ETS- Lindgren Acoustic System’s reverberation chamber suite. The transmission loss of these panels appears in Fig. 6. A structural mount was installed in the ceiling directly above the test specimen location in the center of the floor. A second turntable can be attached to this mounting point to rotate a microphone array around a test specimen when rotating the test spec- imen is not feasible. The inner chamber is lined with wedges that are made from 48.26 cm deep monolithic melamine foam with a base that measures 20.32 cm by 60.96 cm. They are installed in alternating banks of three giv- ing each bank a 60.96 by 60.96 cm foot- print. Each wedge is secured using ETS- Lindgren’s patented wedge clip system. The wedge clip system holds the base of each wedge 10.16 cm off of the inner panel skin, Fig. 5. Inner and outer chamber walls during construction. thereby creating a 10.16 cm airspace. The 18 Acoustics Today, October 2009 Fig. 7. ASTM E1050 normal incidence sound absorption test results for a 30.5 cm panel with and without wedges installed. resulting total depth of treatment is 0.86 m. Figure 7 shows the results of an ASTM E1050 (Standard Test Method for Impedance and Absorption of Acoustical Materials Using A Tube, Two Microphones and A Digital Frequency Analysis Fig. 8. Completed chamber with double doors open. System) normal incidence sound absorption test of the 30.5 cm panel by itself and with the wedges installed. The tests were performed in a 61 cm by 61 cm square cross section impedance tube. Large test specimens are loaded into the chamber though two sets of double doors. The first set of doors is mounted in the outer chamber wall and has a 2.74 by 3.05 m clear opening. The second set of doors is mounted in the inner chamber wall and has a 2.44 by 2.74 m clear opening. Each door has its own automatic door opener. The automatic door openers are syn- chronized to open or close each door leaf in a specific order with the push of a single button. Figure 8 shows the completed chamber with the double door open. Inside the chamber two wedge doors cover the opening. Each wedge door measures 3.05 by 1.83 m (Hx W) and moves out and sideways on a two- pivot hinge system and are shown in Fig. 9. One of the challenges with such a low noise floor was the need for silent environmental control. The chamber’s internal volume is 200 m3 and has an air change every 6 minutes. To accommodate the required air flow rate and eliminate any duct-borne noise, a series of HVAC silencers were installed. Only one inlet path and one outlet path were used on the chamber with two HVAC silencers installed in each path. The first silencer, in the inlet path, is housed in a 10.16 cm thick panelized housing. The outer layer of the housing is 16-gauge steel with a 2.54 cm thick layer of gyp- sum board laminated to its inner surface. Next, 7.62 cm of fiberglass-free sound absorbing material was added and then the inner skin, a second layer of 16-gauge steel, was applied. Inside this housing, a 3.05 m long, high-perform- ance splitter-type silencer with a 0.61 by 0.61 m cross sec- tion was installed. This housing was suspended from the Fig. 9. One wedge basket door in operation. host building’s ceiling with vibration isolation mounts to decouple any vibrations that might be transmitted from the from the chamber’s structure. building or the HVAC system even though it was attached A second HVAC silencer for the inlet path was installed to the HVAC system with flexible duct work. The outer between the outer and inner shells of the chamber. Due to its HVAC silencer housing passes through the chamber’s ceil- location between the chamber walls, a separate housing was ing through an isolated collar, which physically separates it not constructed around this silencer. The inner silencer is Low-Noise Hemi-Anechoic Chamber 19 similar in construction except that it has a cross section of 0.30 by 0.91 m. The end of this silencer terminates with a 90 degree turn through the top of the inner chamber. Acoustic turning vanes are in all 90 degree turns within the HVAC sys- tem to maintain laminar flow and provide sound attenuation. No vent grilles were used at the opening inside the chamber in order to eliminate flow-generated noise. The inlet and out- let path silencer solutions are identical except for the orienta- tion of the silencers due to the air flow direction. Two different sets of lights were installed in the cham- ber. Each style of light met the design requirements of not introducing heat into the chamber and not contributing to the noise floor. Normally, only one style of light is installed but both types were installed for demonstration purposes. The first style of light uses a standard can-type housing with Fig. 10. Control room. a light emitting diode (LED) bulb system. The second light- host slab and not to the isolated pit. The gap between the walls ing system is a fiber optic system. The system consists of was filled with a high-density neoprene foam gasket. The con- two 150-watt metal halide bulbs that are mounted outside trol room measures 8.69 by 3.66 by 3.05 m (Lx Wx H). This of the chamber. Light is delivered by eight fiber optic bun- room is fed by a separate HVAC and silencer system to elimi- dles that pass through the ceiling of both the inner and nate the possibility of noise flanking into the chamber from outer chambers and terminate inside the chamber. All of the the control room through the HVAC system. The control fiber optic bundle penetrations were treated to eliminate room is shown in Fig. 10. flanking paths. Once construction was completed, performance verifi- To accommodate a wide variety of test specimens, a col- cation testing began. The first test conducted was to measure lection of electrical outlets was installed on the chamber. The the chamber’s noise floor to confirm that it met the specifica- electrical outlets range from 110 Vac single phase to 480 Vac tion of a level not exceeding NC-10 at any of the microphone three phase service. Both 50 Hz and 60 Hz service is avail- positions on the measurement hemisphere. The chamber was able. The electrical panel is located on the outside of the outer tested at each location using a GRAS Type 40 HH low-noise chamber and the required electrical lines are passed through microphone. The chamber exceeded the specification with a an acoustically treated penetration in the chamber walls. The noise floor that was > 0 dBA for frequencies ≥100 Hz and reason for mounting the electrical panel outside of the cham- only 5 dBA at 80 Hz. The noise floor is shown in Fig. 11. The ber was to minimize the number of penetrations through the noise floor is verified regularly as part of the laboratory’s chamber walls. accreditation program and has not increased even though To maintain client confidentiality and secure client test substantial highway construction has been completed and devices, a control room was constructed adjacent to the traffic has increased in the area. Also, several construction chamber. The control room consists of three walls and a ceil- projects are ongoing in the area and none of them has affect- ing, which was constructed using 10.1 cm thick panels. These ed the noise floor in the chamber. panels are the same type of panel as used in the outer cham- The next test program was to verify the chamber’s free- ber walls. The fourth wall of control room is a section of the field performance. Using ETS-Lindgren’s automated traversing outer chamber’s wall, which contains a personnel access door rig, several sets of traverse or draw away data were taken in into the chamber. Instead of attaching the control room walls accordance with ISO 3745:2003 (Acoustics—Determination of and ceiling to the outer chamber walls, the control room is sound power levels of noise sources using sound pressure— separated from the outer chamber wall by 1 cm and the ceil- Precision methods for anechoic and hemi-anechoic rooms). ing along this wall is supported by columns that attach to the The chamber was found to have an ISO 3745:2003 compliant Fig. 11. Ambient sound pressure levels. Fig. 12. Deviations from inverse square law at 100 Hz and 80 Hz. 20 Acoustics Today, October 2009 field also accommodates a 3.6 by 3.35 by 3.0 m (Lx Wx H) par- allelepiped for ISO 3744:1994 (Determination of sound power levels of noise sources using sound pressure—Engineer-ing method in an essentially free field over a reflecting plane) noise testing. The completed chamber is shown in Fig. 13. Conclusion Several challenges arose during the design and construc- tion of this project. The low-noise requirements, chamber flexibility, and free-field performance each provided a unique set of challenges. The chamber has been in operation since January 2008 and has tested a wide range of specimens. It has proved to be successful at meeting each of its design goals and has proven useful in the development of new sources for Fig. 13. Inside the completed chamber. hemi-anechoic chamber qualification and free-field per- formance research. Due to the upfront planning and engi- free field for frequencies ≥80 Hz at a distance of 2 m, which met neering, the chamber will continue to be a useful tool well the chamber’s design specification. The free-field traverse data into the future regardless of infrastructure growth and for 100 Hz and 80 Hz are shown in Fig. 12. The chamber’s free expansion in the area.AT Douglas F. Winker is the principal acoustic engineer for ETS-Lindgren. He oversees the design of all of ETS-Lindgren’s acoustic chambers, acoustics projects, and research and development. Doug received a Ph.D. in Acoustics from the Department of Electrical and Computer Engineering at The University of Texas at Austin, and holds a patent for a sound source used in anechoic chamber qualification. He is a member of the Acoustical Society of America, the Audio Engineering Society, the Institute of Noise Control Engineering, the Institute of Electrical and Electronics Engineers, and the Phi Kappa Phi honor society. Low-Noise Hemi-Anechoic Chamber 21 ♦ Verso Filler Page ♦ RECENT ADVANCES in ACOUSTICS & MUSIC Design of an Acoustic Anechoic Chamber for Application in Hearing Aid Research MARC S. RESSL1 PABLO E. WUNDES2 GEDA (Grupo de Electrónica Digital Aplicada) Buenos Aires Institute of Technology (ITBA) Av. E. Madero 399, Buenos Aires ARGENTINA [email protected], [email protected] Abstract: - An acoustic anechoic chamber is a shielded room designed for performing sound measurements under conditions close to free space. This short paper summarises the design and construction of a low cost anechoic chamber, with a focus on hearing aid research. Under these conditions, small scale and a predominant axis of measurement are the major factors of consideration. Insulation, absorption and construction issues are detailed and addressed, preliminary results and proposals for future work are included. Key-Words: - anechoic chambers, sound insulation, sound absorption, sound measurements, hearing aids 1 Introduction 2 Chamber Design Acoustic anechoic chambers are environments with a 2.1 Shape high acoustic insulation from the nearby environ- Large anechoic chambers are usually constructed in ment, used to measure systems under conditions the shape of a rectangular cuboid, mainly due to ar- close to free-space. Due to the generic requirements chitectural limitations. Even though this set-up is of these measurements, anechoic chambers tend to be prone to standing waves, the fundamental frequency of considerable size, in order to attain an acceptable of resonance is usually low enough to be disregarded. response. This leads to high construction and mainte- nance costs, and the requirement of an appropriate In the case of a smaller chamber this effect is not physical space. negligible, thus a non-regular shape is preferred. In spite of this recommendation a rectangular cuboid However, if an anechoic chamber is designed for a was chosen as the shape of the design, as it simplifies specific set of applications, many of these limitations simulation and construction. can be overcome. In the context of the hearing aid project our group is engaged in, a chamber was re- 2.2 Dimensions quired for measuring the response of two, possibly The dimensions of the chamber are a critical design mismatched, microphones from a directional source. factor. If chosen wisely, the standing wave problem can be mitigated and certain parts of ISO 3745 can be In this context the physical size of the chamber can observed. be kept within constraints, since the frequency range of interest is 250 Hz - 4000 Hz, small scales are pre- For determining the internal width, height and length, dominant and a predominant axis of measurement is a common scale was multiplied by three small prime involved. The insulation of the chamber to external numbers. This achieves a mix of standing wave sources of sound and the reflective behaviour of the modes that do not overlap in the frequency range of walls can also be given special consideration. It is interest. even possible to comply with parts of ISO 3745 [1], a standard for performing sound pressure level meas- The chosen dimensions led to a volume of 1.103 m3. urements in anechoic rooms. ISO 3745 recommends a chamber volume (Vc) of at least 200 times the volume of the sound source (Vs), thus devices of up to 5515 cm3 can be measured. ISSN: 1790-5095 18 ISBN: 978-960-474-192-2 RECENT ADVANCES in ACOUSTICS & MUSIC In order to find the chamber’s cut-off frequency, the Transmission loss source is assumed to be directional, facing the major dB d dimension. In this case the geometric cut-off fre- quency can be obtained as [2]: (1) a + 4a + λ/2 + 2lw = d a represents the size of the sound source in the d di- mension, and lw is the wedge height (see 2.4 Absorp- tion). – Single wall (30mm thick) – Double wall (30mm thick, The sound source used in the set-up has a = 0.1 m. 50mm spacing) Solving for λ, the geometric cut-off frequency in the Hz d direction is found to be 211.7 Hz. This falls below the lower frequency of interest. Fig. 1 - Single and double wall transmission loss 2.3 Insulation Transmission loss is also highly dependant on the wall material’s density, increasing with materials of To determine the necessary acoustic insulation, the higher density. A survey of available materials was maximum SPL level at the site of our sound lab was realised, and MDF (Medium Density Fibreboard) was determined with a Brüel&Kjær 2250 sound level me- determined to be the best material matching our con- ter. It was determined that the maximum SPL is 60 straints. dB(A) in the frequency band of interest. The choice of a double wall insulation results in a Simulations were initially carried out with a single chamber that consists of a box inside a box (Figure wall insulation. If it is assumed that wave-fronts are 2). flat, interfaces infinite and there is no dissipation, the transmission loss at the interfaces can be calculated as 2 h (2) TL = -10 log10 ( |T| ) T being the quotient of two phasors that represent the complex amplitude of the incident and transmitted w pressure waves, respectively [3]. The wall width was varied in order to comply with the noise criteria curve NC-10. The resulting width was unacceptably large, thus a double wall insulation was considered. Simulations d were repeated for this configuration, and wall widths and separation were varied. Best values for these pa- Fig. 2 - Chamber structure rameters were determined by constraining the first high frequency transmission zero above the maxi- In order to improve the characteristics of the chamber mum frequency of interest, and forcing the low- even further, glass wool was added to the air gap be- frequency zero to a low as possible value (Figure 1). tween the walls. The quality factor of such zeroes is The values found were 30 mm thickness for both quite high, thus a small amount of absorption is ca- walls, and 50 mm for the air gap. pable of mitigating the effect of these zeroes. 2.4 Absorption To minimise echoes inside the chamber, an internal wall lining is necessary. ISSN: 1790-5095 19 ISBN: 978-960-474-192-2 RECENT ADVANCES in ACOUSTICS & MUSIC At first a survey of anechoic wedges was produced. Several factors were considered: absorption cut-off frequency, absorption coefficients, material, dimen- sions, shape and safety. None of the materials found on the local market was suitable, and importing for- 150 mm eign material was determined to be prohibitive. It was thus decided to design and construct our own an- echoic wedges. 30 mm The design is based on Beranek’s wedge structure A [4]. A very important design factor is wedge height, 100 mm as it is directly correlated to the wedge’s lower cut- off frequency. This frequency can be approximated by the following expression: Fig. 3 - Anechoic wedge (3) fc = c / 4h The response of the wedges is yet to be evaluated. The intended method of measurement is the imped- where c is the speed of sound in the chamber, and h ance tube method. A loudspeaker is located at one the height of the wedges [2]. end of the tube, and the material at the other. Sound waves travelling down the tube are reflected on the In order to meet the lower frequency of interest, the material to be studied. By moving a microphone in- value of h has to be at least 0.34 m. This, along with side the tube, the pressure differences of the standing the geometric cut-off frequency and the requirement wave pattern inside the tube can be measured and of a working space, imposes constraints on the small- used to calculate the amount of sound reflection. est dimensions of the chamber. For the purpose of a preliminary simulation of the Unfortunately the space in our lab was not sufficient chamber, the response of a commercial absorbing for a larger chamber, so it was decided to lower the material resembling our design was used. wedge height to 0.15 m. This raises the cut-off fre- quency of absorption to 571.6 Hz, and is above the 3 Simulation lowest frequency of interest for our research. Even though a certain level of absorption is expected in the 250 Hz - 571.6 Hz region, the effect of this decision The distribution of sound power inside the chamber has yet to be evaluated. was simulated in MATLAB with the source-image method. The method consists of replacing the cham- A second design factor is the wedge’s angle, a pa- ber’s walls with phantom sources, located at the rameter related to base area. A small angle leads to a equivalent geometric points from where the echoes of better acoustical gradient, yet a larger angle simplifies the walls would originate. manufacturing, handling and mounting. It also re- duces the total number of wedges necessary in the Only first reflections were considered in this work, chamber. thus six phantom sources were used. Low-density polyurethane (25 kg/m3) of low-stiffness The pressure distribution of a single source of fre- was chosen as the material for the wedges. This mate- quency f is [5]: rial is commonly used for the purpose of absorbing sound, is easy to cut, easy to adhere to surfaces, and (5) p(r) = A sin (ωt + kr) automatically fills voids, avoiding sound leakage. Fibreglass-based wedges were determined to be too where r is the distance to the source, and A follows expensive and difficult to manufacture. Melamine the inverse square law. foam was also disregarded for being too costly. It is very easy to calculate the power distribution by The design is presented in Figure 3. evaluating the pressure field at some arbitrary time t and at time t + 1/(4f). ISSN: 1790-5095 20 ISBN: 978-960-474-192-2 RECENT ADVANCES in ACOUSTICS & MUSIC double-deflection neoprene mounts of rated capacity Reverberant case, f = 250 Hz Absorptive lining, f = 250 Hz 136-272 kg were used (the inner box’s weight being d [m] SPL [dB] d [m] SPL [dB] 220 kg). Cork mounts were considered, but they did not match the working conditions. The outer box was constructed last. Glass wool was added to the boxes’ air gap. 4.2 Interface In order to access the inside of the chamber some interface is necessary, both for physical access to the working area, as well as for transmitting power and w [m] w [m] signals. Reverberant case, f = 1250 Hz Absorptive lining, f = 1250 Hz d [m] SPL [dB] d [m] SPL [dB] For the purpose of physical access, a double door was devised for the front of the chamber. The inner door consists of a rectangular cut to the inner box’s front panel, a rubber seal and a rectangular panel. This panel can be screwed tightly on top of the inner box, in order to guarantee a good seal. The outer door is constructed likewise but larger, so the inner door can pass through. Due to the larger weight of the outer door, a sliding mechanism was required (Figure 5). w [m] w [m] Fig. 4 – SPL simulations The simulations were carried out for a source located at mid-width, mid-height and at a length of 0.44 m. Perfectly reflective walls (the reverberant case), and walls treated with our absorptive lining were com- pared. The frequency was swept within the range of frequencies of interest, and the horizontal distribution of power for different values of h (height) was calcu- lated. The resulting power distribution was also com- pared to the distribution of the inverse square law. It was determined that the best location for measure- ments is in the area close to the opposite corner of the Fig. 5 – Access door chamber. There, variations in frequency response are minimal. The electrical interface consist of power and shielded signal cables that were routed through sealed holes in the inner and outer boxes. Standard BNC connectors 4 Construction were used for signals. 4.1 Structure 4.3 Anechoic Wedges The parts of the structure were cut from industrial The 450 wedges of polyurethane required for lining MDF panels with a computer controlled saw fence to the inside of the chamber were fabricated using a 0.5 mm precision. The inner box was constructed on pantograph hot-wire cutter. top of the base of the outer box, and properly sealed. Modules of 2x3 wedges were constructed by placing In order to isolate the inner box from the outer box, wedges in alternating orientations on top of a ply- and to keep the necessary air gap of 50 mm, four wood base. The modules are fastened to the walls ISSN: 1790-5095 21 ISBN: 978-960-474-192-2 RECENT ADVANCES in ACOUSTICS & MUSIC through Velcro. It is thereby easy to convert the an- Transmission loss dB echoic chamber to a reverberant one. – Single wall Special care was given to covering the corners as – Double wall well as possible. Where necessary, strips of polyure- – Measured response thane were applied. The wedges were also slightly compressed, in order to avoid sound leakage. 4.4 Stands Small wooden stands were designed in order to locate the objects to be measured inside the chamber’s f working space. Fig. 6 – Preliminary insulation result 5 Results (outer box not sealed) the frequency band of interest. It would also be of Preliminary measurements of the chamber show very interest to determine the optimum chamber parame- promising results. ters through simulations. At this stage the outer box has not yet been sealed, The simulations should be verified experimentally. nevertheless the acoustic insulation response (Figure 6) is superior to the insulation of a single wall insula- 7 Acknowledgements tion in the frequencies of interest. The response above 6 kHz is slightly below expectations. This re- sult can be attributed to an improper seal of the inner The authors gratefully acknowledge Pedro Bontempo box. and Marcelo García Barrese for their work in the de- sign and construction of the anechoic chamber. Preliminary impulse response measurements [6] show a T60 reverberation time of under 50ms. The References: frequency response was found to be flat within 3 dB [1] ISO 3745. Determinación de los niveles de poten- in the frequency range of interest. cia acústica de las fuentes de ruido a partir de la pre- sión acústica. Métodos de laboratorio para cámaras 6 Conclusions anecoicas y semi-anecoicas. A design of a small, low-cost and application specific [2] Blanco M., Herráez M (1993). Calificación de la anechoic chamber was presented. Preliminary results cámara semi-anecoica del laboratorio de acústica y are very promising, but several questions remain to vibraciones de la ETS de Valladolid. SEA- be answered. TecniAcústica. p. 167-170. The first issue to be addressed is whether the effect of [3] Blackstock D. (2000), Fundamentals of Physical the geometric cut-off frequency in the directions of Acoustics, Wiley. width and height do not pose a limitation to the use- fulness of the chamber in low frequencies. [4] Beranek L., Sleeper H. (1946), The Design and Construction of Anechoic Sound Chambers, The The absorption response of the custom-designed an- Journal of the Acoustical Society of America echoic wedges is to be measured. It also remains to be seen what the effect of limiting the wedge height [5] Puep Ortega B., Romá Romero M. (2003), Elec- to 0.15 m is. The preliminary impulse response troacústica Altavoces y Micrófonos, Pearson Prentice measurements look promising, but a proper study of Hall. the absorbing material should be carried out. More simulation experiments should be realised, like calculating the mean and variance of the SPL among ISSN: 1790-5095 22 ISBN: 978-960-474-192-2 RECENT ADVANCES in ACOUSTICS & MUSIC [6] Farina A. (2000), Simultaneous Measurement of Impulse Response and Distortion with a Swept-Sine Technique. AES2000. ISSN: 1790-5095 23 ISBN: 978-960-474-192-2 ! " # ! $ ! ! " # ! $ % & %&# '# ! " ! # ! # ! $ % # & # ! ( '' ' ''' " ' ''( & ' '( ) ' '* + , ( '- & * '-' - '-( . '/ . '/' 0 , " 1 '/( 2 3 '/* 2 3 '/- 2 4 '// 2 '5 '/. $ '5 '. , 6! 7 '' '.' '' '.( '( '1 ) '- '1' " ) '- '1( + 8 '/ '1* & $) '1 '1*' 7 '1 '1*(& '3 '1** , & '4 '1- $) (5 '1-' 9 7 :5;(' '1-( & $ (' '1-* (( '1-- , (( '3 2 ) (* '4 & & (- '4' (- '4( % (1 '4(' 0 # 2 #(3 '4(($)& (4 '4(*0 $)& */ '4* " *1 '4- *3 '4-'& , <& " *3 '4-(& -5 '4-*, -' '4-- 7 -( '4-/6 , " -* ''5 = -- ''5' -- ''5( " -/ ''5* , 7 -. ''5- ) -. ''5/ 2 -. ''' , -. '''' = -1 '''''$ -1 ''''( -1 ''''* " -1 ''''-" -3 ''''/ -3 '''( & -3 '''('$ -3 '''(( -4 '''(* " -4 '''(-" /5 '''(/ /5 ! " ' .1 ( .3 * = .4 - , 15 / , 7 1' . = 1* 1 >& 1* 1' &&= 1- 1( ! &" = 1- 1* &= 1/ ' $ & & % , 7 7 7 8,'5: 88%? ; @ , ! ) 7 # # 7 A A 7 ) # B C ) 7 '5D(5 B C 7 B C $ 9 ) ' $ A A #: ; ('55#'/55#'*55: # # ; *5" ! *5 7 (41 EA # 6 # ) 3 # # '/C# '/C,#:'/? *38; # 7 */5$ 344 ' 7 (/ */55EA 7 45$ '. 7 /55EA (5FA :! (? ; '/? *38 ) #'/? *3 8) 7 '3FA> 7 3>'. .55$ = 7 # 7 3(*5#(*5#'-3: ;E 35$ '5( :' '; 7 155EA (*FA 7 355EA #155#3/5#*55: # # ; # '15 '-5 E '-(3 # 7 :;-*4EA 7 &= * :*;-4/EA 7 7 # 155#3/5#*55: # # ; 3 45 36'*5 -6$ )= :**EA 15EA; , G (5EA (5FA:'5 ; 5'H E ,'+ : , 8 ; * & ) :& ; 7 7 ! * + 7 & F % 7 @ #@ ( A @ - 6 A * # 67 @ ( E A # . # % # $ % 0 @ ( ( ( :'-' '; $ @ #A :; :5'(*-; 7 % % I'I5I' % '5' % 7 B C& % ) !'- ' : ' ; % # % ! ! ! " " 7 7 % 7 ) :'-' (; 7 % '3-4 ! "! # $ @ 0 /55EA'FEA 2 :*; ) # % & 2 7 A 7 A 7 + % A (55EA $ # 7 & ! # & !'- ' & ('55#'/55# '*55: # # ;) !'- ( #% # # 7 # $ " % ! $ J6 7 ) 7 & :'-' *; 7 1-EA 7 "! # # '4*" *1 * , , :2; )&6*1-/ (55 + - "" $ 7 # $ # $ %&& ) 7 :!'/ '; ' '( - ," ' "$ )&6 *-/@(55* B" # @ $ # %# & # C ) E : ;$ ' '55H ) A 58 )&6*/- : 5 (; & $ %&& 6 7 7 '55HJ$ 7 : ; 7 7 $ & 9 7 @ - . " ) ) & & -) . " F ) 6 ) ) $ %&& -5F>*'55F>* '55F>* E $ : ; -* . " / % 7 ! 7 7 KK $ :!'/ (; " ! ( ! ( + +( ! ( ,-(( ) 7 L L 7 M # 7 7 L>- L>-& # 7 7 @ ' :'/- '; - - ) !'/ * & 7 7 6 " ! ( ! ( + +( ! ( ,-./01 * $ %&& # $ $% # 7 # 7 BC >- ( !+ + ( >- ( ! ( :!'/ -; ( /0! BC 7 & 7 7 6 AA " & 8 # B C : ; : ; & $ J # :!'/ /; 3 ' ! 2 $ %&& & : ; : ; 7 7 7 7 '55H 7 7 7 :'.' ';& 7 7 # # # # :'.' ; 0 12' 345 ) 7 )&6*1-/ 6 7 " % )&6*1-/ : G ; 7 7 7 :'.' '; 7 ) ) - - :'.' '; : # # ; A : #; 7 7 & *5 7 544 7 (31/ EA (5FA ' () * +, " %' %$ )&6*1-/ 8=& @ % % # @ ( B* ' 7 ) 8=& & 7 7 # @ :'.( '; ! @ :'.( (; )&6*1-/) :' ; & 7 # @ :'.( *; ' () * +, ) + 0 : ; 0 ) ( ('55#'/55#'*55: # # ; *5 # '/55#455#155 : # # ; (31/EA 7 7 & 7 :'5 ;.'(/EA:8 7 ; 7 (31/EA '1 :*5 ; ! '- 7 - %& 6 "! ) :N; : ;: ; ! # 8,'5 7 7 $ 6 $ ! 7 , @ & : ;@ :'; :(; ) # @ :'1' '; & : ;@ ) # 6 @ 7 # @ :'1' (; $ @ I 7 :; I 7 7 :(; : ;@ # : ; @ - %& :'1' *; $'$( ) @ :'1' -; $ @ &I& :(; I%7 7 :(; E & :& ) ; :& ) & ; 6 %/ 7 & " < & A ,%&2 &,*'5M A :' ; A 8, 8,'5 # 7 7 7 !'1 ' 8,(/ 8,*5 7 7 !'1 ( *5 - %& & 4 ! 5 ! ' !'1 ( *5 , 7 7 7 : ; 7 :' ; 7 & 6 ! 5 ! ' # - %& ! !'1 * 7 7 7 , .*EA -5 .*EA *5 7 (5 & "' ! ! ' 5 6) & 8 ! & : ; 6) 12 5 ! 7 : ; 7 # @ :'1* '; $ @ I= :I(O;:EA; I& :F>(; NI :F>(; & - %& . # # 6) & 2 7 7 9 7 :5; :'1* (; 7 :,; 7 7 :'1* *; $ @ I :>; I$ :; I$ :F>*; PI= %IQ :8>(; 7 7 7 ) ! R R:K5; 7 R5R A R R A 5KK, # B R S ) ) - %& & 8 ! 6)) 7 ' & 2 # F 7 9 7 7 *5 .45F> * # '3'(1(51*5F>* '3 F # 7 *5 :' ; :; :F>*; & ":F>(; *5 .45 (51 '3 1(5 '(4. ('1( '( 1*5 31. & - ! '! ! ! ! 2( ! ) * - %& & !'1 . *5 & # 8 8 !2 ' ! !9 ' , 7 7 : .*EA; 7 E*5 .*EA 7 '5 (5 # # (5 8,'5! 7 *5 (5 -5 5 6* " 8 ! 7 BC 'G(I BC 7 7 ) R 7 CB 7 R 2 - %& 6* # 9 5: ; 0 '( F 7 # @ :'1- '; $ @ FI0 :8>*; ( '(I& :F> ; ) 7 # @ :'1- (; $ @ I :; ) 7 5 7 7 5 7 7 8 7 :'55EA; 6* # & % 2" 8 1 7 # @ :'1- *; :'1- -; $ @ I'(*T I :>; I :; - %& ! 7 I'G( : ; 6 6*) # ) ( 7 ) ( 7 7 ( 7 :,'; :,(; 6** # ' " 2 *5 '3'( ) !'1 1 5 3*EA 7 A : 1 7 ; 7 ) : ; # 7 '3'( 7 - %& & & - 8 !2 ,):1 & 7 @ 5I3*EA: #5 ; 'I*-*(EA:.5 ; (I.3.-EA:.. ; *I'5(4.EA:15 ; -I'*1(3EA:1( ; 7 8,'5 ! 9 7 7 9 7 < ." ! 7 7 $ 7 7 . /& 7 E 7 ! 5>'/ !5 7 7 ( (5EA 7 -/EA = & & # 7 7 3 45 ' 7 = 1 / 7 7 & F : ; 7 A A 7 : ! '4 '; 7 * 7 ( '!! & 7 & # ' ) '/C (C ' & # ? , # 0 % # ) 7 :*/ (5555 EA; !'4 - 7 A # # , # @ ! 7 @ 8 *35'/ 0 ) 3 = , */5 %&( & 44 (3*2 ' ! 7 (/D*/55EA * ! ; )4 $ !9 ' ! 7 @ 0 ) '. = , 45 %& 'FEA & '5/ ' ! 7 /55D(5555EA * ! ; )4 $ !9 ' ) ! '4 - :E(<E*;) = '5 (5 & * 9 ' ( ! !$ (5 @ ( %&( '43-: (55*; 0 % ( , ! 7 '3FEA = , .55 >E! 7 ) 3'. & '( > E! 7 5 '/ * ; 7 '! E 7 & 7 7 # 7 # 7 ) # @ 6 :$#E# ; (*5#(*5#'-3 0 ) 3 = , 35 %& 'FA & '5( ' ' ! 7 155D(*555EA , 7 355EA * # 7<=2> ) !'4 1 # 'FEA (5 FEA 7 A # 355EA 7 -5 * & 9 ' ( ! ! # 0 % = 1 /+ % # # ! # : 7 ; 7 * ) ! = 9 (5 9 % > > & 0 % # ='/5E 7 7 7 7 ! )& !2 = % (. %( # : # # ! '4 4; 7 # # # ) # # R R * * ? ! $ # 7 # E A # # 7 # 7 7 7 7 7 7 7 7 A ! # 7 7 # 7 7 ) 0 % 2 # # # 7 # : ; M :; # # 7 7 # = 8!&" ) $4#>8 Q , 2 # ) * 2 @8-! ! ,- @ * 0 % * @ (5 ( ! (5! B6 C # ,- >& : # 1*; 7 ,# : ' '55; ) * @ 9 ' < ! = (5 2 0 % , # # , 2 : 0 #; - . % =:% = ; ) 2 * @ '(! * * # ? ( ? ( $)& * # - ) $)& ) 8=& ( 8=& 0 % ) . !'4 '1 $)& 2 7 , 7 # * & ?$ ,6 7 !9 '1 = . / % !'4 '3 $ # ## # # # # * ) ?$ ( 0 % ) !# ) ' 3 $ /H :U'1>; ) /H ) 7 7 * * 6 ! ) 0 # = 8=& * 2 6 ( 0 % ! 0 1# / -## !'4 (' * <A , ) 7 7 7 # , ' & # # @ >& @ 2%7 , 2 V 2 ! !0 7 &=& ' # 7 &= # ' ' : 0 #; 0 % * 6 $ ! $ ! = ) " 2 8!& ! # '/? *38 @ ,#'/? *38 V 5(4 V 1/ 2:; */4 V 5* !:EA; ** &=: ; 44 0 :W; .3 = :$; */5 '. '3( ?#:; - :; 5** N:W; 3 &:(; 5533 * ! $ ; )4 0 % $ # *5" ! @ 2 :; '-(3 6 :$#E# M; 1.5#4'5#*55 2 :; ''3 2 :; '15 ! :EA; -*4 * *:EA; -4/ 2 553 * ! ? ! $ ) !'4 (/ # : ;) 7 -*4EA: ; 7 : * ; 7 7 *5EA # 7 # -*4EA : # ; * B! ? ! $ # # 7 !'4 (., 7 7 # 7 :-*4 EA; ) # 0 % 7 7 * # 6 ! - '! ? ! $ =) $ 6 ,' F 7 &80 & &80 8 7 : ; # # & ) # 7 7 7 & 0 % =* 7 " 7 7 ! 3 & ,?'.3 7 =* &2 ' >& $ ,?'.3 ' 3 # ,?'.3 % & = ! # '<( *<- M /<. 1<3 @ , '( , *- , /. , 13 Q X )= ) ! *</ )= * ) % * & ! -8< ( ) 0 % ! ,'<( ,1<3 & " @ & Y XY X& * * % ="- ! = " @= % X * 2 % ="- !( " @ : ,*<, -; - ,?'.3 ,' ( ,* - ,/ . ,1 3 Z X * 0 % X * % ="- ! =* & ,?'.3 #, 2 0 % E & 8 =*) ' 1 ,?'.3 3 7 A & '<'5 I, 68 & '<'5I, % X 6!! ! B C 7 ) , 7 * 7 ( (= ! 7 : 7 ; B C ) $ 7 , Z )68[3 % # # & &% - & &- # # # # 4 # - # # % # # %% # & #%#% -. ## # % - # # %5 0 % =** 125 $ 7 7 7 X 7 7 7 = 7 & 7 7 7 7 ) 7 7 7 ! 7 7 @ & * @ 9 6!! & *@ 9 68 & 3 @ 9 6!! & 3 @ 9 68 & ( @ 9 7 **EA * & (@ 9 7 15EA ( & 4 @ 9 7 **EA !9 '! & 4 @ 9 7 15EA & @ 6!! 7 **EA 7 F " 7 7 7 Z 7 7 # 7 X **EA 15EA 7 :** 15E A; 68 0 % * = !9 '! ( ,C!! >D &2>D1 =*- 3 ' $ : ; ) 0 7 & = " @) ' * ( (5 ( ! # ( " (5 ((( ',: :3 1 0 % " @) !'4 *. ( ) * ' + " # & " " ! * # ? (5 ( ! #, 1" (5 (( ( ',: :3 1 ? ! # % 155#'555#3/5: # #; #:*5; 1.5#'5.5#4'5 : # #; 1 % @ 7 A '1'EA$ # 7 7 % $ # # "" $ ) $ '55 ! /5 !''5 'E 7 7 2 ( !! !2 (' ( /5 :8 : ;; & 6 & & 1 & '(5 $ !''5 ( '55 '(5 !:EA; '(/ (/5 /55 '555 (555 -555 0= :'55; 5-' 53' '51 '51 '5' '55 2 ( !! !2 ! ) ' 5 ! ''5 ( & '55H 0 7 & 7 * "! # :*5; '3'( - ." # # # ' ! A # # ? $ # E B# C 8 % 1.5#'5.5#4'5: # #; *5 '3'( ( & & & # 1.5#4'5*5 * ) "" $ # : #; & # (* # & * $ # ) # 3 # /5 # - - # - % ) ( $ 0 # =2,: #;'1 ''3 # - ) # & 8 % # # # ) # 6 - #& E # - "- # # # # ) !''' ' 7 ( ! ( * ) "" $ # # # 7 & & #& F # * $ ) ) # # ) # - % \ # #& A - 2 ) $ * + ,-$'443 ]^ 7 _# # *+. /0(55' " E +# # "+ E *+- 1'443 E !E6 # 8 *2+ / 3( .'( ( 4% # 1 ((555 ! +# # \$ <&" )& 8@3- *.3 53-5 ' ) % -5 @>> >>"`= @>> > > @>> > > > > ` @>> > 6 #$1 5 0 &" a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iche Technique PYRAMIDE Le panneau PYRAMIDE est une mousse à haut pouvoir d’absorption, en mélamine ou polyuréthane (PU), à relief pyramidal. Les panneaux PYRAMIDE sont particulièrement adaptés à l’absorption acoustique dans les locaux et à la diminution de la réverbération. COMPOSITION : Mousse absorbante à relief pyramidal en mélamine ignifugée ou polyuréthane auto-extinguible (PU). AVANTAGES : - Excellente absorption acoustique sur une large bande de fréquences. - Amélioration du confort acoustique. - Non-fibreux et non-volatile (contrairement aux laines minérales). - Mise en œuvre collée économique et facilitée par son faible poids. - Résistance au feu M1 pour les versions en mélamine. - Possibilité de mise en œuvre en doublage de capotage pour la version plane… CARACTERISTIQUES TECHNIQUES : Polyuréthane (PU) Mélamine Dimensions : 1000 x 1000 mm 1200 x 600 mm Epaisseurs standard : 50 et 100 mm 50 mm Aspect : Gris anthracite Gris clair ou blanc Poids : 30 kg/m³ 11 kg/m³ Combustibilité : M4 M1-B1 PERFORMANCES ACOUSTIQUES : Absorption acoustique 1 0.9 0.8 0.7 0.6 f (Hz) 125 250 500 1000 2000 4000 0.5 Pyra PU 50mm 0.12 0.24 0.40 0.68 0.69 0.69 Abs.Coeff 0.4 0.3 Pyra Méla 50mm 0.13 0.25 0.50 0.75 0.88 0.94 Pyra PU 50mm 0.2 Pyra Méla 50mm 0.1 0 100 1000 10000 f(Hz) MISE EN ŒUVRE : Panneaux à découper au couteau, et à coller directement sur le support propre et sec. APPLICATIONS : - Absorption acoustique dans les locaux (musique, voix,…) - Diminution de la réverbération (acoustique des salles). - Doublage absorbant de capotages et encoffrements de machines. dB MUTE SPRL - INSONORISATION ET TRAITEMENT ACOUSTIQUE - ETUDES / REALISATIONS / MATERIAUX 57, rue de Theux B-1040 Bruxelles Tél. +32(0)2 538.63.10 Fax. +32(0)2 538.81.39 e-mail: [email protected] Website : www.dbmute.com - TVA 873.201.720 - N° enr. 03/16/01 - ING 310-1890717-79 15XA38Nd15XA38Nd C o a x i a l KEY FEATURES • 15” bass loudspeaker and 2” exit compression driver combination • Excellent power handling: 350 w AES (L.F. unit) and 90 w AES (H.F. unit) • Extended and linear frequency response (25-20000 Hz) • High sensitivity: 99 dB (L.F. unit) and 105 dB (H.F. unit) • Low weight (common neodymium magnet system for both units) • Bass loudspeaker designed for compact bass-reflex cabinets GENERAL DESCRIPTION This 15” coaxial loudspeaker is intended for the most demanding professional applications. Its low frequency unit features a 4” (100 mm) edgewound aluminium ribbon voice coil capable of handle 350 w AES. This bass unit has been optimized in order to fit in with the most compact bass-reflex systems. The high frequency unit uses a 2.8” (72.2 mm) aluminium ribbon voice coil and a composite structure diaphragm, that is to say, a titanium dome and polyester surround combination. All these components give as a result a coherent and extended frequency response with low distortion that ranges from 25 Hz up to 20 kHz. Moreover, the use of a common neodymium magnet system for both units reduces the weight to 6.8 kg. FREQUENCY RESPONSE AND DISTORTION CURVES FREQUENCY RESPONSE OUT OF AXIS Note: on axis filtered frequency response measured with loudspeaker standing on infinite baffle in anechoic chamber, 1w @ 1m. PREDICTED LOW FREQUENCY RESPONSE FREQUENCY RESPONSE OF LF & HF UNITS Note: bass-reflex cabinet, Vb=100 l, fb=42 Hz Note: on axis frequency response of low and high frequency units, 1w @ 1m. 15XA38Nd15XA38Nd C o a x i a l TECHNICAL SPECIFICATIONS DIMENSION DRAWINGS L.F. UNIT Nominal diameter 380 mm. 15 in. Rated impedance 8 ohms. Minimum impedance 7.6 ohms. Power capacity* 350 w AES Program Power 700 w Sensitivity 99 dB 2.83v @ 1m @ 2p Frequency range 25-3500 Hz Recom. enclosure vol. 60 / 180 l 2.14 / 6.35 ft.3 Voice coil diameter 100 mm. 4 in. Magnetic assembly weight 4.2 kg. 9.24 lb. BL factor 18.2 N/A Moving mass 0.072 kg. Voice coil length 16 mm. Air gap height 9 mm. X damage 28 mm. MATERIALS H.F. UNIT L.F. UNIT Rated impedance 16 ohms. • Basket: Die cast aluminium Minimum impedance 13.5 ohms.@ 3.5 kHz • Cone: Paper Power capacity 90 w AES above 1 kHz • Surround: Plasticised cloth Frequency range 0.5 - 20 kHz • Edgewound aluminium ribbon Sensitivity 1w @ 1m 105 dB Voice coil: Voice coil diameter 72.2 mm. 2.87 in. • Magnet: Neodymium Flux density 1.6 T BL factor 15.3 N/A H.F. UNIT Dispersion 80° conical • Dome: Titanium • Surround: Polyester • Voice coil: Edgewound aluminium ribbon THIELE-SMALL PARAMETERS • Voice coil former: Kapton Resonant frequency, fs 33 Hz MOUNTING INFORMATION D.C. Voice coil resistance, Re 6.8 ohms. Mechanical Quality Factor, Qms 7.50 Overall diameter 388 mm. 15.28 in. Electrical Quality Factor, Qes 0.30 Bolt circle diameter 370 mm. 14.56 in. Total Quality Factor, Qts 0.29 Baffle cutout diameter: Equivalent Air Volume to Cms, Vas 359 l - Front mount 352 mm. 13.85 in. Mechanical Compliance, Cms 326 mmmm / N - Rear mount 352 mm. 13.85 in. Mechanical Resistance, Rms 2 kg / s Depth 181 mm. 7.13 in. Efficiency, hhho (%) 4 Volume displaced by driver 7 l 0.25 ft.3 2 2 Effective Surface Area, Sd (m ) 0.088 m Net weight 6.8 kg. 14.96 lb. Maximum Displacement, Xmax 4 mm Shipping weight 7.4 kg. 16.28 lb. Displacement Volume, Vd 350 cm3 Voice Coil Inductance, Le @ 1 kHz 1.6 mH Notes: *The power capacity is determined according to AES2-1984 (r2003) standard. **T-S parameters are measured after an exercise period using a preconditioning power test. Program power is defined as the transducer’s ability to handle normal music program material. The measurements are carried out with a velocity-current laser transducer and will reflect the long term parameters (once the loudspeaker has been working for a short period of time). Polígono Industrial Moncada II · C/. Pont Sec, 1c · 46113 MONCADA - Valencia (Spain) · Tel. ( 34 ) 96 130 13 75 · Fax ( 34 ) 96 130 15 07 · http://www.beyma.com · E-mail: [email protected] · FD2XA / FD2XC1 / FD2XC2 FILTERS KEY FEATURES • 2 way filters • 12 dB/oct. attenuation slope • Additional high frequency attenuation • High power components • Specially designed for the use with Beyma coaxials GENERAL DESCRIPTION These 2 way passive filters have been specifically designed to be used with the XA and XC Beyma coaxials, in order to achieve an optimum performance of these drivers. Please, see attached filter selection guide to find the suitable filter model for each coaxial model. These filters offer the possibility of an additional attenuation of the high frequency unit. Anyway, this attenuation is only optional, optimum performance of the coaxial is achieved without using it. However, if you wish to use it you can activate it by simply removing the aluminium bridge labeled as «AT» on the back surface of the board (please see attached picture). TECHNICAL SPECIFICATIONS ADDITIONAL HF ATTENUATION Type 2 way Crossover frequency* 1.8 / 2 / 2.2 kHz Power capacity 600 w Low / High frequency impedance 8 / 16 ohms. Attenuation slope 12 dB/oct. H.F. attenuation 0, -1.5 dB *Note: crossover frequency depends on model (FD2XA / FD2CX2 / FD2XC1). FILTER SELECTION GUIDE FD2XA Use with 15XA38Nd & 12XA30Nd FD2XC1 Use with 12XC30 & 8XC20 FD2XC2 Use with 10XC25 TRANSFERENCE FUNCTIONS Note: FD2XA transference function. Note: FD2XC1 transference function. Note: FD2XC2 transference function. Polígono Industrial Moncada II · C/. Pont Sec, 1c · 46113 MONCADA - Valencia (Spain) · Tel. ( 34 ) 96 130 13 75 · Fax ( 34 ) 96 130 15 07 · http://www.beyma.com · E-mail: [email protected] · TPL-150H PLEATED DIAPHRAGM TWEETER WITH HORN KEY FEATURES • Superior sound reproduction quality • Extended frequency response (0.7 - 23 kHz) • Coverage angles of 80º in the horizontal plane and 30º in the vertical plane • High sensitivity (102dB) • Extremely linear frequency response • Low distortion • Precise directivity control in the pass band TECHNICAL SPECIFICATIONS Rated impedance 8 ohms D.C Resistance 4.9 ohms DIMENSION DRAWINGS Power capacity* 80 w AES above 1 kHz Program power 160 w above 1 kHz Sensitivity 102 dB 1w @ 1m Frequency range 0.7-23 kHz Recommended crossover 1kHz or higher (12dB/oct. min) Horizontal beamwidth 80º(+9º,-20º) (--6dB, 1.2-16 kHz) Vertical beamwidth 30º(+27º,-21º) (--6dB, 1.2-16 kHz) Directivity factor (Q) 27 (average 1.2-16 kHz) Directivity index (DI) 13 dB (+6 dB, -4.5) Cutoff frequency 800 Hz Overall dimensions (WxHxD) 230X230X148 mm. 9.05x9.05x5.8 in. Cutout dimensions (WxH) 195x195 mm. 7.68x7.68 in. Notes: Net Weight 3.68kg. 8.10 lb. *The power capacity is determined according to AES2-1984 (r2003) standard. Program power is defined as the transducer’s ability to handle normal music program material. Shipping Weight 4.08 kg. 8.76 lb. **Sensitivity was measured at 1 m distance, on axis, with 1 w input, averaged in the range 2.5 - 20 kHz. FREQUENCY RESPONSE AND FREE AIR IMPEDANCE CURVE DISTORTION Note: on axis frequency response measured in anechoic chamber, 1w @ 1m. -6dB BEAMWIDTH DIRECTIVITY INDEX Note: Horizontal beamwidth is represented by heavy line. Vertical beamwidth is represented by the discontinuos line. 108 CX168 8-Channel Installed Sound Professional Audio Amplifier User Manual *TD-000095-00* TD-000095-00 rev.C 1 IMPORTANT SAFETY PRECAUTIONS & EXPLANATION OF SYMBOLS CAUTION CAUTION: TO REDUCE THE RISK OF ELECTRIC SHOCK, DO NOT REMOVE THE COVER. NO USER-SERVICEABLE PARTS RISK OF ELECTRIC SHOCK INSIDE. REFER SERVICING TO QUALIFIED PERSONNEL. DO NOT OPEN The lightning flash with arrowhead symbol within an equilateral triangle is intended to alert the user to the presence of uninsulated “dangerous” voltage within the product’s enclosure that may be of sufficient magnitude to constitute a risk of electric shock to humans. The exclamation point within an equilateral triangle is intended to alert the user to the presence of important operating and maintenance (servicing) instructions in this manual. The lightning flashes printed next to the OUTPUT terminals of the CX168 amplifier are intended to alert the user to the risk of hazardous energy. Output connectors that could pose a risk are marked with the lightning flash. Do not touch output terminals while amplifier power is on. Make all connections with amplifier turned off. WARNING: TO PREVENT FIRE OR ELECTRIC SHOCK, DO NOT EXPOSE THIS EQUIPMENT TO RAIN OR MOISTURE. This amplifier has a serial number located on the rear panel. Please write this and the model number down and keep them for your records. Model: CX168 Serial Number:______Date of Purchase:______Purchased From:______ FCC INTERFERENCE STATEMENT NOTE: This equipment has been tested and found to comply with the limits for a class B digital device, pursuant to part 15 of the FCC rules. These limits are designed to provide reasonable protection against harmful interference in a residential installation. This equipment generates, uses, and can radiate radio frequency energy and if not installed and used in accordance to the instructions, may cause harmful interference to radio communications. However, there is no guarantee that interference will not occur in a particular installation. If this equipment does cause harmful interference to radio or television reception, which can be determined by switching the equipment off and on, the user is encouraged to try to correct the interference by one or more of the following measures: - Reorient or relocate the receiving antenna. - Increase the separation between the equipment and the receiver. - Connect the equipment into an outlet on a circuit different from that to which the receiver is connected. - Consult the dealer or an experienced radio or TV technician for help. © Copyright 2002, QSC Audio Products, Inc. QSC® is a registered trademark of QSC Audio Products, Inc. “QSC” and the QSC logo are registered with the U.S. Patent and Trademark Office The Audio Precision logo is the property of Audio Precision, Beaverton OR All trademarks are the property of their respective owners. 2 TABLE OF CONTENTS INTRODUCTION: CX168 Overview ...... 4 Front Panel Illustration ...... 4 Rear Panel Illustration ...... 5 Mounting Dimensions ...... 5 INSTALLATION: What is Included...... 6 Rack Mounting ...... 6 Supporting the Rear of the Amplifier...... 6 Fan Cooling...... 7 AC Mains (AC Power) Connection...... 7 SETUP: Setting the Configuration (MODE) Switches...... 8 Setting Operating Mode (Stereo, Parallel, or Bridge)...... 9 Clip Limiter Setting...... 10 Low Frequency (Subaudio) Filter Settings...... 10 Low Frequency Filtering Tips and CX168 Frequency Response Curve...... 11 CONNECTIONS: Inputs: Connecting to the Input Terminal Block Inputs ...... 12 Connecting to the DataPort Inputs...... 13 Outputs: Connecting the Outputs in Stereo or in Bridge Mode ...... 14 How to use the 8-pin Output Connectors...... 15 OPERATION: Power Switch ...... 16 LED Indicators ...... 17 Gain Controls ...... 18 Security Plate for Gain Controls ...... 19 DataPort Connector ...... 20 DataPort Guidelines for the CX168...... 21 APPLICATIONS: Four Room, Stereo Feed...... 22 Tri-Amplified, Stereo Cabinets with Subwoofer...... 22 APPENDIX: Detailed Explanation of Stereo, Parallel, and Bridge Operating Modes...... 24 Multiple Speaker Loads in Series...... 26 Multiple Speaker Loads in Parallel...... 27 TROUBLESHOOTING ...... 28 SPECIFICATIONS ...... 30 WARRANTY INFORMATION ...... 32 HOW TO CONTACT QSC AUDIO PRODUCTS ...... 32 3 INTRODUCTION Congratulations on your purchase of the CX168 power Alternatively, these DataPorts facilitate connection of signal amplifier. To help you obtain the best results from your processing equipment, such as QSC’s DSP-3, which offers two purchase, we encourage you to carefully review this manual. channels of independent digital signal processing - including With information including connections, configuration and crossover filters, shelf filters, signal delay, compression, peak operation, it offers many useful guidelines. limiting, and parametric filters. The CX168 provides a high level of channel density in a small, As individual channel pairs may be bridged, the CX168 is lightweight form factor ideal for multi-zone audio systems. configurable as a 4-, 5-, 6-, 7-, or 8-channel unit. Further, the Representing the culmination of QSC’s extensive experience CX168 features detachable Phoenix-style input and output in power amplifier development, the CX168 is a highly connectors—enabling the audio system to be wired with versatile, reliable, and user-friendly tool that will likely remain greater ease. Other features include 1-dB recessed and a central component of your system for years to come. detented gain controls and a security cover for tamper-proof installations. Like the entire CX Series, the CX168 is equipped with QSC’s exclusive PowerWaveTM switching power supply technology The CX168 is easy to use. All operating mode switches are to virtually eliminate noise and hum while reducing the unit’s grouped together on the rear panel, with one switchblock per overall weight. Comprehensive protection circuitry includes channel pair. These switches enable clip limiter on/off, LF DC, infrasonic, thermal overload and short circuit protection. filter on/off, LF filter select and operation modes. The amplifier’s front panel includes corresponding LEDs to indicate The CX168 includes 4 HD15 DataPorts (one per channel pair) Bridged or Parallel modes in addition to Signal and Clip for remote management or DSP. These ports allow each status. channel pair to be governed by QSControl, QSC’s audio networking system—enabling the system operator to control Your CX168 power amplifier was designed to provide many amplifier gain levels, check the unit’s clipping and thermal years of trouble-free, great-sounding operation. We hope you status, plus monitor numerous additional system parameters. enjoy your new CX168. FRONT PANEL FRONT PANEL WITH SECURITY PLATE 4 INTRODUCTION REAR PANEL DIMENSIONS AND MOUNTING POINTS 5 INSTALLATION: UNPACKING & RACK MOUNTING What is Included Save the container and packing material Your CX168 shipping container, as shipped from the factory, so the amplifier may be shipped without includes: damage if service is ever required. If the original container is not available, be - CX168 eight channel audio power amplifier sure to use a strong shipping container with enough packing material to prevent the amplifier from being - this user’s manual damaged in transit. - security cover (plate) for gain controls - self-adhesive rubber feet (use for non-rack mount applications) - eight 3-pin terminal block input connectors - two 8-pin terminal block output connectors - #14AWG IEC-type detachable power cord Rack Mounting Use four screws and washers to secure the amplifier to the equipment rack rails. Support the weight of the amplifier while securing it to the rails to avoid bending or distorting the mounting ears. The amplifier may be used in an equipment rack or as a stand-alone unit. Rack mounting is optional. Self-adhesive rubber feet are provided for non-rack mount applications. Supporting the Rear of the Amplifier Unless the amplifier is being installed in its final, fixed location, we strongly recommend supporting the rear of the amplifier. Supporting the rear of the amplifier is required for mobile and portable use. If the amplifier is to be transported in any way, install the optional rear rack mounting ear kit to support the rear of the amplifier. During transport, the shock loads encountered on the chassis and rack can easily damage an unsupported amplifier and the rack rails. With proper support, reliability is enhanced. Optional rear rack mounting ear attachment methods- Rear rack mounting ear kits are an accessory item and are refer to the rear rack mounting ear kit’s documentation available from QSC’s Technical Services Department or from for details your dealer or distributor. Refer to the literature included with the rack mounting ear kit for installation instructions. Two methods of rear rack mounting ear attachment are possible; use the method that best suits you application. 6 INSTALLATION: COOLING & AC MAINS REQUIREMENTS Fan Cooling The CX168 amplifier draws cool air into the rear of the amplifier and exhausts hot air from the front. This is done so the equipment in the rack stays as cool as possible. This method of cooling gives the operator “direct” air tempera- ture feedback at the front of the rack, where it is the most convenient. The front panel’s temperature indicates “how hard” the amplifier is working. The fan varies speed automatically to maintain safe internal temperatures and minimize noise. Keep the front and rear vents clear to allow full air flow. Air flow in QSC amplifiers Do not obstruct the front or rear air vents! Make sure that plenty of cool air can enter the rack, especially if there are other units which exhaust hot air into it. AC Mains Connection The correct AC line voltage is shown on the serial number label. Make sure the AC mains is the correct voltage. Connecting to the wrong line voltage is dangerous and may damage the amplifier. Connect the AC line cord by orienting the IEC plug correctly and pushing the plug firmly into the IEC receptacle. It should seat tightly. It can only be inserted when it is properly oriented. Use the cord supplied with the amplifier, or an equivalent. Insure that the wire gauge of the cord is #14AWG. Use of #16 or #18 AWG can The correct AC line voltage is shown on the be dangerous and is not recommended. serial number label. Use the best possible connection to the AC power source. Avoid extension cords as they will cause some voltage drop between the AC source and your amplifier. If the use of Connecting to the wrong line voltage is an extension cord is required, ensure that it is dangerous and may damage the amplifier or the shortest length possible and is at least constitute the risk of electric shock. Verify the #14 AWG. Ensure that all grounding connec- correct AC line voltage by checking the tions are maintained. specification printed on the serial number label on the rear panel. NOTE: Excessive length or inadequate gauge may result in short muting episodes if all channels are driven to full power. 7 SETUP: SETTING THE MODE CONFIGURATION DIP SWITCHES Setup By using the mode switches and connecting to the amplifier properly, the CX168 amplifier can be configured as a 1 to 8 channel amplifier. This flexibility enables the CX168 to be used for most any multichannel application. The CX168 can be thought of as four separate 2-channel amplifiers on one chassis. Each of the four amplifier sections may be operated in Stereo, Parallel, or Bridge mode independently of the others. For example, channels 1 & 2 could be set for stereo operation while channels 3 & 4 are set for parallel operation; all while channels 5 & 6 and channels 7 & 8 are set for bridge mode. The channels are grouped as follows: Channels 1 and 2 Channels 3 and 4 Channels 5 and 6 Channels 7 and 8 You can set the operating mode for each group by setting that group’s configuration DIP switches in the desired positions. It is NOT POSSIBLE to group the channels differently. For instance, you can not bridge channels 3 & 5 “MODE” configuration DIP switches are located on because they are not in the same channel group. the rear panel and look like the illustration above. Switch positions shown are for example only. Set Before setting the configuration DIP switches, you must first switches as required for your application. decide how the amplifier will connect to the speaker system. “MODE” switch settings and speaker connection diagrams are printed on a rear panel label, as shown, above. 8 SETUP: SETTING THE MODE: STEREO, PARALLEL, OR BRIDGE Modes The following descriptions apply to a channel pair (such as Ch. 1 & 2 or Ch. 7 & 8). It is possible to set each channel pair’s mode differently and customize the system configuration. All possible combinations will not be shown. The following describes the behavior of one channel pair in the three modes of operation and the BRDG and PAR LED’s: Stereo- Stereo mode supports two completely separate audio channels, usually referred to as ‘left’ and ‘right’. Stereo configurations have two separate input signals and two separate output signals. Mode switch settings and LED indication for stereo mode operation. BRDG and PAR indicators should not be illuminated when in stereo mode. Parallel- Parallel mode applies one input signal to both channels. Both inputs of a channel pair are connected in parallel when the mode switch is set for parallel, therefore, connect only one input per channel pair when in parallel mode. The outputs are connected the same as stereo mode. Each speaker will be supplied the same signal, which is still controlled by that channel’s gain control. Mode switch settings and LED indication for parallel mode operation. PAR indicator should be illuminated and BRDG indicator should not be illuminated when in parallel mode. Bridge- Bridge mode combines the two channels of a pair (such as Ch. 3 & Ch. 4) into one higher- powered channel. Like parallel mode, both inputs of a channel pair are connected in parallel when the mode switch is set for parallel, therefore, connect only one input per channel pair when in bridge mode. There are 4 possible channel pairs that can be bridged on the CX168. They are Ch.1-2 Bridge, Ch. 3-4 Bridge, Ch. 5-6 Bridge, and Ch. 7-8 Bridge. Any pair or number of pairs can be independently configured in bridge mode. Use the first channel’s input and gain control when in bridge mode. The Mode switch settings and LED indication for bridge mode operation. BRDG indicator and PAR indicator should be illuminated when in bridge second channel’s gain control is non-operative and mode. should be turned all the way down. Note! Do not connect more than one input per channel pair when operating in parallel or bridge mode. However, in parallel or bridge mode, the unused input terminals may be used for daisy-chaining the input signal to other channels of the CX168 or to other amplifiers. For pictorial description of operating modes, see Appendix. 9 SETUP: MODE SWITCHES: CLIP LIMITER AND LOW FREQUENCY FILTERS Clip Limiter The CX168 amplifier has separate clip limiters for each of the 8 channels. These clip limiters respond only to actual amplifier clipping. Amplifier clipping generates internal error signals which cause the clip limiter to quickly reduce gain and minimize the overdrive. To preserve as much of the program dynamics as possible, limiting occurs only during actual clipping. Each channel’s clip limiter can be switched on or off individually. The clip limiter is internally set to respond as fast as possible after clipping is detected. For program material that is primarily “full-range”, the effect on the overall audio quality should be imperceptible. We recommend using the clip limiters for almost all applications, especially full-range audio applications. Clipping can cause high-frequency artifacts to be output to the speakers, potentially damaging fragile high-frequency drivers. For program material that is primarily low-frequency in nature (low-frequency or subwoofer drive) this may be CAUTION! Clip limiting reduces extreme perceived as a “rubbery” effect on the audio. If this is the overdrive peaks, allowing a higher average case, it may be preferable to turn the clip limiters off and let signal level without distortion. Increasing the the amplifier clip occasionally. With robust, low-frequency gain with the clip limiter engaged until drivers, the occasional clipping should cause no problems. clipping is again audible, can double the average output power. Be careful not to exceed the power rating of the speakers! Low Frequency Filter When driving speakers with limited low frequency response, it is important to limit the low frequency response of the amplifier. Doing so can result in more usable bass response since the speaker is not being overloaded by very low frequencies it can’t handle. Explanation- Low frequency sound waves require much more speaker cone motion to produce the same apparent loudness as higher frequencies. Properly designed speaker enclosures help the speaker to move more air with less motion using techniques like porting. Such enclosures only benefit from porting down to a certain frequency. Below this frequency, the speaker is “unloaded” and is basically free to move around uncontrollably without producing much bass. Limiting the frequency range of the low frequency content enables the speaker to behave the way it was designed to. If frequencies lower than designed are supplied to the speaker, performance will degrade. All low frequency filters will change the character of low frequency transients. For best results, the cabinet design, speaker capabilities, and program material must be taken into account when configuring low frequency filtering. 10 SETUP: MODE SWITCHES: CLIP LIMITER AND LOW FREQUENCY FILTERS Low Frequency Filtering Tips: • The OFF position should be used only for subwoofer systems with rated frequency response below 33 Hz. or if low frequency filtering is provided by other devices. • Know the specification of the speaker cabinet you are driving. Match the low frequency roll off setting to the specified low frequency capability of the speaker cabinet. Do not drive the speaker with frequencies below its rating. • Unless you have low frequency filtering before the amplifier, use the low frequency filter to protect your loudspeakers from cone over-excursion caused by frequencies below the speaker’s limits. • The 33 Hz. rolloff is a good “all purpose” setting. Turn the filter on by setting the appropriate DIP switch to ON position and select 33 Hz. using its 33/70 Hz. DIP switch. This setting is a good starting point for most large, full-range cabinets. • The 70 Hz. rolloff is a good setting to use with smaller, compact speaker cabinets having limited bass capability. Turn the filter ON and select the 70 Hz. setting when using smaller cabinets. • The frequency selection ( 33 or 70 Hertz) has no effect unless the filter is set to the ON position. Frequency Response Curves Filter off Filter ON, 70 Hz. Filter ON, 33 Hz. 11 CONNECTIONS: INPUTS- Using the Terminal Block Inputs Each channel has an active balanced "Euro-style" terminal block input jack. These terminal blocks allow the input wiring to be terminated using simple hand tools and allows for quick reconfiguration when needed. The input impedance is 20k ohm balanced or 10k ohm unbalanced. Balanced connection is recommended for all audio inputs. Balanced signals are less prone to AC hum and other electrical noise. Unbalanced signals can be suitable for short cable runs. The signal source's output impedance should be less than 600 ohms to avoid high frequency loss in long cables. If the DataPort is being used for the input signal source, the terminal block connections should not be used for inputs. However, they may be used for daisy chaining the DataPort input signal to other channels or amplifiers. The signal available from the terminal block input connections will be about 10 dB lower than the signal presented to the DataPort. TERMINAL BLOCK CONNECTORS Terminal Block Connectors: Balanced inputs: Connect the conductors to the connector as shown. shield Terminal block: balanced connections Unbalanced inputs: Connect the conductors to the connector as shown. Make sure that the unused side of the balanced input is connected to ground, as shown. jumper shield Terminal block: unbalanced connections 12 CONNECTIONS: INPUTS- Using the DataPorts If used, the DataPort must be connected to a QSC DataPort product using a QSC DataPort cable. Do not use computer cables; they look similar but can cause operational problems with your amplifier and/or DATAPORT CONNECTORS DataPort accessories. The CX168 amplifier is equipped with four QSC DataPort connections that may be used for connecting to accessory QSC DataPort devices. Each DataPort services a specific channel pair; the channels are paired as: Ch. 1 and Ch. 2 Ch. 3 and Ch. 4 Ch. 5 and Ch. 6 Ch. 7 and Ch. 8 Each DataPort and its associated terminal block input pair can be configured independently (there is no requirement to use a DataPort connection for all channels). Only the Ch.1-Ch.2 DataPort can control the power supply standby mode. The DataPort connection will supply the input signals to the amplifier. If using the DataPort connection for audio input signals, do not apply inputs to the terminal block input connectors. If the DataPort connection is used only for monitoring amplifier operating conditions and NOT for providing audio inputs to the amplifier, then the terminal block inputs may be used to supply the audio inputs to the amplifier. How to Connect to the DataPort: Direct mounting of DataPort accessories is NOT supported by the CX168 amplifier. Devices that are normally mounted directly to the amplifier must be mounted remotely and connected to the DataPort using a DataPort cable. This is due to the high density of connectors on an eight-channel amplifier; there is not enough space to allow for accessories to be mounted directly to the back of the amplifier. If the accessory attaches with a QSC DataPort cable, orient the HD-15 male plug correctly with the DataPort socket on the amplifier (it is “D” shaped and will fit only one way). Push the plug onto the socket firmly and ensure it is seated properly. Finger-tighten the 2 retaining screws. Do not over tighten. Basic DataPort Notes: 1- The amplifier Gain controls will need to be set at their anticipated high-level setting. Use reduced level setting during setup & test. 2- If using the DataPort for audio input signals, control of the audio level will be accomplished with the DataPort accessory device to which the amp is connected. 3- Control of the amplifier’s AC power standby mode will only be possible via the Ch.1-2 DataPort. The amplifier’s POWER switch must be physically set to the “ON” position to use the standby control feature. 4- The order in which the four DataPorts are connected to external devices makes no difference to the amplifier. However, the host controller to which the amplifier is connected will show the channels in the order that cables are connected. 13 CONNECTIONS: OUTPUTS Outputs Refer to the label on the rear panel of the amplifier for proper wiring connections. The output connection for STEREO and PARALLEL modes is on the left side of the label , while the output connection for BRIDGE mode is shown to the right side of the label. Note: Polarity changes from channel to channel and is different for bridge mode. Be certain of polarity of connec- tions before applying power. Reversed polarity may degrade audio frequency response. Speaker connection diagrams and “MODE” switch settings are printed on a rear panel label, as shown, above. Stereo and Parallel Mode In stereo or parallel mode, each speaker is connected to its own individual channel of the amplifier. This connection method is shown on the left side of the connection diagram (rear panel) or in the diagram, right. Use 4 ohm minimum impedance in stereo or parallel mode. Ensure the mode configuration switches are set for stereo or parallel mode when connecting speakers to each channel’s output. Stereo or parallel connection- single speaker shown con- nected to amplifier channel 6 output. Ensure that all speaker Bridge Mode connections maintain proper polarity ( + to +, - to - ). In bridge mode, each speaker is connected to a bridged-pair of outputs. The channels must first be “bridged” by setting the mode configuration switches to the bridge settings. Then con- nect the speakers as shown on the right side of the connection diagram (rear panel) or in the diagram, right. Use 8 ohm minimum impedance in bridged mode. Ensure the mode configuration switches are set for bridge mode when connecting speakers to bridged output pairs . BRIDGE MODE PRECAUTIONS: Do not use 2 ohm or 4 ohm loads in bridge mode! 8 ohms is the minimum impedance for bridge mode Bridge connection- single speaker shown connected to am- operation. This mode puts a high demand on the plifier channels 5 and 6 (bridged with mode switch settings). amplifier and speaker. Excessive clipping may cause Ensure that all speaker connections maintain proper polar- protective muting or speaker damage. Ensure the speaker has ity ( + to +, - to - ). a sufficient power rating. 14 CONNECTIONS: OUTPUTS- USING THE TERMINAL BLOCK CONNECTORS OUTPUT TERMINAL SAFETY WARNING! Do not touch output terminals while amplifier power is on. Make all connections with amplifier turned off. Risk of hazardous energy! When selecting speaker cable (wire), always use the largest wire size and shortest length of wire practical for an installation. Larger wire sizes and shorter lengths minimize power loss and degradation of damping factor. Do not place speaker cables next to input wiring. Terminal Block Connections The output connections are made by attaching speaker wires to the 8-pin terminal block connectors. Once these connectors have been “populated” with the necessary speaker wire connections, plug the terminal blocks into their respective output jack. The output jacks are configured in two discrete connectors. One connector carries the connections for channels 1 through 4. The other, channels 5 through 8. Each connector has 8 pins, each of which connects to the speaker wire using a screw-clamp. Refer to the diagram, right, for the basic connection procedure. Be sure that speaker polarity is maintained. The locations of the “+” and “-” terminals on the amplifier is not the same for each channel. Ensure that all wires are neatly terminated with no loose strands. Do not strip the insulation back excessively. Loose strands and exposed wire beyond the terminals may cause a short circuit and cause protective muting of the amplifier. Once all the required connections have been made to the terminal block connectors, they may be inserted into the OUTPUTS jack on the rear panel. Orient the connector properly (it only fits one way) and push it in until it is fully seated in the receptacle. Note each connector is oriented so its wires face outwards. How to make connections to the terminal block: 1- Strip the wires approximately 0.28” (7mm). 2- Twist any loose strands of wire together. 3- Loosen the appropriate retaining screw fully. 4- Insert the stripped wire into the receptacle. 5- Tighten the appropriate retaining screw to secure. 15 OPERATION: POWER SWITCH and GAIN CONTROLS Power Switch The power switch is a rocker-type switch. It is located on the left side of the front panel. To turn the amplifier on, push in on the top of the switch. To turn the amplifier off, push in on the bottom of the switch. The green power LED should light up when the switch is in the on position. When the power is switched off, the LED may takes several To turn the amplifier ‘on’, push in on the top por- seconds to go out; this is normal. tion of the rocker switch firmly. It should move inward and snap into the ‘on’ position. After switching on, the power LED should illuminate. To turn the amplifier ‘off’, push in on the bottom portion of the rocker switch firmly. It should move inward and snap into the ‘off’ position. After switching off, the power LED should extinguish. If the POWER LED fails to illuminate when the switch is in the ON position: 1- Check the AC cord and insure that both ends are fully inserted into their receptacles. If the power LED still fails to illuminate, 2- Check the AC outlet for voltage with a circuit tester or known good device (lamp, etc.). 3- Make sure that the amplifier is not in STANDBY mode. If the amplifier is being controlled by one of its DataPort connectors, the amplifier may be in standby mode. Verify the AC power status from the DataPort device controlling the amplifier. The ‘POWER’ LED is located at the upper-left of the gain controls section of the front panel. It is a green LED that illuminates when the power is turned on (and the amplifier is properly connected to the correct AC source). The power LED extin- guishes when the power is turned off. 16 OPERATION: LED INDICATORS LED Indicators The LED indicators provide basic operation information to the operator. The ‘nor- mal’ indications of the LED’s are shown, below. NOTE! The BRDG and PAR mode indicator LED’s are discussed on page 9. POWER: The power ‘on’ indicator LED is located at the If no indication: Check AC power cord, AC mains, and upper-left of the gain control cluster. It is green and that the DataPort is not being used to force the illuminates when the power switch is set to the on amplifier into ‘Standby’ mode. position and AC power is present at the IEC cord receptacle. When power is switched off: the LED may take several seconds to extinguish (go out); this is normal. Normal indication: at power-on the LED will illumi- nate. SIGNAL: Each channel has a green ‘signal’ LED If no indication: check gain settings, input cables, located to the right of its gain control. The ‘signal’ LED connections and audio source. If audio source is lights up when the input signal is sufficiently strong. extremely low signal strength, signal LED may not illuminate; this is normal but indicates that the input Normal indication: illuminates when the input signal signal strength should be increased. is present and sufficiently strong. Occasional illumination: normal for weak input signal strength. Fully illuminated (on): normal. CLIP : Each channel has a red ‘clip’ LED that illumi- If no indication: normal if the amplifier is operated at nates when the channel clips. It is located to the upper- nominal output levels. If the input signal is weak, it may right of each channel’s gain control. not be capable of driving the amplifier into clipping, even at full gain. Normal indication: illuminates briefly at extreme output power peaks. Occasional clipping at high power Occasional illumination: Occasional clipping (once levels may be normal. At power-on, the clip LEDs may briefly every few seconds) when operation at high briefly flash during turn-on muting. power levels does not indicate trouble. Continuous operation at high power may trigger the thermal Continuous operation at high power may trigger thermal protection circuitry, shutting down the amplifier and protection circuitry, shutting down the channel pair fully illuminating the clip LED. concerned. The CLIP LEDs for the channel pair will fully illuminate to indicate thermal protection activation. The Illuminated most of the time: Not normal; for cleaner channels are paired as: sound, reduce the output of the amplifier and/or input Ch. 1 and Ch. 2 signal to avoid thermal shutdown of the amplifier and Ch. 3 and Ch. 4 possible speaker damage. If clipping persists at normal Ch. 5 and Ch. 6 levels, check for shorted or abnormal load impedances. Ch. 7 and Ch. 8 17 OPERATION: GAIN CONTROLS Gain Controls The gain controls are located on the right side of the front panel. They are operated by rotating the control clockwise to increase gain or counterclockwise to decrease gain. When turned fully counterclockwise, gain is fully reduced and no output signal will be present. When turned fully clockwise, the voltage gain of the amplifier is +26dB and the output signal will potentially be full strength (provided the input is fully driven). The gain controls are detented in 21 steps for repeatable adjustment. Surrounding the Gain control, the power Location of gain controls on the front panel of the CX168 attenuation level is shown in dB. Each detent is approxi- mately 1dB change in amplifier gain. To operate the gain control, rotate the control’s knob so that the desired level is achieved. A small, flat-tipped screw- driver (#1 blade size) is typically used. Be sure not to use large screwdrivers as it is possible to damage the gain control by applying excessive torque. Gain controls are sensitive, electronic components and should be treated with care. Each channel pair’s gain controls and indicator LED’s are grouped together as shown, right. The channels are paired in the following manner: Channel Pair #1- Ch. 1 and Ch. 2 Channel Pair #2- Ch. 3 and Ch. 4 Channel Pair #3- Ch. 5 and Ch. 6 Channel Pair #4- Ch. 7 and Ch. 8 As detailed in the ‘Setup’ section of this manual, channels can only be paralleled and bridged within their channel pair. Typical channel-pair control and indicator group on the CX168 Example- Channel 3 can be bridged with channel 4 only; it can not be bridged with any other channel. The dB markings around the gain controls are attenuation. They show attenuation from full gain. At 0db attenuation, the voltage gain of the amplifier is 26dB. If the gain control is set at the 7dB attenuation position, then the voltage gain of the amplifier is 19dB (26dB - 7 dB). Typical gain control showing attenuation level markings and LED’s for the channel. Note that there is only one BRDG (bridge) and one PAR (parallel) indicator per channel pair. 18 OPERATION: GAIN CONTROL SECURITY PLATE Gain Control Security Plate After final gain adjustments have been made, the gain controls can be covered with a security plate. This easy-to-use security plate makes it difficult for others to adjust the gain. This may be desirable in instances where the gain controls should not be tampered with. To install the security plate: 3. Line up the tabs with the keyed portion of the ventilation slots, insert into the slot and then slide the panel fully right, locking it in the slot. 1. Use a 9/64” or 3.5mm hex key to loosen the screw. Do not remove; just loosen sev- eral turns. 2. Orient the panel as shown and slide the right end 4. Make sure the LED’s are visible, then tighten the of the plate under the screw. screw with the hex key. Do not over tighten! When properly installed, the security panel will still allow you to monitor all the channel status LED’s and power LED. The amplifier front panel will look like the illustration, right, after installing the security plate. 19 OPERATION: USING THE QSC DATAPORT CONNECTIONS DataPort Connectors QSC’s DataPort equipped amplifiers offer operators of larger audio systems a high degree of system monitoring and operation from remote locations. A system operator in a sports complex might be located hundreds of yards (meters) from the actual amplifiers, making ‘normal’ control and monitoring of the system difficult, at best. By integrating QSC’s DataPort-equipped amplifiers with QSC’s DataPort accessories, a dependable and robust control and monitoring solution can easily be implemented. The DataPort is used for connection to optional QSC DataPort accessories. DataPort accessory devices can provide remote cinema monitoring, DSP, filter and crossover functions. A single DataPort is assigned to each channel pair. Each DataPort is marked with the channel numbers served by that DataPort. Note, The DataPort connections on the CX168 offer full-featured above, that each DataPort is clearly labeled and is located be- implementation of DataPort accessories and supply the tween its respective channel’s terminal block connectors. required accessory voltage to DataPort devices that require it. CX168 amplifiers DO NOT support direct mounting of smaller accessories, like the DSP-3, due to the high density of connectors on the rear panel but they may be connected via cables. Using the DataPort Use a QSC DataPort cable to connect the CX168 DataPorts to DataPort accessories. Consult the accessory’s documentation for recommended mounting and interconnect information. Do not use terminal block connections for inputs when the DataPort is used as the input signal source. If you do, the signals from the DataPort and terminal block inputs will be mixed, possibly with unexpected results. See the following section for general operating notes regarding the DataPorts on the CX168. Connect to the DataPort using a QSC DataPort cable. Align the connector with the socket housing and insert fully. Finger-tighten the retaining screws. 20 OPERATION: USING THE QSC DATAPORT CONNECTIONS What Can I Connect to the DataPort? Consult your QSC representative for up-to-date accessory compatibility. QSC has several DataPort-based accessories available like the: CM16a Amplifier Controller & Monitor DSP-3 Digital Signal Processing Module DSP-4 Digital Signal Processor Module XC-3 Crossover SF-3 Subwoofer Filter LF-3 Low Frequency Filter DCM Crossovers & Monitors General Guidelines to using the DataPorts on the CX168 1. Each channel pair ( Ch.1-2, Ch.3-4, Ch.5-6, and Ch.7-8) has its own dedicated DataPort connector that applies control and monitoring functions to its specific channel pair. 2. The DataPort for channels 1 and 2 has a special function on the CX168. It is the ‘master’ DataPort for controlling AND reporting power supply status of the amplifier. It is only possible to control and report the status of the power supply using the Ch.1 & Ch.2 DataPort. 3. In order for the Ch.1 & Ch.2 DataPort to control the power status of the amplifier, the front-panel power switch MUST be in the ON position. If the power switch is in the ‘off’ position, the amplifier will not respond to any com- mands sent to its Ch.1 & Ch.2 DataPort. 4. The DataPort connection can be used for amplifier status monitoring only, using the terminal block inputs for supplying input signals to the amplifier. Ensure that the DataPort connection supplies NO INPUT SIGNAL when operating in this fashion. If signal is applied to both the DataPort and the terminal block input connector at the same time, the signals will be summed and amplified. The result may be undesirable. 5. If using the DataPort for supplying the audio input signals to the amplifier, we don’t recommend using the terminal block input connectors, even if they are connected to non-operating audio sources. The DataPort and terminal block inputs are connected in parallel and isolated by about 20k ohms of resistance. Even with this isolation, it is possible that connection to other signal sources may significantly alter the input impedance at the DataPort. This could affect output levels as connections are altered. 6. If using the DataPort for supplying the audio input signals to the amplifier, do not use the unused terminal block inputs for daisy-chaining the DataPort signals to other channel pairs or other amplifiers. The DataPort device supplying the audio signal will not accurately report gain and levels will be effected. 7. The heat sink temperatures of the CX168 are reported on the first channel of each channel pair’s DataPort. In other words, to monitor all four heat sink temperatures, all four DataPorts must be connected to the monitoring device (such as QSC’s CM16a). Temperatures will be reported on channels 1, 3, 5 and 7. 21 APPLICATIONS: TYPICAL APPLICATION EXAMPLES Four Room Stereo Feed A simple application of the CX168 would be an installation with 4 rooms of coverage; each room to be provided with a stereo audio feed. All speakers are 4 ohm or greater impedance and have a frequency response rated down to 45 Hz., making them suitable for the 70 Hz. low frequency setting. Clip limiting will be used with these full-range speakers. Operating mode of all four channel pairs will be set to ‘stereo’ because the audio inputs are stereo and a stereo result is desired in each room. Input connections can vary too much to provide meaningful examples. For this example, assume that each input has its own discrete signal provided at the terminal block input connectors. Routing would make sense if kept in room number order: Room #1: Fed by channel pair #1 (Ch.1&2) Room #2: Fed by channel pair #2 (Ch.3&4) Room #3: Fed by channel pair #3 (Ch.5&6) Room #4: Fed by channel pair #4 (Ch.7&8) MODE switch configuration: Ch.1/ Ch.2 mode switch: Clip Limiters ON Low Frequency Filter ON Low Frequency Setting 70 Hz. Mode: stereo Ch.3/Ch.4 mode switch: same as first mode switch Ch.5/Ch.6 mode switch: same as first mode switch Ch.7/Ch.8 mode switch: same as first mode switch Tri-Amp’d Stereo Cabinets with Subwoofer Driven by Bridged Channel Pair (next page) Two discrete 3-way (tri-amp) cabinets are to be connected to the CX168 along with an 8 ohm subwoofer cabinet. The subwoofer will require the most amount of power and therefore will be driven by two channels bridged into one. Its impedance meets the minimum requirement for bridged output. All other drivers are 4 ohms or greater and meet the single channel minimum impedance requirement. Each of the tri-amp cabinets will be driven with three discrete channels (totaling 6 channels). The total required channel count is 8; the CX168 is the perfect tool for the job. 22 APPLICATIONS: TYPICAL APPLICATION EXAMPLES Tri-Amp’d Stereo Cabinets with Subwoofer Driven by Bridged Channel Pair (continued) Routing is a function of ‘easiest wiring’ order and is: LF Driver Left: Ch. 4 Mid Driver Left Ch. 3 HF Driver Left Ch. 2 HF Driver Right Ch. 1 Mid Driver Right Ch. 5 LF Driver Right Ch. 6 Subwoofer Bridge Ch. 7 & 8 MODE switch configuration: Ch.1/Ch.2 Mode: Stereo Ch.1 Clip Limiter: ON Ch.1 LF Filter: OFF Ch. 2 Clip Limiter ON Ch. 2 LF filter OFF Ch.3/Ch.4 Mode: Stereo Ch. 3 Clip Limiter ON Ch.3 LF Filter: OFF Ch. 4 Clip Limiter OFF Ch. 4 LF filter ON Ch. 4 LF frequency 70 Hertz Ch. 5/Ch. 6 Mode: Stereo Ch. 5 Clip Limiter ON Ch.5 LF Filter: OFF Ch. 6 Clip Limiter OFF Ch. 6 LF filter ON Ch. 6 LF frequency 70 Hertz Ch. 7/Ch. 8 Mode Bridged Ch. 7 Clip Limiter OFF Ch. 7 LF Filter OFF Ch. 8 Clip Limiter no effect in bridge Ch. 8 LF Filter no effect in bridge Channels 1 through 6 are configured for ‘stereo’ mode because all channels are completely separate from one another. Channels 7 and 8 are bridged to combine their power to drive the subwoofer (sub’s are the most inefficient driver and therefore require the largest portion of the system power). Clip limiters are used on the mid range drivers and the HF drivers because they are the most fragile and are damaged easily by clipping events. The LF drivers and the sub’s are rugged enough to handle even prolonged clipping events without damage and thus no clip limiting is used. Low frequency filtering is used only on the LF drivers. The HF and mid range drivers have filtering before the amplifier and need no filtering from the CX168. The subwoofer cabinet can handle the low frequencies without problem, so no filtering for them. The LF drivers response specification will dictate at what frequency (33 or 70 Hz.) the low frequency filter is set for. 23 APPENDIX: DESCRIPTION OF STEREO & PARALLEL OPERATING MODES NOTE! In the following diagrams, channels 1 and 2 are shown as examples only. Any channel pair can be configured as shown here. The channels pairs on the CX168 are : Channels 1 and 2, Channels 3 and 4, Channels 5 and 6, and Channels 7 and 8. Therefore, in any example below, replace channel 1 and 2 with the channel pair numbers you are setting up. Stereo Mode: Each input signal is sent to its respective channel. Each channel has independent low-frequency (subaudio) filtering, clip limiting, gain control, and output connection. When to use STEREO input configuration: Use stereo mode for stereo sources (L-R inputs) and any other situation that requires each channel to be completely separate from the other. How to use STEREO mode: 1- Set the MODE switches for stereo mode operation. STEREO MODE OPERATION 2- Connect the two input signals to CH1 and to CH2 (or NOTE: Ensure that the MODE SWITCHES are set signals may be provided from DataPort connection). to STEREO when feeding two separate signals to the two channels. 3- Connect the two speakers; one to channel 1’s output terminals, one to channel 2’s output terminals. Parallel Mode: The channel 1 and channel 2 inputs are connected together, applying a single input signal to both channels of the amplifier. A signal into any input jack will drive both channels. Each channel's low frequency filtering, clip limiting, and gain control still function independently. Each channel drives its own speaker load. You can patch the input signal on to additional amplifiers (daisy chain) by using any of the remaining input jacks. This feature eliminates the need for “Y” cables. When to use PARALLEL input configuration: Use parallel mode when you need one signal to drive both channels; each channel having its own control (gain, clip limiter, low- frequency filter). How to use PARALLEL mode: 1- Set the MODE switches for PARALLEL mode. PARALLEL MODE OPERATION 2- Connect the one input signal to either channel’s input connector. The unused input may be used for daisy chaining the signal to another device or channel pair. 3- Connect the two speakers: one to channel 1’s output terminals, one to channel 2’s output terminals. 24 APPENDIX: DESCRIPTION OF BRIDGE OPERATING MODE Bridge Mode Bridge mode combines both output channels into one output. This mode is for driving a single, high-power-rated load with twice the “normal” voltage swing. This results in about 4 times the peak power and about three times the sustained power of a single channel. It is also common to call this bridge mono mode. When to use BRIDGE mode: Use BRIDGE mode when you need to deliver the power of two channels to a single 8 ohm load, such as a subwoofer. Do not use less than 8 ohm loads in bridge mode. Refer to the CONNECTIONS section of this manual for details. How to use BRIDGE mode: BRIDGE MODE OPERATION- 1- Set the MODE switches for bridge mode. Note that speaker connection for bridge mono mode is 2- Connect the one input signal to channel 1 or channel 2’s different than other modes. See section on Connections: input. Outputs for proper bridge mode output connections. 3- Connect the one speaker load to the bridge output terminals 4- For a bridged pair, use the first channel’s gain control, clip limiter and filter. The second channel’s controls will have no effect. 5- Turn the second channel’s gain to zero gain. Filters and limiter settings of second channel do not matter. Normal Operating Levels The amplifier’s protective muting system guards against excessive internal temperatures. With normal ventilation and loads, the amplifier will handle any signal level, including overdrive. Lower load impedances and higher signal levels produce more internal heating. Loads below 4 ohm loads may trigger thermal shutdown. A Note on Sound Pressure Levels Sound is heard as the ear converts vibration from sound waves into impulses in the nerves of the ear. Sounds above 90 decibels, particu- larly if the sound is prolonged, may cause such intense vibration that the inner ear is damaged. 90 decibels is about the loudness of a large truck about 5 yards away. A jackhammer emits sounds of about 120 dB from 3 feet away, and a jet engine emits sound of about 130 dB from 100 feet away. Motorcycles, snowmobiles, and similar engines range around 85 to 90 dB, and a rock concert may approach 100 dB. A general rule of thumb is that if you need to shout to be heard, the sound is in the range that can damage hearing. Some jobs in the entertainment industry involving loud noise from music carry high risk for hearing loss. In the U.S., the maximum job noise exposure is regulated by law. Both the length of exposure and the extent (decibel level) of exposure are considered. If exposure is at or greater than the maximum exposure, protective mea- sures must be taken. The table at right is referenced from OSHA’s Permissible Noise Exposure table G16. 25 APPENDIX: MULTIPLE SPEAKER LOADS IN SERIES Multiple Lo-Z Loads in Series: Series connection is where the same signal current flows speaker cabinets are NOT wired for series connection ability; through each of the speaker loads. The signal goes into one consult your speaker’s documentation for connection information. terminal of the first speaker and out its other terminal; then the signal goes into one terminal of the next load, and so on. If any It is not recommended to connect speakers of different imped- of the speakers in a series-connected branch of a system fails, ances in series because power will be divided unequally the signal is prevented from flowing through any of the loads (no between them. If you are experienced with mixing and matching sound). Total impedance for series connections is the sum of speaker loads of varying impedances, no damage to the amplifier each of the impedances in the chain (see illustration, below). will result as long as the total impedance is within 4 to 16 ohms per stereo/parallel channel or 8 to 16 ohms, bridged. Series connection schemes are usually used with “raw” speakers (not often with speaker cabinets) for wiring conve- Below are examples of series connections for 2, 4, 8, and 16 ohm nience. The exposed speaker connections make this method speaker loads. Series connections resulting in an impedance of easy for small “strings” of speakers in one area. Most PA-type 250 ohms or greater are not recommended. 26 APPENDIX: MULTIPLE SPEAKER LOADS IN PARALLEL Multiple Lo-Z Loads in Parallel: If each speaker load is connected across the same signal source Most PA-type speaker cabinets are provided with two or more (the output of the amplifier), then the loads are in parallel with input jacks that are wired in parallel. If one cabinet is connected one another. If one of the speakers fails in a parallel-connected to the amplifier, the load impedance is equal to the one cabinet’s system branch, the remaining speakers will continue to operate. specified impedance. If you attach another similar cabinet to the If all speaker loads are the same impedance, the total imped- remaining jack of the first cabinet, the two cabinets are now in ance of a parallel-connected system branch is the impedance of parallel with each other and the impedance will be half of the one speaker load divided by the number of loads in the branch one-cabinet value. Consult your speaker’s documentation for (example: three 8-ohm loads in parallel: 8/3=2.667 or roughly 2.7 connection details as there are many connection methods in use. ohms, too low to use on the CX168). Below are examples of parallel connections of 4, 8 and 16 ohm It is not recommended to connect speakers of different imped- speakers. 2 and 4 ohm parallel loads are not usable because the ances in parallel because power will be divided unequally parallel impedance of 4 ohm loads will always be 2 ohms or less; between them. If you are experienced with mixing and matching this is too low an impedance for the CX168 amplifier. speaker loads of varying impedances, no damage to the amplifier will result as long as the total impedance is within 4 to 16 ohms per channel, dependant upon operating mode. 27 TROUBLESHOOTING: NO SOUND Problem: NO SOUND • INDICATION: POWER LED INDICATOR NOT ILLUMINATED • Check both ends of the AC line cord for proper connection. Both ends should be fully seated in their connectors. • Check that the AC outlet works by plugging in a known-good device or by testing with a voltmeter. If too many amplifiers are used on one outlet, the building's circuit breaker may trip and shut off power. If this is the case, unplug some of the amplifiers from the outlet, reset the building’s breaker and try again. For the other amps, use an outlet that is connected to another circuit breaker to distribute the load current. • If using the DataPort connections, the channel 1 and 2 DataPort controls the power supply of the amplifier. Check whether or not the DataPort accessory connected to the CX168 is forcing the power supply into standby mode. • An amplifier which keeps tripping the AC circuit breaker may have a serious internal fault. Turn it off, remove AC power, and have the amplifier serviced by a qualified technician. • INDICATION: SIGNAL LED’S NOT RESPONDING TO INPUT SIGNAL LEVEL (NOT LIGHTING) • If the green power indicator LED is lit, but the signal LEDs do not light up at all, check the input. Make sure the signal source is working and try another input cable. Connect the source to another channel or amplifier to check that it is working. • If you are using the terminal block inputs, be sure the connectors are fully inserted at both ends of all interconnecting cables. If you are using the DataPort for your input signal, ensure the DataPort cable is properly connected at both ends and that the device supplying the audio signal to the DataPort is properly configured. • Check gain controls. • INDICATION: SIGNAL LEDS RESPONDING TO SIGNAL LEVEL • If the green signal LED indicators are lighting normally, the fault is somewhere between the amplifier and the speakers. Check the speaker wiring for breaks. Check for proper connection at each end. Try another speaker and cable, if necessary. If your system has multiple speakers connected in series, any one speaker failure will cause all other speakers (in the same series circuit) to lose their signal. • INDICATION: CLIP LED FLASHING • If the red clip indicator flashes when signal is applied, the amplifier output may be shorted. Check that the speaker cable connec- tions are tight and reliable at both ends. Check the speaker cables for stray strands or breaks in the insulation. Trim off all loose strands of wire and insulate any exposed wire with electrical tape. All connections should be made with great care and all wire terminations should be carefully checked to ensure quality workmanship. • If the red clip indicator flashes when connecting the speaker cable between the + and - terminals, then that speaker cable (or speaker) may be shorted. Check the cable and the speaker. • INDICATION: CLIP LEDS BRIGHT AND STEADY The amplifier is in protective muting. • Three seconds or so of muting is normal when the amp is turned on. After power-up muting, the Clip LED‘s should go out. • Overheating will cause protective muting and the Clip LED to fully illuminate. If the amp is overheated, the fan will be running at full speed and the chassis will be very warm to the touch; sound should resume within a minute as the amplifier cools to a safe operating temperature. DO NOT TURN THE AMP OFF! This would stop the fan from cooling the inside of the amplifier. Instead, reduce the input signal ( or amp gain setting) and wait for the amp to resume operation automatically. • Check for proper ventilation. If the fan isn't running at all, the amplifier requires servicing. 28 TROUBLESHOOTING: DISTORTION, NO CHANNEL SEPARATION, HUM, HISS, FEEDBACK PROBLEM: DISTORTED SOUND • INDICATION: CLIP LED FLASHING • If the red clip indicator flashes before the signal indicator does, the load impedance is abnormally low or shorted. Unplug each speaker one-by-one at the amplifier. If the clip LED goes out when you disconnect a cable, then that cable or speaker is shorted. Try another cable and speaker to locate and/or remove the fault. • INDICATION: CLIP INDICATOR NOT FLASHING • This could be caused by a faulty speaker or loose connection. Check the wiring and try another speaker. • The signal source may be clipping. Keep the amplifier gain controls at least halfway up (-10 dB or less) so that the source does not have to be overdriven. PROBLEM: NO CHANNEL SEPARATION • Check the mode switch settings on the back of the amplifier. If the mode switches are set for BRIDGE or PARALLEL mode, there will be no channel separation. The mode switches must be set for STEREO for each channel to operate independently. • Make sure other equipment in the signal path to the amplifier, such as mixers, preamps, etc., is set for stereo, not mono. PROBLEM: HISS • Unplug the input cables to the amplifier. If the hiss goes away, then the problem is with the equipment or cables leading to the amplifier. • If the hiss is present with no audio input cables connected, check that the AC line cord is properly grounded at its connection to the line. If the ground connection is OK and the hiss continues with no input cables connected, then the amplifier requires servicing. • To keep the normal noise floor low, operate the primary signal source at full level, without clipping, and avoid boosting the signal further between the source and the amplifier. PROBLEM: SQUEALS AND FEEDBACK • Microphone feedback should be controlled with mixer controls. If noise continues to build up with zero mic gain, there is a serious fault in the signal processors or cables. Working in succession from the signal source towards the amplifier, check each device in the signal path by reducing its gain or unplugging it. PROBLEM: HUM • Use a common AC ground connection for all audio equipment. • Check the input cables for broken ground (shield) connection. • Use balanced connections for the entire signal chain. • The design of the amplifier eliminates internal hum fields, but external transformers or other magnetic devices may cause hum. Move cabling and signal sources to identify "hot spots" in the Magnetic field from power supplies in equipment can system; then avoid those spots. Cables with faulty shielding are a induce hum into cabling that is located in the field. If common entry point for hum. Use top quality cabling. Another common source of magnetic fields are “wall warts” or pluggable hum is a problem, try relocating cabling so that is away transformers; keep input wiring away from them. from power supplies, transformers and other magnetic field producing devices. 29 SPECIFICATIONS CX168 OUTPUT POWER in watts FTC: 8 ohms per channel (20 Hz.-20 kHz., 0.05% THD) 90 4 ohms per channel (20 Hz.-20 kHz., 0.1% THD) 130 EIA: 8 ohms (1 kHz, 0.1% THD, 1 channel driven) 120 (100 watts, all channels driven) 4 ohms (1 kHz, 0.1% THD, 1 channel driven) 180 (140 watts, all channels driven) Bridged Mono: 16 ohms, 20 Hz.-20 kHz., 0.1% THD 180 8 ohms, 20 Hz.-20 kHz., 0.1% THD 260 DYNAMIC HEADROOM 2 dB at 4 ohms DISTORTION SMPTE-IM Less than 0.02% Typical 20 Hz.-20 kHz., 10 dB below rated power Less than 0.05% Typical 1 kHz. and below, full rated power Less than 0.02% FREQUENCY RESPONSE +0.0, -3.0 dB: 8 Hz. to 50 kHz. ±0.2 dB: 20 Hz.-20 kHz. DAMPING FACTOR >200 for 8 ohm load (5 kHz. and below) SIGNAL to NOISE (unweighted, 20 Hz.-20 kHz.) -107 dB VOLTAGE GAIN (gain control set to 0 dB attenuation) 26 dB (20 X) INPUT SENSITIVITY for rated power into 8 ohms 1.35 Vrms INPUT CLIPPING, Vrms 10 Vrms (+22 dBu) INPUT IMPEDANCE 10k ohms unbalanced , 20k ohms balanced 30 SPECIFICATIONS CX168 AMPLIFIER PROTECTION Full short circuit, open circuit, thermal, ultrasonic and RF protection. Stable into reactive or mismatched loads. COOLING Continuously variable speed fan; back-to-front air flow through tunnel heat sink INDICATORS Front Panel: POWER LED (green), 8 each (1 per channel) SIGNAL (green) and CLIP (red) LEDs 4 each (1 per channel pair) BRDG LED (yellow), 4 each (1 per channel pair) PAR LED (orange) CONNECTORS Input: (8) 3-pin terminal block (“euro” or “Phoenix”) Output: (2) 8-pin detachable terminal block connectors Control & Monitoring: (4) QSC DataPort connectors LOAD PROTECTION Turn-on/turnoff muting AC coupling (DC fault blocking) Clip limiting OUTPUT CIRCUIT TYPE AB + B POWER REQUIREMENTS SEE SERIAL NUMBER LABEL ON REAR PANEL FOR THE SPECIFIED OPERATING VOLTAGE Configured at factory for either 100, 120 or 230 VAC, 50- 60 Hz. CURRENT CONSUMPTION @ 120 VAC, typical (all channels driven) in Amperes Idle 0.6 1/8 power, 8 ohms* 6.2 1/8 power, 4 ohms* 9.2 1/3 power, 8 ohms* 9.2 1/3 power, 4 ohms* 14.2 NOTE: 1/8 power is representative of current draw with typical music program material with occasional clipping. *Pink noise 1/3 power is representative of program material with severe clipping. CONTROLS Front: AC POWER switch, (8) gain controls Back: (4) DIP switch blocks with 10 poles (1 per channel pair): Clip Limiters, Stereo/Parallel/Bridge mode selection, low frequency filter on/off and frequency selection. DIMENSIONS 19.0" (48.3 cm) wide, 3.5" (8.9 cm) tall (2 rack spaces) 14" (35.6 cm) deep (from front mounting rails) including rear support ears WEIGHT 21 pounds ( 9.5 kg) net, 27 pounds ( 12.3 kg) shipping SPECIFICATIONS SUBJECT TO CHANGE WITHOUT NOTICE 31 WARRANTY INFORMATION & HOW TO CONTACT QSC WARRANTY (USA only; other countries, see your dealer or distributor) Disclaimer QSC Audio Products, Inc. is not liable for any damage to speakers, or any other equipment that is caused by negligence or improper installation and/or use of this amplifier product. Product Warranty QSC Audio Products, Inc. (“QSC”) guarantees its products to be free from defective material and / or workmanship for a period of three (3) years from date of sale, and will replace defective parts and repair malfunctioning products under this warranty when the defect occurs under normal installation and use - provided the unit is returned to our factory or one of our authorized service stations via pre-paid transportation with a copy of proof of purchase (i.e., sales receipt). This warranty provides that the examination of the return product must indicate, in our judgment, a manufacturing defect. This warranty does not extend to any product which has been subjected to misuse, neglect, accident, improper installa- tion, or where the date code has been removed or defaced. QSC shall not be liable for incidental and/or consequential damages. This warranty gives you specific legal rights, and you may also have other rights which vary from state to state. This limited warranty is freely transferable during the term of the warranty period. HOW TO CONTACT QSC AUDIO PRODUCTS Mailing address / Adresse postale / Postanschrift / Dirección postal: QSC Audio Products, Inc. 1675 MacArthur Boulevard Costa Mesa, CA 92626-1468 USA Telephone Numbers / Numéros de téléphone / Telefonnummern / Números de teléfono: Main Number / Numéro principal / Hauptnummer / Número principal +(714) 754-6175 Sales Direct Line / Ligne directe ventes / Verkauf-Direkt / Línea directo ventas +(714) 957-7100 Sales & Marketing / Ventes & marketing / Verkauf u. Marketing / Ventas y marketing (800) 854-4079 (toll-free in U.S.A. only) (sans frais aux É-U seulement) (zollfrei nur beim USA) (sin costo en EE. UU. solamente) Technical Service / Service à la clientèle / Kundendienst / Servicio a la clientela +(714) 957-7150 (800) 772-2834 (toll-free in U.S.A. only) (sans frais aux É-U seulement) (zollfrei nur beim USA) (sin costo en EE. UU. solamente) Facsimile Numbers / Numéros de télécopieur / Telefaxnummern / Número de FAX: Sales & Marketing FAX / Télécopie ventes & marketing / Telefax der Verkauf u. Marketing / FAX ventas y marketing +(714) 754-6174 Technical Service FAX / Télécopie service à la clientèle / Kundendienst-Telefax / FAX servicio a la clientela +(714) 754-6173 World Wide Web: www.qscaudio.com E-mail: [email protected] [email protected] 32 This page intentionally left blank. 35 Assessment of Errors in Sound Pressure Measurement in a Large Anechoic Chamber S.E. Keith, M.G. Davidson, S.H.P. Bly Health Canada, Radiation Protection Bureau, 775 Brookfield Rd., Ottawa, Ontario, K1A ICI Introduction faces of a dodecahedron with an average spherical radius of 19 cm. A large acoustical anechoic chamber has recently been put into The higher frequency source was a lead zirconate sphere with a operation at the Canadian Radiation Protection Bureau. The walls, 96.52 mm O.D. and a 4.92 mm wall thickness (Channel Industries, floor and ceiling of the anechoic chamber are lined with flat-tipped Santa Barbara, CA). fibreglass wedges designed for a cut-off frequency of 50 Hz. The Each sound source was supplied with a simultaneous mix of 11 interior (wedge tip to wedge tip) is 13 m long, 9 m wide and 8 m computer generated sinusoids at 1/3 octave band centre frequencies high. spanning the range appropriate to each source. For the low In order to create well-defined noise exposures and measurements of frequency measurements, the signal was supplied from a Nagra IV- sound power and sound pressure in the chamber, its free-field SJ tape recorder through a Brüel & Kjær (B&K) 2706 power performance must be quantified. This provides baseline estimates for amplifier. The driving signal for the higher frequency source was errors in sound pressure measurements and a technique for error output by computer through a National Instruments AT-DSP2200 estimation as measurement configurations change. This paper digital signal processor board and amplified using a Yamaha PC describes characterization of the free field performance from 50 to 5002 M 1000 W power amplifier. 5000 Hz along the chamber axis. Implications for measurement For directivity measurements, the low frequency source was rotated uncertainties in the anechoic chamber are discussed. by mounting it on a tripod centred on a B&K 3922 turntable. The Method and Apparatus higher frequency source was rotated by supporting it in the end of a The technique used was similar to that recommended in ISO1 and nylon stocking, hung 2 m below the spindle of a B&K 3923 rotating ANSI2 standards and used to characterize other large anechoic microphone boom. The sources rotated about their geometric chambers3,4 centres, and measurements were taken every 4 degrees. Repeated revolutions gave a total averaging time of 4 seconds per 4 degree Free-field performance was characterized by measured deviations sector. Nine directivity measurements were made at distances from the llr dependence of pressure,^, on distance, r, for a point between 31.5 and 200 cm from the centre of each source. source in a free field (inverse square law). This can be expressed as: Deviations from the inverse square law were obtained with the 1 Ip-rlA (1) sources hung from the ceiling and centred vertically in the chamber. Measurements were made along a horizontal line parallel to the long where A is proportional to the source strength. axis of the chamber. The mid point of this measurement traverse The deviation, e(r), from the inverse square law was obtained from: was near the centre of the chamber. Low frequency measurements were taken from 50 cm to 560 cm in 2 cm to 10 cm steps using 64 second linear time averaging. Measurements of the higher frequency e(r>201og^7r (2) source were averaged for 8 seconds at each position. At frequencies above 2500 Hz the microphone was positioned from 50 to 530 cm in where pmeas was the measured pressure at position r. The quantity 1 cm steps. For frequencies from 500 Hz to 2500 Hz, the PfU> at any value of r, was the pressure given by a linear least squares microphone was positioned from 50 to 545 cm in 5 cm steps. r fit to equation (1) of the pressures measured over a limited range of All measurements of the dodecahedron were made using a B&K values. 3545 intensity probe, with 12 mm 4181 microphones and a 50 mm The data range to be fitted was arrived at as a compromise between spacer. Microphone calibrations were made before and after several competing criteria. Ideally, to reduce the effects of echoes, measurements using a B&K 3541 system (with pistonphone). For the least squares fit should be determined from data measured as the piezoceramic sphere, all directivity measurements and deviation close to the source as possible. However, the measurement also has measurements above 2500 Hz were made with a B&K 4165 1/2" to be far enough away that the field is omnidirectional within ±1 microphone, attached to a nosecone, and powered by a B&K 2807 dB1,2. This limit on source directivity means that estimation of the microphone power supply. A B&K 4228 pistonphone was used for maximum measurement error (emax) can be determined microphone calibration for this latter system. To obtain an adequate experimentally within ±0.2 dB, for emax values less than 2 dB3. In signal to noise ratio at frequencies from 500 Hz to 2500 Hz with the addition, the measurement has to be far enough away that, to a good piezoceramic sphere, deviation measurements were made with a 1 " approximation, the free-field pressure field due to the source varies diameter B&K 4179 low noise microphone with nose cone. All with 1/r. This was diagnosed by measurements of sound source pressure measurements were made with a signal to noise ratio of at directivity as a function of distance as described below. Finally the least 30 dB. range must be large enough to sufficiently reduce statistical All positioning of the measuring microphone was controlled using a uncertainties in the linear least squares fit. B&K 9654 robot. A B&K type 2133 1/3 octave band frequency Generation of the acoustic pressures to be measured was done with analyzer was used for data analysis. two novel sound sources; a low frequency dodecahedral array of For both directivity and deviation measurements, all ancillary speakers for 50 to 500 Hz5 and a higher frequency piezoceramic structures within the anechoic chamber that could act as acoustical sphere for 500 to 5000 Hz. Both sources were symmetric on three reflectors, were carefully wrapped with fibreglass batts. axes, about a reference point. The low frequency source was constructed from twelve, 6 inch, 100 watt loudspeakers set in the - 33 - Results and Conclusions The least squares fit to equation (1) was made over the range /- 0.5 to 5.5 m anechoic limit r 1 m, where was taken as the separation between the microphone u 0.5 to 5.5 m and the source geometric centre. This was justified by the spatial o LO symmetry of the sources described above. □ to 4.0 m m2, Measurements suggested that the minimum source receiver c □ 0.5 to 2.0 m .2 2 separation for the least squares fit should be 50 cm. At distances <3 > ■ 0.5 to 1.0 m greater than 50 cm, the change in directivity with distance was 0.1 to a) 0.4 dB for the dodecahedron and 0.2 to 0.5 dB for the piezoceramic Q sphere. In most frequency bands, the inverse square law was validated to within ±0.4 dB for the sources used in this work. Furthermore, at distances greater than 50 cm, the maximum deviation from omnidirectionality was less than 0.8 dB. in The results shown in Fig. 1 are of maximum absolute deviations co versus frequency for four source-receiver separation distances up to Frequency (Hz) 1,2,4 and 5.5 m. In the frequency range from 100 to 5000 Hz, the maximum deviations from free field ranged from 0.4 dB at 100 Hz to Figure 1 : Measured maximum deviations from inverse square 1.0 dB at 5 kHz. At 50 Hz the maximum deviation rose to 3.6 dB. law for four source-receiver separation distances. Below 100 Hz, the source receiver separation must be reduced to about 2 metres in order to reduce errors to 1 dB. The maximum deviations described above, gave a "worst case" error estimate for pressure measurements in the anechoic chamber. These References errors could only occur for a source with strong tonal components. 1. Anon., IS03745, "Acoustics-Determination of Sound Power The maximum deviation, or error (emax) can be calculated by the Levels of Noise Sources-Precision Methods for Anechoic and Semi- summation of pressure amplitudes of direct and reflected sound6: Anechoic Rooms," International Organization for Standardization, Geneva, 1977 2. Anon., ANSI SI .35-1979, "Precision Methods for the 201og (3) u dh Determination of Sound Power Levels of Noise Sources in Anechoic and Hemi-Anechoic Rooms," American National Standards Institute, where R is the chamber wall reflection coefficient (<0.1 above the New York, 1979 cutoff frequency for an anechoic chamber), r is the distance between 3. F. Ingerslev, O.J. Pedersen, P.K. Moller, J. Kristensen, "New source acoustic centre and microphone, di is the distance travelled Rooms for Acoustic Measurements at the Danish Technical from source to microphone by a wave reflected from the ;'th wall. University," Acustica 19,185-199, 1967/68 The deviations presented here were consistent with the chamber 4. W. Koidan, G.R. Hruska, "Acoustical Properties of the National being anechoic down to 50 Hz. Measured deviations only reach the Bureau of Standards Anechoic Chamber," J. Acoust. Soc. Am. 64, deviations calculated using equation (3) (withi?=0.1) at the chamber 508-516, 1978 cutoff frequency of 50 Hz. 5. G.L. Basso, R.A. Williams and L.C. Hurtubise, "Dodecahedron Acoustical Source Strength Calibration," NRC Institute for The error estimate in equation (3) is conservative for spatially Aerospace Research, 1992 averaged measurements, a broadband source using a large analysis 6. J. Duda, "Basic Design Considerations for Anechoic Chambers," bandwidth (i.e., A-weighted totals), or a large device containing Noise Control Engineering, Sept.-Oct, 1977 multiple incoherent sources. In these cases the cross terms in equation (3) tend to cancel and in the limit the deviations are reduced to an energy type summation of the direct and reflected sound. =101og (4) The difference between the two measurement situations is seen, for example, when the deviation from equation (3) is 1 dB. Then the estimated error in equation (4) will be in the range 0.02 to 0.1 dB (the latter value is associated with a source or receiver position close to a single wall). For an omnidirectional source, initial estimates of expected errors can be made using the data from figure 1. For example, using equation (3), R can be estimated as 0.1 at 50 Hz, and ranges from 0.02 to 0.06 above 100 Hz. Substituting values for djs in equations (3) and (4) can give upper and lower bounds on pressure measurement errors in any position in the chamber. If the source is not omnidirectional, the djs in equations (3) and (4) must be weighted to obtain an error estimate. - 34 -