NO136017B - - Google Patents

Description

The present invention relates to acoustic surface waves

and apparatus for such bolts. The term "acoustic surface waves" will be used in the following to include both acoustic interface waves and acoustic surface

waves .

Application of acoustic surface waves is now proposed for

more and more purposes in electronics, and filters and delay lines for acoustic surface waves are likely to find important applications in the future. Such devices usually include a converter for sending out acoustic surface waves along a predetermined path, which can be made up of a surface or interface for materials in which surface acoustic waves can be propagated, but need not have any special design and limitation; as well as at least one further converter for detecting the acoustic surface waves and producing electrical signals depending on these waves. The converters usually used comprise comb-like electrodes, the teeth of which mesh with each other. If such electrodes are placed on a piezoelectric material,

application of electrical alternating current signals of a suitable frequency across the electrodes could cause the propagation of a surface acoustic wave perpendicular to the teeth in the aforementioned comb-like electrodes. Likewise, an applied acoustic surface wave perpendicular to said teeth will induce a corresponding electrical alternating current signal between the electrodes. It is also

known that such converters can also work effectively on an electrostrictive material, if an electric bias field

the material is pressed under the converters. Said converters can be designed to provide a filter effect.

It is an object of the present invention to provide means for coupling acoustic surface waves in such a way that a desired proportion or practically all of the energy in an acoustic surface wave in a first area can be transferred to acoustic surface waves in another area. It is a further purpose of the invention to indicate the design of various new devices which comprise one or more such coupling devices and constitute components with practically applicable properties, as said devices can be used either to achieve new or improved technical effects or as alternatives to known electronic components.

The present invention thus relates to a device for acoustic surface waves and which comprises a first and a second area of a material in which acoustic surface waves can be propagated as well as a coupling device composed of several wire-shaped conductors, the distinctive feature of the invention being that the wire-shaped conductors of the coupling device are arranged separately from each other and electrically isolated from each other and each extending over at least part of both the first and the second area to produce mutual influence between acoustic surface waves propagating across

said wire-shaped conductors in resp. the first and the second area.

Surface waves that are propagated across the filamentous ones

conductors in the second area, by means of electrical alternating current signals induced in said conductors. Said material may be a piezo-electric material, in which case the coupling device may simply consist of a series of wire-shaped conductors extending over the first region in a direction perpendicular to the propagation direction of the surface acoustic waves in this region, as well as further extending over the other area in a direction perpendicular to the propagation direction

for the acoustic surface waves in this area. The wire-shaped electrical conductors do not need to have any mutual electrical contact connection.

Alternatively, said material can be an electro-strictive material/ in which case the connection device must also include means for applying an electric bias field to the material under the wire-shaped conductors in said first and second area. Devices that use electrostrictive material in a similar way have been described in more detail in British Patent No. 1,330,142.

As another alternative switching device, electromotive action can be used. In this case, the wire-shaped conductors are interconnected at their ends to form closed circuits, and means are provided for maintaining a magnetic field perpendicular to the respective wire-shaped conductors over each of the areas in which mutual influence is to be created.

As a further alternative, the coupling device can utilize the magnetostrictive effect. In this case, said material must be a magnetostrictive material of such a nature that it does not short-circuit the electrical signals induced in the wire-shaped conductors, while these conductors are interconnected at their ends to form closed circuits and means are arranged to apply a magnetic bias field on the material in the aforementioned first and second area.

The device can be formed on the surface of any piece of a suitable material, or on a thin layer of a usable material on a substrate. The device can also be designed on any substrate in which acoustic surface waves can be propagated, and with a thin film of suitable material to achieve the desired coupling effect, arranged on the substrate only over the areas where said

coupling effect is desired.

The device may be covered with a film or layer

of protective material, so that the surface on which

the conductors are arranged, covered. Each protective

material that will cause significant damping of the acoustic surface waves should be carefully avoided.

The coupling device can be arranged for each other

coupling of acoustic surface waves acting in

two areas of a single propagation path for such waves,

or for interconnecting acoustic surface

waves that appear in separate areas of two separate propagation paths for surface acoustic waves, and which

does not need to be of the same width, although coupling between propagation paths of the same width gives maximum efficiency

tight.

Connection sections of the wire-shaped conductors can be

arranged over a material which absorbs or in which acoustic surface waves cannot propagate; as this

material can advantageously be made up of a piece of material with

low dielectric constant.

The simplest and preferred form of connection is the piezo-electric connection. The explanatory descriptions that will be given in the following will refer to embodiments with piezoelectric coupling, which means embodiments with at least one layer of piezoelectric material or a thicker piece

of such material above or below each person concerned

.converters and areas in which electroacoustic coupling is to be

take place; if no special reference is made to another form of connection. However, it should be clear that .

in most cases, corresponding structures can be created for the use of the above-described, alternative connection

forms.

Different embodiments of the invention in use

of piezo-electric switching devices will now be described with reference to the attached drawings, in which:

Fig. 1 is a plan view of a coupling device adapted to transfer acoustic wave energy from a propagation path to a nearby propagation path on the same substrate, Fig. 2 is a plan view of a coupling device adapted to transfer acoustic wave energy from a propagation path to form convergent surface acoustic waves in a nearby propagation path on the same substrate, Fig. 3 is a plan view of a coupler arranged to transfer the energy in an acoustic surface wave on a substrate to a nearby substrate, ' Fig. h is a plan view of a coupler arranged for splitting acoustic wave power between two separate propagation paths, to create surface acoustic waves with a 90° mutual phase shift in the two paths, Fig. 5 is a plan view of a beam switch for surface acoustic waves and arranged to produce a surface acoustic wave in one or the other of two output paths, depending on the sign of a 90° mutual phase shift ng on the input side, Figs. 6 and 7 are schematic plan sketches of alternative beam concentrators for surface acoustic waves, and arranged to produce surface acoustic waves with a small beam width, Fig. 8 is a plan sketch of a processing circuit for surface acoustic waves, Figs. 9 and 10 are plan sketches a/ alternative delay lines for acoustic surface waves and provided with drains, Fig. 11 and 12 are plan sketches of broadband path changers for acoustic surface waves-, Fig. 13 is a plan sketch of a resonator or recirculating delay line for acoustic surface waves and with two path changers, Fig. 1<*>+ is a plan view of a delay line for surface acoustic waves and with angle couplers, Fig. 15 is a plan view of a folded delay line for surface acoustic waves, Figs. 16 and 17 are plan views of alternative broadband unidirectional converters for surface acoustic waves , Fig. 18 is a plan view of a reflector for acoustic surface waves , Fig. 19 is a plan view of an alternative embodiment of a path changer for surface acoustic waves, Fig. 20 is a plan view of a unidirectional converter for surface acoustic waves, Fig. 21 is a plan view of a delay line for surface acoustic waves and with taps ,

Fig. 22 is a schematic representation to facilitate the explanation

of the operation of the tapped delay line for surface acoustic waves in FIG. 21, Fig. 23 is a plan sketch of a delay line for acoustic surface waves, and which includes means for suppressing third-order propagation of the signals along the relevant path, Fig. 2h is a plan sketch of a reflective delay line for acoustic surface waves,

Fig. 25 is a plan view of an amplifying path changer,

Fig. 26 is a plan view of a directional filter,

Fig. 27 is a plan view of a variable directional coupler,

Fig. 28 is a connecting core for a transducer arrangement arranged for the emission of surface acoustic waves alternatively in symmetrical or anti-symmetric mode, Fig. 29 is a diagram illustrating alternative transducer arrangements for the emission of surface acoustic waves in anti-symmetric mode, Fig. 30 is a schematic plan view of an anti-symmetric beam splitter that is applied to a signal in anti-symmetric mode, Fig. 31 is a schematic plan view of the anti-symmetric beam splitter in fig. 30■impressed a signal in symmetrical mode, Figs.. 32 and 33 are plan sketches of coupler adaptations arranged to reduce arbitrary reflections, Fig. 3^ is a plan sketch of a light-controlled coupler for acoustic surface waves, Fig. 35 is a plan sketch of an electrically controlled coupler for surface acoustic waves, Fig. 36 is a circuit diagram of part of the coupler shown in fig. 35, Fig. 37 is a perspective sketch of an electronic component in the connection diagram in fig. 36) °g Figs 38 and 39 are respectively a plan view and a wiring diagram for an alternative electrically controlled surface acoustic bubble coupler. Fig. 1 shows a plan view of a coupler arranged to transfer surface acoustic waves from a propagation path A to a nearby parallel propagation path B on one and the same substrate 1. The surface acoustic wave substrate 1 can be a piezoelectric material such as quartz, lithium- niobate or lithium germanate, a thin film of aluminum nitride applied to a non-piezoelectric single crystal substrate, or a thin film of piezoelectric material, e.g. zinc oxide sputtered on a non-piezoelectric, amorphous substrate, e.g. glass. Alternatively, the various transducers and coupler elements shown may be formed on a non-piezoelectric substrate (e.g. glass) in which surface acoustic waves can be propagated and which is provided with a thin film of piezoelectric material, e.g. zinc oxide sprayed or otherwise applied over or under the converters and coupling elements to increase their efficiency. A converter 3 with comb-like electrodes, the teeth of which mesh with each other, is arranged on the substrate 1 in a suitable position for sending out acoustic surface waves along the propagation path A. An acoustic surface wave coupler 5 is placed on the substrate 1. The coupler 5 is made of several vaporized wire-shaped conductors, each of which has a length of 2b and is arranged mutually separated and parallel and at right angles to the propagation paths A and B. The dashed line S indicates a line of symmetry which bisects the coupler 5 and extends parallel to the paths A and B. The wire-shaped conductors in the coupler 5 may be separated by equally large spaces, monotonically varying spaces or spaces whose width varies in any regular or arbitrary manner. Another cam converter of the above type is formed on the substrate 1 on top of the propagation path B, near the end of the path B which is farthest from the converter 3- The converters 3 °S 5 are provided with ordinary electrical connections (not shown) with external circuits, but the wire-shaped conductors in the coupler 5-needs not necessarily to be provided with electrical connections, and must be electrically isolated from each other. It should be noted that Fig.-1 and the other plans are schematically drawn, so that they make no attempt to show the width of each wire-shaped conductor or the required number of such conductors with any degree of accuracy.

It has been found that when surface acoustic waves are coupled

a series of wire-shaped conductors extending across the propagation path of the surface waves in question created electric fields between nearby conductors, so that these fields in turn can induce acoustic surface waves in any other propagation path of such waves and which is crossed by the concerned array of wire-shaped conductors. In the simplest embodiment of such a series, which is indicated in the coupler 5? the two halves of the array on opposite sides of the symmetry line S act as coupled structures, and will tend to exchange energy between waves propagating under each half of the coupler.

This effect can be explained by the theory that acoustic surface waves can propagate in piezoelectric material under a series of wire-shaped conductors across the direction of propagation, in two main modes, namely a symmetric mode and an anti-symmetric mode. In the symmetrical mode, the waves under both halves of the array are in phase with each other and their amplitude is constant over the entire width of the wire array. In the anti-symmetrical mode, the signals under the two halves of the array are of the same amplitude, but are mutually antiphase. If a wave in antisymmetric mode is combined with a wave in symmetric mode and of the same amplitude, the result will be an acoustic surface wave under only one half of the wire array, as the two modes will cancel each other under the other half. When exciting a surface acoustic wave that arrives only under one half of the wire row, this will be equally distributed between the symmetric and the anti-symmetric mode. The wave in the anti-symmetrical mode will, however, cause currents to flow along the wire-shaped conductors, and this wave mode will therefore propagate more slowly than the symmetrical mode. The phase relationship between the symmetric and the anti-symmetric mode will therefore change successively as the signals propagate forward", and this has an effect corresponding to energy transfer from said surface acoustic wave arriving along the propagation path A under one half of the coupler, to a new acoustic surface wave in propagation path B under the other half of the coupler When both waves have propagated a distance which will henceforth be

called L and is sufficient to produce a phase shift on

TT radians between the signals in the symmetric and anti-symmetric mode respectively, practically all of the wave energy that originally propagated along the path A will be transferred to the propagation path B. If the string of wires extends further along the propagation paths and the waves are allowed to continue their propagation unhindered a further distance L along the paths, if power losses in the paths are disregarded, practically all energy will be transferred back to path A. It follows that in order to transfer energy, from the surface acoustic wave in path A to path B, the coupler 5 extend over a distance L (or an odd multiple of L) in the direction of propagation of the waves. The length L can at least be approximately calculated in the following way in the case of a connector with evenly distributed conductors on piezoelectric material-

where - 9 = a <J d/s ,

Npp. is the number of conductors required to provide maximum energy transfer,

( jJ is the angular frequency,

d is the space between the center lines for nearby,

stranded conductors,

s is the propagation speed of the acoustic surface waves,

K is the electromechanical coupling constant, as well

F and a are factors depending on the "material" in question and the relationship between the thickness of the wire-shaped conductors and their mutual spacing.

For Y-cut lithium niobate with conductors Uke as wide as their mutual spacing, and arranged to propagate the surface acoustic waves parallel to the Z axis of the crystal, a = 0.75 and F = 0.85.

Under the same conditions, the general characteristics of a connector with N wires will be indicated by a scattering matrix M:

This coupling effect can take place over a wide frequency range, which is limited by a cutoff frequency that occurs when the space between the conductors becomes approximately equal to half the wavelength of the acoustic waves in the material in question. The formulas above do not apply in the stop band beyond the cut-off frequency. The band width can be increased by unequal or arbitrary distribution of the conductors.

In such a case, the formula for Np should be modified slightly, but

will still apply approximately. L will be equal to NT times the average spacing between the conductors.

When electrostrictive coupling or coupling with motor power is used, the system constants will be different. In the electrostrictive case, the constants will be functions of the applied bias field.

The coupling effect for the wire row will only be slightly changed

if the path is curved or the active parts of the row are separated from each other, which means that the wire-shaped conductors are provided with sections that only serve as electrical connections between the parts of the wire row that are respectively located above the first and the second area. However, a complete energy transfer is only possible if the effective width of path A is equal to the effective width of path B (it is assumed that the paths are of the same material. If the two paths are of different width, a modified theory must be applied and similar, but less good results are achieved.

In some of the devices that will be described in the following,

it is advantageous that the middle parts of the wire-shaped conductors have little or no connection to the substrate on which they are arranged. Such sections will hereafter be called connecting sections or C-sections.

There are different ways to achieve this. A method that can be used on an anisotropic substrate is to design the device so that the electromechanical coupling constant is large in the propagation directions for the acoustic surface waves, compared to the value of this constant in directions perpendicular to the C sections.

Alternatively, it will be possible to design the device so that K is zero during the C sections. It is e.g. possible to perform certain piezoelectric ceramic substrates with selected areas without

■piezo-electric coupling.

Another, alternative method is based on mismatched velocity ratios between acoustic surface waves produced during C section. Such mismatch can arise from anisotropy in the crystal or can be achieved by adjusting the space between the threads in the C sections.

Alternatively, the C sections can be applied over pieces of silicon oxide or other non-piezo-electric material - with a low dielectric constant, said pieces themselves being applied to the substrate.

For additional insulation, the low dielectric constant pieces can be applied on top of a metal film on the substrate. This shields the substrate from the electric fields between the wire-shaped conductors.

Those sections of the conductors which only serve as electrical connections will inevitably constitute a capacitive load on the coupler. This extra load can be counteracted by increasing the number of conductors in the coupler, and full compensation is possible by using this technique. However, the relevant load can also be reduced by using pieces of silicon oxide under the relevant section of the conductors, which will further produce the beneficial effect that the connection between the conductors and the substrate is reduced, as stated above.

Regardless of the coupler's length, the symbol L will be used to indicate the length which results in maximum energy transfer from one propagation path to another. In other words, the length hereafter denoted by L will also include any additional length that is necessary in that case, due to the effect of the capacitive load described above. The term "full-length connector" will be used in the following to denote a connector with length L.

It is possible to design a connector so that it supplies energy in

one path is distributed equally between two output paths, and this requires a length of half an L. Couplers designed in this way will in the following be called 3dB couplers.

It is also possible to design a coupler of suitable length to transfer any desired proportion of the added energy to another propagation path. Couplers designed to transfer a share of less than half of the added energy to another path will hereafter be called partial couplers.

Fig. 2 is a plan view of a device comprising a coupler 6, designed to transfer acoustic surface waves from a parallel path A to a nearby convergent path on the same substrate. This coupler is of the same type as the coupler 5 in fig. 1" except that the parts of the wire-shaped conductors that cross path B are given curvatures corresponding to a series of circular arcs with a common center 0. On an anisotropic substrate it may be better to use curves that do not have a circular shape, since acoustic waves will be generated in directions perpendicular to the conductors. : The operation of this device is as follows. Acoustic surface waves emitted along path A from the converter 3 create electrical

fields between nearby wire-shaped conductors in the coupler 6, and these fields are transferred to the circular parts of the conductors. This results in the generation of acoustic surface waves that propagate along path B in a direction perpendicular to the arcs of the circle, so that the produced acoustic surface waves will converge towards the focus at point 0. A fine focus can be achieved at point 0 by suitable selection of curve shapes for the wires over path B One application for a coupler of this type is for the supply of acoustic surface waves to a waveguide (not shown) for such waves and connected to point 0.

The two active areas are connected by means of a

connectors of the type described above do not need to be on the same substrate, as long as the conductors over one area are appropriately connected to corresponding conductors over the other area. Fig. 3 shows a plan view of a device comprising a multiwire coupler mounted to transfer surface acoustic wave energy from one substrate to another. The device comprises a first comb converter 9 attached to a first substrate 11 for propagation of acoustic surface waves, as well as another comb converter 13 arranged on another substrate 15 arranged for propagation of acoustic surface waves. Substrates 11 and 15

are mounted in close proximity to each other (e.g. by attachment to a common base) and a full-length multiwire coupler 17 is formed across the substrates 11 and 15 between the converters 9 and 13. If the substrate 15 is of the same nature as the substrate 11 in all ways,

the distance between the conductors on both substrates must be the same,

but otherwise it may be necessary to arrange different inter-

space and the length b will not be the same over each of the propagation paths on the two substrates.

Fig. h is a plan sketch of a device comprising a coupler 19 arranged to divide the received acoustic surface wave effect between two output paths and with a 90° phase difference. This is a half length or 3dB coupler. A third comb converter 21 is attached to the substrate 1 at the end of the path A which is farthest from the converter 3.

The device works as follows. Acoustic surface waves are emitted along path A by the converter 3« The designation a^ represents the amplitude of these waves. When the waves reach the coupler 19, their energy is distributed with equal values between the symmetric and anti-symmetric modes. The waves will thus propagate as a signal in symmetrical mode and amplitude ia^ plus a signal in anti-symmetric mode and likewise with amplitude -g-a^. The two mode signals start in phase with each other at the leading edge of the coupler 19 in path A. In path B, the signal in anti-symmetric mode will be as large as the signal in symmetrical mode, but have the opposite phase. The length of the 3dB coupler 19 is just sufficient to cause the anti-symmetric mode signal to lag behind the symmetric mode with a phase of TT/2 radians, when the signals reach the trailing edge of the coupler.

The resulting acoustic waves that leave the coupler along path A and path B, respectively, will thus both have amplitudes equal to a^/"V2, as the wave in propagation path B will be ahead of the wave in path A with a phase difference of 17/2 radians. The wave in path A is detected and converted into electrical signals using converter 21, and the wave in path B is detected and converted into electrical signals using converter 7.

Fig. 5 is a plan view of a beam switch for a surface acoustic wave and arranged to produce an output wave along one or the other of two propagation paths, depending on the sign of the present 90° phase difference between the two input signals. This beam switch is similar to the device in fig. <!>+, but is equipped with a fourth cam converter 23, continued - the substrate 1 near the end of the path B which is farthest from the converter 7-

The beam converter works in the following way. It is assumed that the converters

3 and 23 produce signals with amplitudes of respectively a^ and 9- 22' the coupler 19 acts in such a way under the influence of the signal a^ that signals of lower amplitude a^AV 2~ are produced in both path A and path B, - as the signal in path B is ahead of the signal in path A with a phase difference of Tf /2 radians. In the same way, the signal ag-^ will influence the coupler 19 tii to produce signals in path A and path B, both with amplitude equal to ag-^/Y 2^", but the signal in path A will be ahead of the signal in the other path with a phase difference of ff/2 radians. If now the original signals a^ and a^ have the same amplitudes, but are 90° apart from each other, the resulting output signals from the coupler 19 will cancel each other in one or the other of the two paths, depending of whether the signal from converter 3 is in phase before or after the signal from converter 23. The emitted output signal from the converter can thus be switched from converter 21 to converter 7 and vice-versa, by means of reversing the present 90° phase difference between the signals supplied respectively from converter 3 °g from converter 23.

Fig. 6 is a schematic plan view of an acoustic wave width concentrator arranged to produce a narrow beam of acoustic surface waves on the output side. An acoustic surface wave substrate 23 forms two propagation paths A and B of the same width on opposite sides of a line S. A source 25 for acoustic surface waves and of width 2b is arranged on the substrate 33. for the emission of acoustic surface waves both along the path A and the path B. A multi-wire coupler 35 of half length is arranged on the substrate 33 to cover both path A and path B. A receiver device 37 is attached to the substrate 33 in path B on the opposite side of the coupler 35 in relation to the source 25. The device is designed so that the signals that reach the coupler 35 along path A and path B respectively are of the same amplitude, but have a 90° mutual phase difference. This can be achieved in any of four alternative ways, which will now be described.

The first method to ensure a 90° phase difference between the relevant signals is to lower or increase the propagation speed of acoustic surface waves along one of the paths, - by placing a piece of suitable material, e.g. a metal, which has elastic properties different from the corresponding properties of the substrate material, on one of the webs.

The second method consists in letting the source 25 consist of two converters, which are mutually offset in relation to each other by a distance corresponding to a quarter wavelength for the acoustic surface waves.

The third method involves letting the source 25 consist of two converters at the same distance from the coupler 35, but driven by electrical signals with a 90° mutual phase difference.

The fourth method involves making the coupler 35 with a kink on each conductor so that one half of the coupler is effectively displaced a quarter of an acoustic wave length along the propagation direction, as shown in fig. 7 which will be described later.

With each of these methods, it is ensured that the signals which

when the coupler 35 along path B is Tf/2 radians ahead of the signals that reach the coupler 35 along path A. In the case of a mutual influence of the type described with reference to fig. h, the coupler 35 will effectively concentrate the wave energy that is initially propagated both along path A and path B, to form a single wave along path B on the output side of the coupler 3.5-

It will be obvious that several width concentrators can be connected

in cascade one after the other, so that the width of the acoustic . surface wave is halved in each step. Fig. 7 thus shows a three-stage width concentrator which comprises three couplers ^•3?<*>+5 and k- 7. Each of these couplers is provided with a step shift in the middle of each of the coupler's wire-shaped conductors . Block ^-1 represents a source, and block ^9 represents a receiver for surface acoustic waves. The receiver ^-9 can be a coupler or a converter, or possibly a waveguide for acoustic surface waves.

In the case of successive width reductions as described above, i

essentially all the energy from the wide source <*>f1 be concentrated into an orbit with only one-eighth of the width of the source ^1.

This device will work equally well in the opposite direction, as a width expander, if ^9 is a narrow source and *+1 - a wide receiver. The main application for such a device is the matching of acoustic impedances.

Fig. 8 is a plan view of a connecting device designed to act as a branch circuit. Branch circuits are known both for low-frequency equipment (in the form of inductive circuits) and for microwave equipment (in the form of T-shaped wave-conductor connections),

but it is difficult to provide an appropriate and practical embodiment of an electrical branch circuit designed to

work in the widely used intermediate frequency ranges. Branch circuits for acoustic surface waves should be very applicable and appropriate in these frequency ranges, where purely electrical or electromagnetic branch circuits are inappropriate or impractical.

Fig. 8 shows the same components as in fig. 5, except that the half-length coupler 19 is extended with an offset step on it

a quarter-wavelength in the middle part of each of the coupler's conductors, whereby a displacement of half of the coupler 19 is achieved in relation to the other half, with a distance equal to a quarter-wavelength for the acoustic surface waves, so that when waves are emitted in mutual phase along the two paths A and B, the waves in path A will reach the first conductor in the coupler 19 TT/2 radians for the waves in path B.

When signals in the same phase and with amplitudes respectively a^ and a^ are propagated from the converters 3 °g 23, each of the signals a^ and a^ will be split to form signals with the same amplitude in the two paths A and B on the output side of the coupler 19. The phase for the contribution

of the signal a^ to the output signal along path A in a plane P on the output side of the coupler 19, will be taken as reference.- Relative to this signal, the phase of the contribution of the signal' a^ to the output signal along path B will be accelerated by 7T/2 radians by means of the step displacements in the coupler 19, as well as accelerated further

Tf/2 radians by means of the coupling effect. The contribution from the signal ag-^" to path B will be in phase with the reference signal. The contribution from the signal a^ to the output signal along path A will

be reset or delayed 77/2 radians by the step offsets in the coupler, but the acceleration of TT/2 radians effected by the coupler will exactly compensate for this. The output signal in path A is thus a summation of the signals a^

and a2^5 but in path B the signal contribution from the signal a^ will be inverted and the resulting output signal will be the difference

between the signals a-^ and &23* This device will thus form a branch circuit in which the converters 3 °S 23 form the input ports, while the converters 21 and 7 respectively form the summation gate and the difference gate.

Fig. 9 is a plan view of a delay line for acoustic surface waves and which is provided with drains. A substrate 63 for acoustic surface waves has a line of symmetry S between two propagation paths A, B which both have a width b on either side of the line S. A comb converter 65 with teeth that engage each other and a width b is attached to the substrate 63 in an appropriate position for sending acoustic surface waves along path A. A series of partial wave couplers 67a, 67b, 67c of the same type as the coupler 5 in fig. 1" but with a smaller number of conductors, is placed on the substrate 63. A comb converter 69 is attached to the substrate 63 in the propagation path A on the opposite side of the couplers 67a, 67b,

67c in relation to the comb converter 65. Other comb converters 71a, 71b, 71c of the same type as those mentioned above are arranged on the substrate 63 in path B beyond each of the couplers 67a, 67b, 67c, counted from the converter 65. Some pieces of suitable acoustic absorbing material 72a, 72b, 72c are placed in path B between the switches 71a, 71b -, 71c.

The device works in the following way. Acoustic surface waves

which is emitted along path A by means of converter 65 is received by converter 69 after a delay corresponding to the time (fet takes for acoustic surface waves to travel along path A between converter 65 and converter 69, in the same way as with a conventional acoustic delay line surface waves As the acoustic surface waves pass the couplers 67a, 67b, 67c, however, each coupler transfers part of the received wave energy to path B, so that this transferred energy will be detected and transformed into an electrical output signal from the nearest of the transducers 71a, 71b, 71c. The absorbing pieces act to reduce arbitrary signals. The remaining energy in path A will produce a signal at the farthest converter 69. The lengths of the propagation path between converter 65 and each of the converters 71a, 71b, Tic determine the relevant delay periods,

Fig. 10 is a plan view of an alternative embodiment of a tapped delay line for surface acoustic waves. This tapped delay line is similar to the delay line in Fig. 9, except that the sub-couplers 67a, 67b, 67c with straight conductors are replaced by sub-couplers 73a, 73b, 73c with deflected conductors to direct the output waves from the respective couplers at a certain angle to path A. The output switches 75a, 75b , 75c are arranged to receive the output waves from the couplers 73a, 73b, 73c respectively. This arrangement provides a means of reducing the acoustic wave energy reflected by the transducers, which can form arbitrary signals in earlier transducers. Couplers of the type described here can also be used as mode discriminators, since they are very sensitive to acoustic surface waves, but relatively insensitive to waves that are propagated inside the material. If, for example, the converter 3 in fig. 1

■produces unwanted acoustic waves inside the material, the coupler 5, which has a full length, can be used to simply isolate the acoustic surface wave, which will be transferred to path B, while the wave that propagates inside the material will be relatively unaffected by the coupler 5 and continue along path A. Such couplers can also be used in a similar way to discriminate between different acoustic surface wave modes that can occur when the acoustic waves propagate in a thin material film on a substrate of another material.

Fig. 11 is a plan view of a broadband coupler 79 which serves as a path changer and is placed on a substrate 77 for acoustic surface waves. The path changer 77 is made of several J-shaped conductors, arranged within each other in such a way that all conductors run straight and mutually parallel at one end which constitutes a first propagation path A for acoustic surface waves, while all the conductors are likewise straight and parallel to each other as well at the other end, which forms another propagation path B for surface acoustic waves. The length of the path exchanger: over both surface acoustic wave paths is L. The two propagation paths of surface acoustic waves are mutually parallel, but due to the curvature of the conductors and the fact that they are arranged inside each other,

the order of the conductors in one lane will be reversed in relation to the order in the other lane. If the substrate 77 is constructed of

an anisotropic piezoelectric material, it will be possible to design the device so that the direction of the conductors in all other parts of the path-changing coupler 79, where the conductors are mutually parallel, is such that the direction perpendicular to these parts corresponds to the piezoelectric inactive direction , so that no significant acoustic surface wave will be able to propagate in the relevant direction. The operation of the path-changing coupler is as follows. Acoustic surface waves incident on the path exchanger 79 in the first propagation path A for acoustic surface waves cause the creation of electric fields between the adjacent conductors. These fields are transferred from the aforementioned first propagation path A to the second propagation path B for acoustic surface waves. Since the alignment of the conductors over the aforementioned two propagation paths is mutually opposite due to the design of the conductors that make up the path exchanger 79, the surface acoustic wave emitted along the second propagation path B will travel in the opposite direction to the original surface acoustic wave in the first lane A. The coupler 79 can thus be used to transfer energy from lane A to lane B as indicated by dashed arrows in fig. 11, or in the opposite direction as indicated by fully drawn arrows.

Fig. 12 is a plan view of an alternative path changer for acoustic surface waves and which includes an alternative design for preventing the emission of acoustic surface waves from the parts of the path changer located between its active end sections. The substrate 77 is glass or another elastic, non-piezo-electric material, on which the J-shaped conductors which make up the coupler 79 are placed. A thin film 81 of piezoelectric material such as e.g. zinc oxide is sprayed or otherwise applied over the conductors at the end of the path exchanger 79 where they cross propagation path A, while a thin film 83 of zinc oxide is likewise sprayed or otherwise applied over the conductors at the end of the coupler 79 where the conductors cross path B It ef only in the areas covered by the piezo-electric films 81

and 83 as a connection can be established between acoustic surface waves and electric fields, so that surface acoustic waves are only emitted and detected in these areas, while the other parts

of the conductors in the coupler 79 only serve as electrical conductors.

Fig. 13 is a plan view of a surface acoustic wave resonator or recirculating delay line, comprising two path-changing couplers which will hereafter be referred to as path changers. The two path exchangers 85 and 87 of the same form as described above with reference to fig. 11 and 12, a substrate 89 is attached in such a way that the two acoustic propagation paths C and D that are connected by the path changer 85 are the same two propagation paths that are interconnected by means of the path changer 87.

A sub-connector 8k- with several conductors is arranged for connection of track

C to another path E, on which two converters 86 and 88 are placed on opposite sides of the coupler 8h.

The device works as follows. The energy in the acoustic surface waves emitted from the converter 86 is partially transferred to path C by the coupler 8<>>+. The acoustic surface wave which thus propagates along path C is connected to path D by means of path changer 87, and also connected back to path C by means of path changer 85. The device constitutes a resonator with an oscillation period equal to the combined delay in path C and path D , as the signal supplied to the loop formed by paths C and D together with the path exchangers 85 and 87 can be propagated around the loop a greater or lesser number of times. Each time the signals pass the coupler 8<*>f a proportion of their energy is transferred to the converter 88 by the coupler'8^. It should be noted that it is the short length of the partial coupler 8^ which makes the device act as a resonator. If this partial coupler were replaced by a full-length coupler, the resonator would become a delay line, since all the wave energy originally emitted from the converter 86 would be returned to path C by the path changers and then completely

. extracted by coupler 86 after a single propagation along the delay line. Fig. 1<*>f is a plan view of a f orzincing line "for acoustic surface waves with couplers forming an angle. A substrate 91 for acoustic overfite waves is attached to three angle couplers 93? 95 and 97 in full length and placed at each of one

equilateral triangle, the whole being arranged in such a way that each angle coupler will receive acoustic surface waves from one of the other angle couplers, as well as forward said waves in the direction of the

third angle coupler. The connecting parts for the angle couplers 93,

95 and 97 are mounted over pieces of silicon oxide, respectively 9<*>+, 96 and 98, in order thereby to minimize the coupling between the substrate 91 and the parts of the couplers 93, 95 and 97 which are not used for receiving or sending acoustic surface waves. In itself this coupler arrangement forms a surface acoustic wave resonator of triangular shape, but a fourth angle coupler 99 is provided for the supply of surface acoustic waves

to the delay line and connection of such waves from the line. The angle coupler 99 is fed by a first cam converter 101 and feeds

even another cam converter 103.

The operation of the device is as follows. Acoustic surface waves emitted from the comb converter 101 are received by the angle coupler 99 and fed by this into the triangular circuit. The bolts are received by the angle coupler 93 and transferred by this to the angle couplers 95 and 97 in sequence. Acoustic surface waves emitted from the angle coupler 97 are again received by the angle coupler 99 and emitted from this in the direction of the comb converter 103. In this way, a delay line is obtained with a delay time equal to the total path length between the converters 101 and 103 via the angle couplers 99,

93, 95, 97 and back to 99. It should be noted that the coupler 99

which makes a strong coupling due to its full length, acts so that the device acts as a delay line.

If this connector was replaced with a part connector that is not

of full length, the delay line would become a resonator in which each added signal could travel several times around the circuit.

When using angle connectors., longer folded delay lines can also be arranged on relatively small strips of material.

Fig. 15 is a plan view of such a folded acoustic delay line. In this figure, the angle couplers used are not shown and only the folded path for the acoustic surface waves is indicated, as this is drawn on a much smaller scale than in the other sketches. The track is made up of a series. triangular paths each overlapping the neighboring path to a certain extent, thereby achieving

a long overall path length on a relatively small substrate.

Fig. 16 is a plan view of a broadband, unidirectional converter,

for acoustic surface waves. A cam converter 105 and a .

coupler 109 for acoustic surface waves is attached to a piezoelectric substrate 107. The coupler 109 is made of several U-shaped conductive wires with long parallel sections at their outer ends, the wires being placed within each other so that the coupler 109 itself becomes U- shaped. The cam converter 105 is placed between the side branches of the U-shaped coupler 109, so that said long parallel sections of the conductors that make up the coupler 109 are of the same length and arranged parallel to the teeth in the comb converter 105. The coupler 109 is placed so in relation to the center line through the comb converter 105 that surface acoustic waves emitted by the converter 105 and traveling in mutually opposite directions will reach the inner edges of the innermost conductors in the coupler 109 with a mutual phase difference of 90°. This can be achieved by positioning the comb converter 105 in such a way that its teeth are placed symmetrically about a line which is one-eighth acoustic wavelength offset to the side of the axis of symmetry of the coupler 109. Each side of. the coupler 109 has a width equal to half the transfer length L, which means that the coupler 109 is a folded coupler of half the transfer length.

The device works as follows. The acoustic surface waves that are emitted from the converter 105 in both directions when the innermost wires of the coupler 109 are in phase with each other by 90°, so that the coupler 109 works in the same way as the converter 35 in fig. 6. Acoustic surface waves will thus only be emitted from the one straight section of the coupler 109. The converter will thus have a directional effect and only emit signals from the side of the U-shape that receives the signal that is first in phase.

Fig. 17 is a plan view of an alternative embodiment of a broadband, unidirectional converter for surface acoustic waves. Like the converter described above with reference to fig.

16, a comb converter 105 and a coupler 111 for acoustic surface waves are attached to a substrate 107 for such surface waves. The coupler

111 is made up of several oblong U-shaped conductors with long parallel parts along opposite sides, the relevant conductors being arranged within each other, so that the coupler 111 itself is also 0-shaped. The converter 105 is placed inside the coupler 111 like the converter 105 in fig. 16 was placed between the branches in the coupler 109.

This device works in the same way as the device in fig. 16, but each of the straight sections of a conductor in the coupler 111 is connected to the corresponding straight section on the opposite side of the coupler 111 over two wire sections instead of one. This creates current paths with less resistance and reduces the adverse effects of an unwanted break in a conductor. A disadvantage, however, is that larger conductor lengths are required, and this entails a greater capacitive load on the coupler 111.

Connectors with directional action of the type shown in fig. 16 or fig. 17 can advantageously replace the simple comb converters shown in several of the devices described above, and can e.g. are used instead of the converters 3, 23, 21 and 7 in the branch circuit in fig. 8.

Fig. 18 is a plan view of a wave reflector 113 for. acoustic surface waves and applied a substrate 115 for such waves. The reflector is made up of a folded 3dB coupler, substantially similar to the coupler 109 in fig. 16, except that there is no gap between the two branches of the relevant U-shape. The way this reflector works is as follows. It can be considered as a half-length coupler (of the same kind as the coupler 19 in fig. h) which is bent back towards itself. Acoustic surface waves incident on the half-length coupler 19 along path A produce two acoustic output signals of the same amplitude, but with a 90° phase difference. In the coupler 113, these two bolts will each be supplied by the opposite branches of the U-shape. The coupler will therefore effectively find itself in a similar situation as the coupler in fig. 5, as its two halves receive equal signals with a mutual phase difference of 90°. An output wave is thus emitted from only one half-side, and in the folded version in fig. 18 the output wave is reflected in the opposite direction to the incident wave. The device therefore acts as an effective reflector for acoustic surface waves.

Fig. 19 is a plan view of an alternative embodiment of a path changer for acoustic surface waves. A substrate 119 for acoustic surface waves has a line of symmetry S between two mutually adjacent propagation paths A and B, both of which have a width B. A half-length coupler 117 is placed across the two paths A and B, and two reflectors for surface acoustic waves 121 and 122 of the type shown in fig. 18, lane A and lane B are connected respectively, both on the same side of the coupler 117.

The operation of the lane changer in fig. 19 is a combination of the effects described in connection with fig. fig. 5 and fig. 18. When a surface acoustic signal reaches the coupler

117 in lane A, will the coupler 117? which works in the same way as the coupler 19 in fig. ^, effectively divide the incident energy between two waves that propagate from each half of the coupler 117 on its output side. The two reflectors 121 and 122 throw back these two waves to the respective halves of the coupler 117. The coupler 117 will now be in a similar situation to the coupler 19 in fig. 5? which receives signals with a mutual phase difference of 90°, and therefore emits an output signal in the path that receives the signal with a leading phase. The signal received by the path changer 79 in path A is thus effectively thrown back along path B and the device can equally work in the opposite direction, whereby a surface acoustic wave signal from path B is reflected along path A. The device thus actually forms a path-changing reflector.

Fig. 20 is a plan view of another form of unidirectional converter for acoustic surface waves. A comb converter 123 of width •5-b and which is attached to a substrate 12h for acoustic surface waves, is connected to a propagation path A for such waves by means of a coupler 125 of full length. The coupler 125 is made up of several J-shaped conductors, each of which has two straight lines parallel to each other

arms of different lengths. These conductors are placed within each other

in such a way that the coupler 125 itself takes on a J shape. The shortest arm of the J-shape runs parallel to the teeth of the cam converter. 123, but does not reach track A. Track A has a width of b and does not overlap converter 123. The longest arm in the J shape extends across the entire width of track A at right angles to it. The converter 123 is positioned so that acoustic surface waves emitted from the converter 123 in opposite directions will reach the innermost conductor in the coupler 125 in the same phase.

The coupler 125 is essentially similar to the basic coupler 5 in fig. 1, except that one half is folded back on itself. Although the arrangement looks different as far as the operation of the coupler is concerned, it is excited in the same way as the coupler in fig. 1, and as the present coupler is of full length, it transfers substantially all of its input energy to the other side of its other half. The signals from the coupler's outer conductor are therefore only transmitted to lane A.

The coupler in fig. 20 can just as well be used for receiving acoustic surface waves incident along path A towards the outermost conductor of the coupler 125. The converter .123 will not receive acoustic surface waves incident towards the innermost conductor in the coupler 125. The unidirectional sensitivity of this arrangement makes it applicable in many devices.

If the converter 123 is omitted, the coupler 125 will act as a reflector for acoustic surface waves incident on the outer conductor of the coupler 125, but it will tend to split the surface waves incident on the inner conductor of the coupler 125 in two wave trains that are propagated in mutually opposite directions from the folded part of the coupler.

The curved parts of the coupler 125, which do not cross the path A or the path of the acoustic surface waves emitted or received by the converter 123, are preferably extended as connecting sections in the manner indicated above.

Fig. 21 is a plan view of a delay line for surface acoustic waves' and provided with drains. A substrate 126

for surface acoustic waves is provided with a comb converter 127 arranged for emitting surface acoustic waves along a path A of width b. The substrate 126 is also provided with several delay line taps comprising several couplers, of which only one coupler 128 is shown. The coupler 128 is a modified embodiment of the coupler 125 described above with reference to fig. 20, as it deviates from the former links on two essential points which are of particular importance fa? present application. First, the coupler 128 is positioned to transfer a certain amount of energy to a comb converter 129 of width b (instead of width -g-b) and which is located in a propagation path B

parallel to and close to path A. Second, coupler 128 is a partial coupler. The parts of the coupler that do not cross either lane A or lane B are laid out as connection sections in the manner described above.

At the delay line in fig. 21, acoustic surface waves emitted along path A are received by the converter 127 according to different

delay times at different tapping points along the line, e.g. the tapping which includes the coupler 128. Each tapping 3'can only extract a small part of the energy that is propagated along the line, so that sufficient wave energy remains for subsequent extraction with subsequent tappings. ■ The coupler 128 can be thought of as a folded coupler which, if straightened, would be equivalent to a coupler 128' for connecting waves from a narrow path A to a path B' twice as wide as path A. These are shown in fig. 22. Such a switch cannot be made to achieve complete energy transfer, but it should be remembered that taps from a delay line are not designed to extract all the energy propagated along the line.

However, it is possible and appropriate to extract a signal with a level of e.g. 20 dB less than the level of the signal emitted from the converter 127. The signals over the entire width of the wide path B<1> (fig. 22) are emitted in mutual phase towards the converter 129, so that this converter 129 will receive a signal 3 dB higher than it would have received from a simple coupler of the same length, as indicated in fig. 8. This arrangement is also advantageous because it makes it possible to extract the desired amount of energy by means of a connector with fewer conductors than the respective comparable simple connector.

It would be possible to extract more power from converters of the same type as converter 129, by tuning in a known manner by means of a series inductance to create a series resonant circuit,

as the converter in itself constitutes a capacity. This extra power draw occurs at the expense of power that is not normally absorbed by the converter, and not at the expense of the power that is propagated along the delay line.

The drains along the delay line may be designed to enable the transducers to be deflected away from the wave front for the surface acoustic waves traveling along the delay line,

in the same manner as angle taps for delay lines described with reference to fig. 9.

An important property of delay line taps such as the combination of coupler 128 and converter 129 is that they are unidirectional and will only be affected by surface acoustic waves traveling in a particular direction. This makes delay line taps of this form particularly applicable for folded delay lines and similar devices in which a signal or an unwanted reflection can be thought of as being thrown back past a tap, for example if it is desired to add further taps to the delay line which will be described later with reference to fig. 2h.

•Fig. 23 shows a plan view of a surface acoustic wave delay line including means for suppressing reflected third pass signals. In the case of a delay line made up of a single signal emitter, a delay medium and a signal receiver, most of the signal transmitted to the delay medium by the signal emitter will be absorbed by the signal receiver. Experience shows, however, that a small proportion of the signal that falls on the signal receiver can be reflected and a part of this reflected signal can reach back

to the signal transmitter. A small proportion of the signal incident on the signal emitter can in turn be reflected, and this signal can be received once more by the signal receiver. Since this signal will have passed through the delay medium three times it is termed a triple transit fe signal, and although its power level will be low compared to the originally received signal, it may be of sufficient speed to form annoying, unwanted echo signals.

In fig. 23 is a surface acoustic wave substrate 131 provided with two parallel, adjacent paths A and B. A comb converter 133 is arranged for sending surface acoustic waves along

path A. A 3 dB coupler 135 is placed across both paths A and B in the propagation path of surface acoustic waves emitted from the transducer 133. An absorber 137 for surface acoustic waves is placed in path B on the same side of the coupler 135 as the transducer 133• A comb converter 139 is placed at the end of path A which is farthest from converter 133• A comb converter 1<!>+1 which is in all respects similar to converter 139, is placed in path B at exactly the same distance from coupler 135 as the converter 139-The converter 139 is electrically connected in series with an inductance 1139 and a load resistance R139, while the converter 1<>>+1 is electrically connected in series with an inductance LI^-1 which is identical to the inductance L139 and a resistance R1 <*>f1 which is identical to the load resistor R139-

The device now works in the following way. Acoustic surface wave signals are emitted along path A by the converter 133- These signals are distributed equally between path A and path B, using the 3 dB coupler 135? and the output signal is taken from the load resistor R139.

The 3 dB coupler 135 thus emits signals with the same amplitude, but

with a mutual phase difference of Tf/2 radians, and these signals will fall on the converters 139 Cg 1<*>+1. Since converters 139 and 1^1 are electrically identical, any signal reflected by converter 139 in path A will have an exact counterpart in a signal reflected by converter 1*f1 in path B. The signal in path B will retain its phase at TT/2 radians ahead of the signal in path A, and the effect of the 3 dB coupler 135 on these two reflected signals will thus be to combine them and feed them back

along path B, where they will be absorbed by the absorber 137.

The converters 139 and 1^1 can advantageously be designed as unidirectional converters, of the type that has previously been described with reference to fig. 16, fig. 17 or fig. 21. As long as the converters 139 and 1<*>f1 are mechanically and electrically identical, however, in any case all unwanted reflections will be directed towards the absorber 137.

Fig. 25 is a plan view of a reflective delay line for surface acoustic waves, and which can be used to double the available delay time at a given substrate length for surface acoustic waves.

A substrate 151 for surface acoustic waves comprises two parallel, adjacent propagation paths A and B. A comb converter 153 is arranged for sending surface acoustic waves along path A. A comb converter 15<*>+ is placed in path B

near comb converter 1 53 • The working function of the converter 1 5*+ is to extract the delayed acoustic surface wave. A 3dB coupler 155 is placed across paths A and B near the converter 153 and the converter 15<*>+. A path changer 156 of the type described above with reference to fig. 19, is placed across path A and path B at the end of the substrate 151 which is farthest from the converters 153 and A^ h. An absorber 157 for surface acoustic waves is placed in path B near the coupler 156 and at the side which is farthest from the converter A reflector 158 of the type described above with reference to fig. 18, is placed in path B near the absorber 157 and at the end furthest from the converter 15^.

The operation of the device is now as follows. The signal is emitted along path B by means of the converter 153• By the action of the 3dB coupler 155, the signal energy is divided to produce equal signals along path A and path B. The energy thus transmitted along path B is absorbed by means of the absorber 157- The signals along path A is connected to path B by the effect of the path changer 156, is reflected by means of the reflector 158, is connected back to path A by means of the path changer 156, and is transferred again to the coupler 155 - With the help of the coupler 155, the returned signal energy is divided to path A again between path A and path B, so that the signal energy is received by the converter 15*+. Since the signal energy is halved twice, there will be an inherent loss of

6dB in this device. An alternative is to use a pair of unidirectional converters and omit the coupler 155- Unidirectional or bidirectional taps (not shown) can be placed along track A if desired.

Fig. 25 is a plan view of a reinforcing web vector, which can be considered as an improvement of the web vector described above with reference to fig. 19.

A surface acoustic wave substrate 161 comprises two parallel, adjacent propagation paths A and B. A 3dB-

coupler 163 is placed on the substrate 161 across the two lanes A and B, each lane A and B being provided with a reflector comprising a unidirectional converter of the type described above with reference to fig. 16. The reflector 165 in path A comprises a U-shaped acoustic surface coupler 171, which partially encloses a comb converter 169 which is electrically connected in series with a tuning inductance 173 and a device 175, with negative resistance. This device with negative resistance can be of conventional design bg negative resistance characteristic, e.g. a tunnel diode circuit.

The reflector 167 in path B is constructed and connected in a similar way. The reflectors 165 and 167 are at the same distance from the coupler I63. The device with negative resistance and the tuning inductance are also identical for the two reflectors.

The mode of operation of the reinforcing path vexiser will be similar to the mode of operation of the simple path vexiser as described above with reference to fig. 19, except that the converters and their connected devices with negative resistance amplify the reflected signals.

Fig. 26 is a plan view of a directional filter. The directional filter is known from microwave technology and is available in resonator form. A known embodiment of a directional filter for micro-bubbles is constructed in the following way. A microwave source is connected to a first matched load via a first directional coupler. The first directional coupler is in turn connected to another directional coupler through a circulating cavity. This other directional coupler is connected to a cable conductor which is connected to another adapted load. If it is assumed that the source has a broadband output signal, the frequency spectrum of the output signal from the first matched load will have a series of depressions separated by frequency differences that depend on the phase shift in the circulating cavity, because these frequencies are those most easily absorbed by the circulating cavity. These frequencies will appear as peaks in the frequency spectrum on the output side of the second adapted load. The width of said depressions and peaks is determined by the losses in the circulating cavity and in the two directional couplers. Fig. 26 shows a plan view of an analogous device for acoustic surface waves.

In fig. 26, a substrate 181 is provided with four parallel, adjacent paths A, B, C and D in the aforementioned order. A comb converter 183 is arranged for sending out acoustic surface waves along path A. A coupler 185 is placed across paths A and B, in such a way that some of the energy emitted by the converter 183 is propagated along path B. The rest of said energy, which is transferred along path A, falls on a comb converter 187. Paths B and O are provided with reflective path exchangers 189, 191 at each end, whereby energy in path B is transferred by the path exchanger 191 to path C, and energy in lane C is transferred by lane changer 189 to lane B. A further coupler 193 is placed across lanes C and D, close to coupler 185? whereby part of the wave energy that is propagated along path C is transferred to path D. Single-comb converter 195 is arranged to receive the energy that is transferred to path D.

The operation of the device is now as follows. Paths B and C together with the reflective path exchangers 189 and 191 form a resonator in which the acoustic surface waves can circulate, and energy connects into and is taken out of this resonator respectively. using couplers 185 and 193- The resonator will have a range of resonant frequencies determined by the phase delay in the complete circuit

which is formed by the paths B and C. These resonant frequencies will build up in the resonator far more strongly than the other frequencies that are supplied, so that the output signal from the coupler 193, and which is taken out with the help of the converter 195, will have a frequency spectrum with a series of peaks at mentioned resonance frequencies. The output signal from converter 187 will make up the rest of the output signal from converter 183, and the frequency spectrum will thus in this case have a number of depressions at the resonance frequencies.

The degree of coupling into and out of the circulating resonator is adjusted during the construction work by setting the lengths of the couplers 185 and 193 to satisfy the desired coupling conditions.

Fig. 27 is a plan view of a variable direction switch. A substrate 201 for acoustic surface waves is provided with two parallel, adjacent propagation paths A and B. Two converters 203 and 205 are placed close to each other at one end of the substrate 201, converter 203 being placed in path A and converter 205 being placed in path B Two further converters 213 and 215 are placed at the opposite end of the substrate, whereby converter 213 is placed in path A and converter 215 in path B. Two couplers 207 and 209 are placed across both paths A and B, and an area 211 with controllable propagation speed for the acoustic waves are formed over a •part of path A between the couplers 207 and 209. The area 211 can e.g. is made up of an area that includes a material with electrically or magnetically controllable piezoelectric or electrostrictive properties, whereby the propagation speed of the acoustic surface waves can be set by varying an electric or magnetic bias field, or by other appropriate external control.

The operation of the present variable directional coupler is as follows. The area 211 can add a controllable phase delay to the acoustic surface waves in path A, in relation to the signals propagated along path B, between the coupler 207 and the coupler 209.

If both couplers 207 and 209 are 3dB couplers, and area 211

is inactive, the entire power emitted along path A by the converter 203 will be supplied to the converter 215, while the entire power emitted along path B by the converter 205 will be supplied to the converter 213. By introducing a phase shift of Tf radians in path B by controlling the area 211, the entire power emitted along path B by the converter 205 will be supplied to the converter 215 in path B. Phase shifts of less than Tf in the area 211 will produce a directional coupler effect between these extremes. Anti-symmetric and symmetric modes have been mentioned in connection with fig. 1. Since the next devices to be described utilize these modes and the effect of the modes under a series of wire-shaped conductors, it will be appropriate here to return to this feature of the operation of such devices. Fig. 28 shows a circuit diagram of a transducer arrangement for generating surface acoustic waves either in a purely symmetrical mode or in an anti-symmetrical mode. This arrangement comprises two identical comb converters 216, 217 placed end to end, for the transfer of acoustic surface waves along adjacent, parallel propagation paths. The converter 216 is connected directly via electrical connection terminals 218. The converter 217 is connected to the terminals 218 through a switch 219 - Although the switch 219 is shown as a conventional switch, it is preferred in practice to use an electronic switch, which can conveniently be made of an integrated circuit . With switch 219 in the position shown, a signal applied to terminals 218 will excite both converters identically and output a signal in symmetrical mode. With the switch in the opposite position, the inverter 217 will be energized

in opposite phase in relation to the converter 216, so that an aati-symmetric wave will be emitted along the two adjacent, parallel paths.

Fig. 29 shows an alternative arrangement for the emission of acoustic surface waves in anti-symmetric mode by means of a pair of equals. comb converters 220, 221 arranged end to end on a substrate 222, so that surface acoustic waves are emitted along nearby parallel paths. The used interconnection of the comb electrodes in question in mutual engagement ensures that a signal supplied to said converter pair will excite acoustic surface waves which will be mutually opposite in phase, so that, in other words, a signal is formed in anti-symmetrical mode.

The line AD2 forms a graphic representation of the amplitude progression

for the signal in anti-symmetric mode across the adjacent parallel propagation paths. Fig. 29 also shows a number of mutually spaced wire-shaped conductors, which are placed over the propagation path of the anti-symmetrical signal and directed perpendicular to the propagation direction of the signal.

It is possible to choose the material and dimensions of the wire-shaped conductors so that signals in symmetrical mode will propagate under the row of wires at the same speed as they travel along an open part of the surface of the substrate 222.• This is achieved by arranging the wire-shaped conductors so that the short-circuiting effect exerted by the width of each conductor wire on the piezoelectric fields in the propagation direction is compensated by mass loads.• Signals in anti-symmetric mode, however, will always travel more slowly as they cause the creation of currents along the wire-shaped conductors across the width of the propagation path, and the conductivity of the relevant conductors in this direction substantially reduces the material's effective piezoelectric stiffness, so that signals in anti-symmetrical mode are retarded.

Figs. 30 and 31 schematically show a device for dividing signals in anti-symmetrical mode and which comprises a number of parallel wire-shaped conductors 22^, designed to unweave the speed adaptation mentioned above. The first manager

is the longest and each successive wire is slightly truncated at both ends, so that the profile of the wave splitter forms an isosceles triangle placed symmetrically on the propagation path of the acoustic surface waves. When this structure is fed a signal in the anti-. symmetric mode, the part of the surface acoustic waves that travel under the said structure will have their propagation speed lowered in relation to the parts of the surface acoustic waves that travel in the material outside the structure. The effect of

said structure is thus to deflect the acoustic surface waves

in two rays away from the original direction of propagation for the waves. This is shown in fig. 30. However, when, on the other hand, symmetrical waves are applied to the structure, no velocity variation will take place, and surface acoustic waves of this nature will thus not change direction. This is shown in fig. 31• The line AD2 in fig. 30 is a graphical representation of the amplitude progression for the added anti-symmetric waves across the width of the track. The line AD1 in fig. 31 is a corresponding graphical presentation of the amplitude progression for the added waves in symmetrical mode. In this wave-splitting device, the path for the surface acoustic waves is not completely determined by the location of the surface acoustic wave components on the surface of the substrate, as the path is also controlled electronically by the signal supply to the converter or pair of converters in question that serve to emit the surface acoustic waves.

Unfortunately, however, any structure that introduces a propagation velocity discord is likely to cause reflections of surface acoustic waves, which can give rise to "random" signals. Random signals may, for example, arise due to reflections from coupler ends or similar acoustic wavelength structures.

Figs 32 and 33 are plan sketches of fitting sections for couplers, and intended to reduce random reflections.

In fig. 32, the front conductor 225 in a coupler or waveguide row 223 is V-shaped and symmetrically arranged about the line of symmetry between the two coupled propagation paths for surface acoustic waves. The angle 9 that the branches of the V-shape form with a line perpendicular to the propagation direction of the acoustic surface waves is given by the formula tg 9 = ?\/ w, where }\ is the wavelength of the acoustic surface waves along the substrate and w is half of the coupler's width. Subsequent conductors are also V-shaped, but form angles 9 which successively decrease towards zero. The distance d between the angle points in the V-shapes can be the same as the mean distance between the conductors in the main part of the coupler. The operation of this connector is as follows. The first conductor 201 will not be connected to the coil to a significant extent. Successive V-shaped conductors connect to an increasing degree as the angle 9 decreases, until the conductors become straight when 9=0. The coupling strength for the conductors near the front edge of the coupler thus decreases steadily towards zero. If a sufficient number of V-shaped conductors are placed between the first conductor 225 and the first straight-fault conductor 226, this will result in an acceptably low level

for said resulting echo signals. A similar conductor pattern can also be used at the rear edge of the coupler.

In fig. 33, the conductors near the leading edge of the coupler are increasingly truncated, while still remaining symmetrically arranged about the line of symmetry between the two coupled propagation paths for acoustic surface waves. Any one of these truncated conductors shall now be considered. The conductor will be shorter in length than one of its neighboring wires due to a slight shortening at each end. Each truncation gives a reflection equal to the reflection fraction that can be expected from a comparable untruncated, single coupler. The reflected acoustic surface waves are the vector result of the small reflections in question. It is possible to design the truncation steps so that the resultant of the reflections is minimized at a selected frequency (or several frequencies) below the stopband's

•limit frequency.

Usually, the shortening steps will be arranged symmetrically, so that half the length of the shortest conductor has the same extent as the shortening steps between each end of the longest conductor and its neighboring conductor, and so on. For a three-step transition, (not shown) on each side of the coupler, the length of the shortening steps will be given by x, y and x respectively, as

where w is the width of the coupler. For a four-step transition (as shown in Fig. 33), the length of the steps in each of the . propagation paths given by p, q, q and p, respectively, idet

A third method for reducing the reflection from a coupler involves setting the width and location of a sufficient number of conductors at each end of the coupler. Each conductor in the coupler can be regarded as a separate reflecting element, and by setting the position and width of the conductors, it will be possible to achieve such relative phase shifts and amplitudes for the reflections from each of the conductors that their vectorial resultant becomes sufficiently small over the desired bandwidth.

Fig. 3^ is a perspective sketch of a light-controlled acoustic surface wave coupler. Three converters 227, 229, 239 and a multi-wire coupler 233 are placed on a substrate 231 for acoustic surface waves, in the same way as in the device in fig. k. The coupler 233 is, however, a multi-wire coupler of full length and extended over a portion 235 of the substrate 231 beyond the outer edge of the propagation path B. Photoconductive material 231 is deposited by vapor deposition or other processes on the portion 235 of the substrate 231 either before or after the coupler 233 has been brought into place.

The operation of this device is now as follows. If the photoconductive strip 237 is not illuminated, the coupler 233 will work in a similar way to the coupler 5 described above with reference to fig. 1. If, however, said photoconductive strip is illuminated, the conductors in the coupler 233 will be short-circuited, and their coupling effect will thus be cancelled, so that part of the acoustic energy emitted from the converter 227 is received by the converter 229. In this way, the amount of energy received can of the converter 229 is controlled by the amount of incident light towards the strip 237, so that the output energy from the converter 229 can be used as a measure of the strength of the incident light towards the strip 237.

Fig. 35 is a plan view of an electrically controlled coupler device for acoustic surface waves. This device differs from the device in fig. 3<*>+ in that the photoconductive strip 237 is replaced by an electrical control device 2^1. The electrical control device 2^1 can e.g. is made up of a series of P-I-N diodes, or several bipolar transistors or field-effect transistors. The device must be capable of electrically connecting the conductors in the coupler 233 under the control of an electrical signal.

Fig. 36 is a connection diagram of a possible design of the control device 2h^ , and Fig. 37 shows a perspective sketch of an integrated circuit corresponding to the circuit diagram in fig. 36. The individual conductors 2<>>+3, 2^-5, 21+7...2!+9 in the coupler 233 are each separately connected to the source electrodes of several MOS transistors, 251, 253, 255 respectively. ..257. The MOS transistors 251, 253, 255...257 have their drain electrodes interconnected and connected by a grounded return connection, while their gate electrodes are collectively connected to a terminal 259. By these means,

can a suitable voltage on the terminal 259 control the transistors and effectively, interconnect and ground all conductors in the coupler 233.

The physical arrangement of the device 2^-1 shown in fig. 37, all conductors 2^3, 2V?, 2<1>+7...2<l>f9 have an insulating layer 260 applied to

a semi-conducting substrate 261, each conductor, such as e.g. 2^3, forms contact with an assigned, heavily doped region, such as e.g. 263, on the substrate 261. A single grounded conductor electrode 265 makes contact with a heavily doped area 267 of the substrate 261.

A film 269 of insulating oxide is applied over the ends of the conductors 2V3, 2^5, 2^7...2M?, the edge of the electrode 265 as well as the intermediate area, while a metallic strip electrode 271 is applied over the film 269.

The operation of the device corresponds to the operation of a conventional MOS transistor. A control voltage of the correct polarity applied to the metal strip electrode 271 forms a low-impedance connection between the conductors 2V3, 2<>>+5, 2V7...2^9 and the grounded electrode 261, whereby the conductors are grounded to suppress the coupling effect.

Fig. 38 is a plan view and fig. 39 a coupling core for an alternative, electrically controlled coupling device for acoustic surface waves. This device differs from the device in fig. 35 only in that the substrate 231 is provided with an area 273 close to and parallel to the propagation path A for acoustic surface waves along the side facing away from the propagation path B. The areas

273 and 235 comprise several diodes with variable capacitance, and each conductor in the coupler 233 is connected between a diode with variable capacitance in the area 235, and a diode of the same type in the area 273 and connected with the same conductor direction. The terminals for the diodes with variable capacitance in the area 235 outside the coupler 233 are connected with a common terminal 275, and the terminals for the diodes with variable capacitance in the area 273 and outside the coupler 233 are connected with a common terminal 277.

The operation of this device is now as follows. When pressed

and variation of a voltage between the terminals 275 and 277, the capacity between the conductors in the coupler 233 and the terminals 275 and 277 can be varied, whereby the capacities between the conductors in the coupler 233 itself will vary. This change in the capacity between the conductors will necessarily change the connection between the conductors, whereby the proportion of energy received by the converters 229 and 239 varies in a controlled manner.

Claims (21)

1. Device for acoustic surface waves and which comprises a first and a second area of a material in which acoustic surface waves can be propagated as well as a coupling device composed of several wire-shaped conductors; characterized in that the connecting device's wire-shaped conductors (e.g. 5, 6, 35, 19, 67a, 73a, 73, 109, 111, 113, 115 or 128) are arranged mutually separate and electrically isolated from each other and each extend over at least part of both the first and the second area in order to produce mutual influence between acoustic surface waves that propagate across said wire-shaped conductors in resp. the first and the second area. 2. Device as specified in claim 1, with said first and second area being made respectively a first propagation path (A) with a first converter device (3) arranged for sending acoustic surface waves along the first path, and a second propagation path (B) with another converter device (3) arranged for receiving and detecting acoustic surface waves propagating along the second lane; characterized in that the conductors (5) of the coupling device are designed and arranged on said propagation paths (A, B) for transferring at least part of the energy in surface waves that are emitted along the first propagation path (A) to the second propagation path (B) for creation of surface waves that propagate towards the other converter device (7). (Fig. 1) . 3. Device as stated in claim 2, characterized in that the section of the wire-shaped conductors (6) located above the second propagation path is curved for the formation of converging acoustic surface waves in the second propagation path. (Fig. 2). 4. Device as stated in claim 2, characterized in that the first converter device (3) and propagation path (A) as well as the coupling device (35) are extended and arranged in relation to each other in such a way that acoustic surface waves emitted from one half of the first converter device (25) will reach the coupling device every quarter of a period for surface acoustic waves emitted from the second half of the first transducer device (25). (Fig. 6). 5. Device as stated in claim 4, characterized in that the half of the first converter device (25) is half a wavelength closer to the coupling device (35) than the other half of the first converter device. (Fig. 6) . 6. Device as specified in claim 4, characterized in that each thread-shaped conductor in the coupling device (43, 45, 47) comprises two essentially equal sections, one section of which is a quarter of an acoustic wavelength closer to the first converter device (41) than the other mentioned section. (Fig. 7). 7. Device as specified in claim 2, which forms a delay line with branches, characterized in that it comprises several branch-connecting devices (67a-67c) which are arranged to form branches from the first propagation path and extend over each section of the first and the second propagation path (A,' B) as well as a number of converter devices ( 71a-71c) arranged in the second propagation path (B), between successive branch connection devices. (Fig. 9). 8. Arrangement as specified in claim 7, characterized in that each branch connection device (75a-75c) is designed so that the part that lies above the second propagation path is arranged at an angle with the part that is located above the first propagation path (A), so that the branch connection device can transmit signals to each with a separate output lane. (Fig. 10). 9. Device as stated in claim 1, and with several coupling devices (85, 87) and several areas that form propagation paths for acoustic surface waves, characterized in that the coupling devices and the propagation paths (C, D) are interconnected to form a closed circuit, while at least one further connecting device (84) is arranged for connecting surface waves in this circuit to a separate propagation path, which is equipped with an input converter device (86) and an output converter device (88). (Fig. 13) . 10. Device as stated in claim 1, and which forms a directed converter device for acoustic surface waves, characterized in that said first area and said second area lie on a common propagation path for acoustic surface waves, and a converter device (105) is arranged between the first area and the other area and at such a distance from these areas that signals emitted from the converter device (105) in mutually opposite directions will reach the coupling device (109) with a 90° mutual phase difference. (Fig. 16). 11. Device as specified in claim 10, characterized in that the wire-shaped conductors (109) are U-shaped. 12. Device as stated in claim 10, characterized in that each wire-shaped conductor has elongated 0-fprm. (Fig. 17) . 13. Device as specified in claim 1, which forms a reflector for acoustic surface waves, characterized in that said first and second areas lie in a common propagation path for acoustic surface waves and the coupling device (113) is designed to form a 3dB coupler. (Fig. 18) . 14. Device as specified in claim 1, which constitutes a rectified converter device, characterized in that the coupling device (125) above the first area is U-shaped with a converter element (123) placed between the branches of the U-shape in such a way that acoustic surface waves emitted from the converter element in mutually opposite directions will reach the branches in the same phase, while an extension (A) of one of the branches of the U-shape extends over the gnawed area, (Fig. 20) . 15. Device as specified in claim 7, and which comprises a number of rectified converter devices of the kind specified in claim 14, characterized in that each converter device forms a branch connection device (128) with a connected converter element, (129), the connection device's U-shaped part and the converter element is arranged over the second propagation path (B), while said extension of one of the U-shaped branches extends over the first propagation path. 16. Device as stated in claim 1, characterized in that the front wire-shaped conductors (225) in the coupling device are V-shaped and form opening angles which successively increase towards 180° in the direction of propagation of the surface waves. (Fig. 32). 17. Device as specified in claim 1, characterized in that the front wire-shaped conductors in the coupling device (223) increase monotonously in length in the direction of propagation of the surface waves. (Fig. 33). 18. Device as stated in claim 1, characterized in that the wire-shaped conductors in the connection device (233) also extend over an area with controllable electrical impedance (237). (Fig. 34). 19. Device as specified in claim 18, characterized in that said area with controllable electrical impedance is made of a foot-conducting material (237). (Fig. 34). 20. Device as stated in claim 1, characterized in that the wire-shaped conductors in the connection device (233) are electrically connected to a series of field-effect transistors (241). (Fig. 35 and 36) . 21. Device as specified in claim 1, characterized in that two rows of diodes are connected to opposite ends of the wire-shaped conductors (233), so that a series of connections is formed, each comprising two diodes connected in series with a wire-shaped conductor in the connection device. (Fig. 38 and 39) .