вход по аккаунту



код для вставкиСкачать
reduction of transmission delays in time-encoded speech
systems by an order of magnitude, compared with constantlength code allocations at identical transmission bit rates. They
should be considered as an alternative to linear codes in TES
system design. The effects of these codes on receiver delay and
their tolerance to transmission errors are matters currently
under examination.
Acknowledgment: The willing support of P. Cooper, in setting
up the Miproc TES link and devising and implementing the
algorithms for the simultaneous real-time comparison of competitor codes, is gratefully acknowledged.
ducer controlled by external voltage.
In this work the first experimental observation is made of
the transduction of the acoustic waves into an electric signal
using the principle described above. Configuration used in the
experiment is shown in Fig. 1. The SAW of the frequency
co/2n = 560 MHz was excited on the YZ-cut surface of
LiNbO 3 crystal by the conventional interdigital transducer.
HF input signal
ohmic contacts
pulsed or DC
ef power supply
2nd April 1981
School of Electrical Engineering
University of Bath
Claverton Down, Bath BA2 7A Y, England
KING, R. A., and GOSLING, w.: Time-encoded speech', Electron.
Lett., 1978, 14, (15), pp. 456-457
KING, R. A., and GOSLING, W.: 'Time encoded speech (TES)'. IEE
Conf. Publ. 180, 1979, pp. 140-143
Fig. 1 Configuration of acoustoinjection transistor using SA W and its
TURNER, L. F., FRANGOULIS, E., and ALCAIM, A.: 'Some considerations
relating to the performance of variable-information-rate-source to
constant-transmission-rate schemes of data compression', IEE J.
Comput. & Digital Tech., 1979, 2, (3), pp. 134-141
4 MASON, D. c , and BALSTON, D. M.: 'Relationship between system
delay and transmission rate in time-encoded speech', Electron.
Lett., 1980, 16, (4), pp. 128-130
output signal
to oscilloscope
- load resistance
On the surface of LiNbO 3 there was deposited the photosensitive semiconductor film of CdSe, which allowed one to
investigate the AIT effect as a function of the electron concentration controlled by the light intensity. The thickness of the
film was 2 /im. Ohmic contact to the film was made in the form
of interdigital transducer, the width of electrodes and interelectrode spacings being equal to half a wavelength of the
One may easily understand that in this case the changes of
the conductance of all interelectrode spacings due to propagation of electron bunches produced by SAWs are in phase with
each other. Thus, propagating SAWs changes the resistance of
all the structure which leads to the appearance of an alternating current of frequency co in the collector circuit. The
amplification coefficient K of the device, determined as a ratio
of the power on load resistance Z to the power applied to the
input SAW transducer, is given by the expression
TURNER, L. F., FRANGOULIS, E., and ALCAIM, A.: 'Further results on
relationship between system delay and transmission rate in time
encoded speech-type systems', ibid., 1980, 16, (25), pp. 947-948
HUFFMAN, D. A.: 'A method for the construction of minimum redundancy codes', Proc. IRE, 1952, 40, pp. 1098-1101
Indexing terms: Signal processing, Acoustic devices, Semiconductor devices and materials
A new principle of signal amplification is experimentally
demonstrated which is based on the effect of modulation of
conductivity of a semiconductor due to bunching of electrons
by the acoustic wave. For the layered structure LiNbO 3 -CdSe
good agreement between theory and experiment is obtained.
In Reference 1 a new principle of signal amplification was
proposed. It is based on the effect of modulation of the resistivity of a semiconductor due to electron bunches accompanying
the bulk (BAW) or surface (SAW) acoustic waves in piezoelectric semiconductors or in layered structure piezoelectric
semiconductors. The semiconductor is put in series with battery and a load resistor ('collector circuit'). An alternating electric signal applied to the input transducer excites the acoustic
waves which modulate the resistivity of the semiconductor
directly proportional to the deformation ('emitter circuit'). By
proper choice of semiconductor parameters and configuration
of the device the power on the load resistance may exceed the
input power for excitation of acoustic waves and one has the
amplification of input signal. In analogy with a conventional
transistor this device could be called an 'acoustoinjection transistor' (AIT). The level of output signal, and consequently
the efficiency of transformation of acoustic energy into electric
one, is determined by the battery voltage. This allows one to
consider this semiconductor with the corresponding
configuration of electrodes also as an acoustoelectric trans396
and Vs = sound velocity, q = o)/Vs, x = xM £, r = rDx, T M and
rD = Maxwellian relaxation time and Debye length, respectively, n and S = effective electromechanic coupling constant
and acoustic beam cross-section, e = dielectric constant of the
piezoelectric, £ and x = constants, taking into account properties of the layered structure, 2 x = a numerical constant of the
order of unity, Z = terminal resistance of the film, a = NWd/l,
N = number of interelectrode gaps, W — width of the gap,
/ = width of the sound beam, fi = electron mobility, y and
aL = conversion efficiency of input transducer and lattice
attenuation (in dB), respectively. The above expression is valid
also for devices using BAW in piezoelectric semiconductors if
one puts N = 1, b. = b + , S = Wd and £ = % = 1.
The experiments of the observation of the AIT effect were
performed in the pulse regime of work of the input HF oscillator, the collector battery power supply being both pulsed and
DC (at V < 15 V). The oscillograms of the observed signals are
presented in Fig. 2. The first RF pulse corresponds to parasite
throughfeed signal from the output of HF oscillator. The
second pulse corresponds to the AIT output signal delayed by
the transit time of SAW between input and 'collector' circuit
transducers. It appears only after switching in the collector
battery voltage V, and depends on its value.
The oscillograms of Figs. 2a and b illustrate the growth of
the signal due to change of V from 1-5 to 12 V, respectively.
Notice that the value of the level of the output signal does not
11 th June 1981
Vol. 17 No. 12
depend on the polarity of the collector battery. Relative change
of the output signal level as a function of the collector battery
voltage is presented in Fig. 3. The solid curve corresponds to
the theoretical calculations (eqn. 1). The dots represent the
results of the experiment. One may see that by changing the
collector battery voltage it is possible to change
the amplification coefficient in wide limits. At V > 10 V one
observes the saturation of amplification, which may be related
to reduction in the electron bunching in strong electric fields
when electron drift velocity Vd $> Vs.3 In the work we also investigated the dependence of the amplification coefficient of
AIT on the electron concentration in the film, on the value of the
load resistance and on the number of interelectrode spacings.
The results obtained as well as the data of Fig. 3 are in a good
agreement with the theory. By using the LC circuit as a load
resistance the voltage amplification coefficient of the device
reaches unity.
In this work the results of the first experiments are presented
on the observation and investigation of a new type of acoustoelectronic device—the acoustoinjection transistor. It is shown
that by varying the collector battery voltage it is possible to
change the conversion efficiency of SAW energy into an electric signal by hundreds of times. Estimates along eqn. 1
show that using conventional materials and appropriate constructions of a device it is possible to obtain the power
amplification in AIT both using surface and bulk acoustic
The authors are grateful to S. V. Boritko, I. M. Kotelyansky,
A. I. Krikunov and E. N. Mirgorodskaya for helpful cooperation and discussions.
24th March 1981
Institute of Radio Engineering & Electronics
Academy of Sciences of the USSR
Moscow, USSR
1 GULYAEV, YU. v.: USSR patent 101275, 6 July 1973
2 KINO, G. s., and REEDER, T. M.: 'A normal mode theory for the
Rayleigh wave amplifier', IEEE Trans., 1971, ED-18, pp. 909-920
3 WHITE, D. L.: 'Amplification of ultrasonic waves in piezoelectric
semiconductors', J. Appl. Phys., 1962, 33, pp. 2547-2554
0013-5194/81/120396-02$! .50/0
Indexing terms: Radar, Radar cross-section diagram
The use of electromagnetic waves for the recognition of a
structure represented by point scatterers is a fundamental
problem. Much research is done on this subject, and the study
of aircraft observed in the yaw plane gives interesting results.
But to apply these methods it is necessary to use many sophisticated acquisition systems. In the letter, we give a new
method which can be applied for plane structures composed
of isotropic scatterers. This method is quite interesting because it uses power measurements only and necessitates only
a classical tracking radar.
Fig. 2 Oscillograms of output signal
(a) V = 1-5 V
(/>) V = 12 V
Horizontal scale: 2 //s/div.; vertical scale: 01 V/div.
Introduction: We have in a former letter1 the analytic expression of the RCS diagram in the far field for a radar target
represented by point scatterers. In the case of a plane target
consisting of N isotropic point scatterers with RCS <r7separated by the relative distances \djk\ we obtain
ff.v(0) =
j. + 2 £
Ajk cos [BJk cos (0 + </>,,)]
= \/(<Tj<rk) and
Fig. 3 Dependence of relative change of amplification coefficient of AIT
on collector battery voltage
11th June
No. 12
BJk =
In this relation \jiik represents the angle between the bipoint
(j, k) and the reference bipoint {/, j] which is observed on the
angle of incidence 0 by the radar.
The problem of structure recognition consists of determining the parameters o-}, \djk\ and \\i)k from a record of the function CT.v(0).
For any multipoint of order N the problem is equivalent to
the research of the [N(N — l)/2] bipoints of which it is
For a bipoint (j, k) it is also necessary to find the two geometrical parameters \\d\j and ij/jk and to establish the relation
/( f f j , ok) = 0.
Без категории
Размер файла
320 Кб
Пожаловаться на содержимое документа