Augmented Passive Radiator Loudspeaker Systems, Part 1 PDF
Augmented Passive Radiator Loudspeaker Systems, Part 1 PDF
THOMAS L. CLARKE
394 © 1981 Audio Engineering Society, Inc. 0004~7554/81/060394-11500.75 d. Audio Eng. Soc., Vol. 29, No. 6, 1981 June
PAPERS AUGMENTED PASSIVE-RADIATOR LOUDSPEAKER SYSTEMS, PART I
RAL
UL2 1 =
=acoustic resistance
t°tal v°lume of driver enclosure leak-
°fpassive-radiat°renclosureleakage
vel°city _x"]j (Rg_ %'qj/
AB,_ [ Sr_flh 1_2 I
age losses
RAm = acoustic resistance of driver enclosure inter- Fig. 2. Complete acoustical analog circuit of the augmented
ual absorption losses passive-radiatorloudspeakersystem.
RIL e = acoustic resistance of passive-radiator en-
closureleakagelosses MASCAS
RAB 2 = acoustic resistance of passive-radiator eh- , 0 "8,
closure internal absorption losses ' T Mi;_
S D = effective area of driver diaphragm : RAT _ ._
Sp] = enclosurearea
of passive radiator coupled to driver CABIT/ Cfi['
P-j-
SE2 = passiVe-radiator
area °f passive
enclosure
radiat°r baffled bY _'_" g u°'_ 7
This circuit can be simplified by combining RAS with Fig. 3. Simplified acoustical analog circuit of the augmented
the resistance due to the electromagnetic damping of the passive-radiator system.
In particular, if the circuit values are identical, the driv- resonance due to losses from absorption, leakage, and
ing point impedance and the displacement functions of passive-radiator losses, respectively:
both the driver and the passive radiator are identical for
the two systems. Because of its construction, there is 1
QA - (]0)
additional volume velocity Up2 associated with the aug- roBCABiRABi
mented passive radiator, and the response functions of
the augmented and conventional passive-radiator sys- QL = roBCAmRALI (11)
tems are in general not identical. With proper choice of
1
system parameters, however, the response function of Qp - . (12)
an augmented passive-radiator system can be made roBCABIRAPT
driver with the passive radiator has also been neglected ,. _ - CAp (13)
These approximations are justified in [2]. CAm
It is useful to introduce characteristic times and di-
mensionless ratios which parameterize the loudspeaker the system tuning ratio
system. For the driver these are
fB rob Ts
h = - - (14)
1 fs % TB
Ts2 = --= CAsMAs (3)
ros2 and the passive-radiator
tuningratio
rosMas
The above value is an upper limit and may be closely then thesecondpairofzerosisalsolocatedatthe origin.
approached by making the passive radiator physically Iftheapproximation (22) is used, condition (28) reduces
large and thus increasing CAps. to
From the analysis of Fig. 3, using the methods of[6] 1
and substitutingthe above dimensionlessparameters, 7 _ _ (29)
the system describing functions can be derived. Allow-
ance must be made for the volume velocity Up2 which Thus if, Eq. (28) or Eq. (29) is satisfied, the response
does not appear in Fig. 3. For the augmented passive- function becomes similar to that of a vented-box system.
radiator system the total output is given by the vectorial The response notch which causes difficulty in drone-
sum conesystems isremoved.
The response function is The frequency response G(jto) of Eq. (23) is examined
in more detail in the Appendix (included with Part II),
G(s) = s2Ts2(s2Tp2+ 1 - 6(7 - 1)) (23a) andausefuldesignprocedureforavarietyofalignments
D(s) is given. These alignments are similar to those given for
the drone-cone system in [2] and for the vented-box
where system in[7].
and s = o + jto is the complex frequency variable, steeper slope at cutoff. If 3' % l/e, two response zeros
are located on the real o axis. The response in this case
The displacement function of the driver diaphragm drops off less steeply than even a pure fourth-order
normalized to unity at zero frequency is given by
response, and as the zeros recede from the origin, the
response approaches that of a closed-box system [8].
X(s) = (oF 3' (5 3' l)(S2rB2 + l) (24) This last form of response has no direct analog in pre-
D(s) vious loudspeaker systems and might be called quasi
second order.
and the displacement constant is
*zeros on the o axis: The points at a =x/2-, 3' = i corre- f3/fs=[ [.5 2 3
on a' and
enclose is useful when isthe
the 7passive-radiator volume unlimited
essentially available to /_ \," \,
that
passive-radiator system, so I /._' I!/ %
61_
_,I m ///_ \I_ \_.o_'4
3_'_fXx_ -i
I
.5 I 2 ;:5 5 I0
CAB2 (21{
CAp = CAPS (< (T -- 1) 2 ' (30)
Fig. 4. Curves of constant e, (5,andre/rs for maximally fiat
augmented passive-radiator system alignments plotted against
In this case CAp and (5are determined entirely by the pas- a and 3'.The curve BW is the locus of Butterworth alignmer/ts.
sive-radiator compliance and e is irrelevant.
Alignments with 7 < 1 have not been investigated.
With proper.choice ofSpl.andSe2anyvalueof 7 should 6,4S=o9 I .1.5 2
1.67 and 3' = 2.33. System responses for these cases are
the charts. The responses tend to be smoother for 5' < Ot=o6_ ,3
l/e, especiallyvalues
intermediate for small values
of e and ofa.be However,
3' can interpolatedthe from
case _.._
3/ _ 1/_ tends to produce a lower cutoff frequency. _ _10
mO
9 (:z=.6
z- .5
-I0
,.C//-,oo /// /_,o
/ / / / _=,:_,
---I0 t.5
/ / / / _:5't
_
-5>0
////
, , , , -20
///
, , , ,
.5 I 2 4 .5 I 2.. 4
cots wTs
Fig. 7. Responses obtainable from the augmented passive- Fig. 10. Responses obtainable from the augmented passive-
radiator system for 3' = 1.67, e = 0.40. radiator system for 3' = 2.33, e = 0.55.
0 - --ELLIPTIC I 'QB3
O[ =o6-- f3/fs
h
J.AudioEng.Soc.,
Vol.29,No.6,1981
June 399
CLARKE PAPERS
augmented passive-radiator system is similar to that for shown in Fig. 17. Passive-radiator losses have the least
passive-radiator systems. The effect of these losses on effect on the system and absorption losses the greatest
the response of the a = 5 Butterworth alignment is effect.
Rather than presenting four more alignment charts
........ EL!IPTIC' , , for the lossy augmented passive-radiator system, we
.6 _e=[. ,C=.75 3 f3/fs duce adjusted values of h,f3, and QT in the presence of
h losses:
[ QT¢'i_ [l__s
I ! found the following equations to be useful as they pro-
.4 2 y ] _ ] + ] 1
QTI _ _ I Qi_' QA _-L + 2Qp (3la)
Fig. 12. Alignment chart for lossless augmented passive-fa- Eq. (3lb) is derived by considering the effect of losses
diator system with 7 = 1.67, e = 0.75.
on the passive-radiator volume velocity at the enclosure.
resonance frequency. Shifting the enclosure resonance
t Ltl PHC' frequency upward by the amount given in Eq. (3lb)
[ '_, ........ I_/_/_f3 ? restores the passive-radiator output at the box reso-
,6 =2_3_ ' _='_ _ f3/fs nance frequency to its lossless value. The first term in
c>( h
Fig. 13. Alignment chart for lossless augmented passive-ra- Ct = 5 +50% -_x'x
diator system with }, = 2.33, e = 0.33. +20%L_
_'= m.5 x,-[--..
/
....... _ ' ' ' [ mlJm_ ,_ I 03 0 ¢=.67 /
.6 [ I _'--*-2.33, ¢=.55 I / jf ._ -.o
r I , ,Ira .... ' 11[/,/13 lO f3/fs _ -"_
'_1 _ _ Ii
olq7 I111!1
I Io
.3 .s.zl 2 3 57 _o -2o ,
.5
,
I 2
, 4
,
c< o_Ts
Fig. 14. Alignment chart for lossless augmented passive-ra- Fig. 15. Effect of misalignment of h on augmented passive-
diator system with 7 = 2.33, e = 0.55. radiator system response.
overvolume the box by 40% can be used to compensate 3.2 System Response Factor
for losses (as in [11]). Values of K,(c) were calculated for the lossless align-
3 EFFICIENCY ments presented in Figs. 6 and 11-14. They are plotted in
Fig. 19 together with values for drone-cone and vented-
3.1 Efficiency Factors box alignments as a function of a. For 3' _ 1/e the
system response factor is up to 50% greater than values
The piston range reference efficiency of the driver is for vented-box or drone-cone alignments having the
given by [10, Eq. (32)] same a. Alignments with 3' < 1/e (including the But-
4rr2 fs 3VAS terworthalignments)havesystemresponsefactorscom-
r/° - c3 QEs (33) parable to or a little smaller than those for the vented-
box or drone-cone systems.
Using the method of [7, sec. 5], Eq. (33) may be put into The Butterworth alignments (line BW in Fig. 19) illus-
the form trate that for identicalalignmentsthe systemresponse
factors are identical. As shown in the Appendix, Part II,
770= K f3 3 V B (34) any vented-box or drone-cone system alignment can be
extended to an entire family of augmented passive-radi-
where V_ is the net internal volume of the system enclo- ator system alignments; thus each alignment correspond-
sure, and K_ is an efficiency constant consisting of two
factors:
QT
Kv(Q)- QEs (37)
QT +100%
ir =1.5 _-_
oD 0 ¢=,67 rrlO- a=5 LOSSLESS__
'_ _=5 * 20/_7/_ -_ _= I.5 /.,'/It/
9 9 _ ¢=.67
1A/l-
-50% _ COMPENSATED
I0 - -I0 -
-20 , , , , -20 , , , ,
.5 I 2_ 4 .5 I 2 4
cots cots
Fig. 18. Augmented passive-radiator system response after
Fig. 16. Effect of misalignment of QTon augmented passive- alignment has been adjusted to compensate for enclosure
radiatorsystemresponse, losses.
d.Audio
Eng.Soc.,
Vol.29,No.8,1981
June 401
CLARKE PAPERS
lng to a point on the vented-box curve is the starting volume of the driver; here XMA X is the peak linear dis-
point ufa family of alignments. If the efficiency constant placement of the driver diaphragm. Kp is a power rating
of this family were plotted, it would be a horizontal line constant which depends on the alignment. In [7] and [2]
extending to the right of the starting point, it is shown that the value K v = 3 is a good approxima-
tion for all drone-cone and vented-box alignments. This
4 LARGE-SIGNAL RESPONSE value is also suitable for rating augmented passive-rs-
4.1 Driver Diaphragm Displacement diator systems. In practice the value Kp = 3 is very
conservative and can be increased several times in sys-
The driver displacement function for augmented pas- rems designed for ordinary domestic music reproduc-
sive-radiator systems [Eq. (24)] and the displacement tion:.
constant K x [Eq. (25)] are similar to those for drone- The displacement-limited electrical power input rat-
cone systems [2, Eqs. (23), (24)]. Because augmented lng is obtained by combining Eq. (38) with Eq. (34):
passive-radiator alignments generally have smaller val-
LOSSLESS
.... i-'" ' ,n -to _c,:=l
%
/rz,.O_.k[ 5 ',,
/ ........ 4
IV, ENT,ED, III _%'?--- Kzt(G) _'-20
I 3
· [lll[[ [ [
I D0¢Jq4>C-L
[ ilr: 1.67 c:,40 / 2
The exact displacement requirements of the passive ra- the driver in the low e case for low values of a, rather
diator, of course, depend upon the spectrum of the than to a very much larger passive-radiator displace-
driving signal. Assuming a signal limited to the system ment in the high e case. This reflects the conservatism of
passband, Fig. 22 shows the ratio of maximum passive- the values of Kp given earlier. A value of VpR/Vr) = 2
radiator displacement volume VpR to driver displace- can thus be adopted for all alignments of augmented
ment volume for two alignments versus a. VpR is the passive-radiator systems. This value will conservatively
maximum linear volume displacement of the passive predict the required passive-radiator displacement.
radiator, that is, the sum of Spl and Sp2 times the maxi-
mum linear displacement. Forlarge a the twocurves are 5 PARAMETER MEASUREMENT
similar, and the passive-radiator displacement volume The small-signal parameters a', 6, h, QT, Ts, and 3' of
must be about twice that of the driver. The ratio ap- the augmented passive-radiator systems may be mea-
proaches 1 for the low e high 7 alignment at low a, and
sured by the method given in [2, sec. 8]. The augmented
increases for the higher e, lower 3/alignment. This diver-
gence is due more to the larger displacement required of passive radiator has additional parameters 'y and e. Also
the value of CAps or 8' = CAPs/C/_m may be desired.
The value of 3/ may, of course, be obtained approxi-
0 _'=I,5 mately
measure from physical the
accurately measurement, but it is difficult
effective diameter of a dia-to
o-_
-,o 3/ = 1 + S
')
6' ' (40)
_ 6 REFERENCES
ua
,,N [1] H. F. Olson,"Loud Speakerand Methodof
-J PropagatingSound,"U.S.Patent1,988,250, application
_-20 1934 Feb. 17; patented 1935 Jan. 15.
tv-
O :'R [2] R.H. Small,"Passive-Radiator LoudspeakerSys-
z tems, Part 1, Analysis," J. Audio Eng. Soc., vol. 22,
pp. 592-601 (1974 Oct.).
[3] Y. Nomura and Z. Kitamura, "An Analysis of
Design Conditions for a Phase-Inverter Speaker System
-3.0 with a Drone Cone," IEEE Trans. Audio Electroacoust.,
,5 I 2 4 vol.[4]AU-21, pp. 397-407
T.L. Clarke, (1973 Oct.).
"Augmented Passive Radiator Loud-
W / 2 11_'3 speaker," U.S. Patent 4,076,097, application 1976 Aug.
4, patented 1978 Feb. 28.
Fig. 21. Displacement functions of driver and passive radia- [5] E. Hossbach, "Loud Speaker System," U.S. Pat-
tor as a function of normalized frequency, ent 3,772,466, application 1971 Nov. 24, patented 1973
Nov. 13.
[6] R. H. Small, "Direct-Radiator Loudspeaker Sys-
,3 ,5 2 I 2 3 5 7 10
pp63 _639 ,9730
J. AudioEng. Soc., vol. 20, pp. 798-808 (1972 Dec.); vol.
21 pp. 11-18 (1973 Jan./Feb.).
[9] A. N. Thiele, "Loudspeakers in Vented Boxes,"
Proc. IREEAustralia, vol. 22, pp. 487-508 (1961 Aug.);
republished in J. Audio Eng. Soc., vol. 19, pp. 382-392
(1971 May); pp. 471-483 (1971 June).
C_ [10] A. Budak and P. Aronhime, "MaximallyFlat
Low-Pass Filters with Steeper Slopes at Cutoff," IEEE
Fig. 22. Required ratio of passive-radiator displacement vol- Trans. Audio Electroacoust., vol. AU- 18, pp. 63-66 (1970
ume YPRto driver displacement volume VDas a function of a Mar.).
for two augmented passive-radiator alignments. [11] D. B. Keele, Jr., "A New Set of Sixth-Order
O.AudioEng.
Soc.,Vol.29,No.6,1981
June 403
CLARKE PAPERS
Vented-Box Loudspeaker System Alignments,"./. Audio er," CQ, vol. 20, pp. 101-104 (1979 Feb.).
Eng. Soc., vol. 23, pp. 354-360 (1975 June). [13] L. Weinberg, Network Analysis and Synthesis
[12] T.L. Clarke, "Build a Superthruster Loudspeak- (McGraw-Hill, New York, 1962), chap. 11.
THE AUTHOR
University of Virginia in 1975. He is currently a Ph.D.
candidate in applied mathematics and ocean engineer-
ingat the University of Miami in Coral Gables, Florida,
and expects to receive his degree in 1981 December.
From 1971 to 1973 Mr. Clarke was a radio engineer
for WFUN Miami after serving two years in the U.S.
Army. Since 1975 he has been a mathematician at the
Atlantic Oceanographic and Meteorological Laborato-
ries of the National Oceanographic and Atmospheric
Administration (NOAA), first at Marine Geology and
Geophysics Laboratory and currently at the Ocean
Acoustics Laboratory.
Mr. Clarke's research interests include modeling of
Thomas L. Clarke was born in Miami, Florida, in sediment transport in the ocean, as well as acoustics. He
1948. Florida International University in Miami has published over 15 technical and scientificpapers. He
awarded him a B.S. degree in mathematics in 1973, and is a member of the Audio Engineering Society and the
he received an M.S. in applied mathematics from the Society of Industrial and Applied Mathematicians.