WO2012155284A1

WO2012155284A1 - Waveguide-mems phase shifter

Info

Publication number: WO2012155284A1
Application number: PCT/CH2012/000110
Authority: WO
Inventors: Jan Hesselbarth; Dimitra PSYCHOGIOU
Original assignee: Eidgenoessische Technische Hochschule Zurich ETHZ
Current assignee: Eidgenoessische Technische Hochschule Zurich ETHZ
Priority date: 2011-05-18
Filing date: 2012-05-16
Publication date: 2012-11-22
Anticipated expiration: 2013-11-18

Abstract

The disclosed invention relates to a device to efficiently realize (that is, with low dissipative loss) a large variable phase shift in a millimeter wavelength electromagnetic wave. Accordingly, a phase shifter device (1) is disclosed which comprises an axially extended waveguide (2) with a cavity (20) for conducting the electromagnetic wave, a MEMS-actuation mechanism (3), and a plurality of conductive fingers (4, 4a, 4b) which - due to their mechanical movement by means of the MEMS-actuator (3) - adjust a distributed electrical capacitance in the waveguide (2) and thus cause the adjustable phase shift to the electromagnetic wave. Specifically, at least two conductive fingers (4a, 4b) are arranged at an axial distance in a side wall (22) of the waveguide (2). By rotating opposing ends of these fingers into the cavity (20) of the waveguide (2) towards the opposing surface, which can have a predefined elevation profile, a large adjustable phase shift is achieved due to the distributed interaction between the fingers and the electromagnetic wave. Furthermore, by integrating the MEMS device directly into an air-filled metallic waveguide, a low-loss, high-power handling system is realized.

Description

Waveguide-MEMS Phase Shifter

Technical Field

The present invention relates to a phase shifter device for electromagnetic waves in the millimeter range and a use_: and an operating method for such a device .

Introduction and Background Art

Phase shifter devices (or simply phase shifters) find many applications in millimeter-wave systems (i.e., systems working with electromagnetic waves in the millimeter range, at frequencies between 30 GHz and 300 GHz) . Fields of applications for phase shifters include, e.g., phased array antennas, ultra-high data rate communication systems, and a growing number of imaging and sensing applications. Circuit and control components working at these frequencies are inherently costly (because of the high mechanical accuracies needed) and power- inefficient (because of the large conductive and dielectric losses of the materials involved) . Furthermore, efficient adjustable phase shifters (i.e., with low losses and large phase shift ranges) are much more difficult to realize [1,6] than, e.g., switches [2,5] which rely on a switchable short-circuiting of a transmission line.

The overwhelming majority of current MEMS- based phase shifter devices is based on planar transmission lines, such as micro-strip lines or coplanar waveguides. Often, the aim is the integration of the MEMS device and a transmission line structure on the same substrate. [1] discloses such a device.

However, the described implementations have the disadvantages of high dissipative losses and rather modest power-handling capabilities. Disclosure of the Invention

Hence, it is a general objective of the present invention to provide a low-loss phase shifter device for higher-power millimeter electromagnetic waves that is capable of imposing a larger phase shift. Other objectives are to provide a use and an operating method for such a device .

These objectives are achieved by the device and methods of the independent claims.

Accordingly, a phase shifter device for an electromagnetic wave with a wavelength λ in the millimeter range is provided. The phase shifter device comprises a waveguide extending along the z-direction with a cavity for conducting the electromagnetic wave. Furthermore, the phase-shifter device comprises a plurality of conductive fingers which can - depending on their adjustable position - locally interact with the (cavity of the) waveguide and thus cause a phase shift to the electromagnetic wave. An actuation mechanism adjusts the imposed phase shift to the electromagnetic wave in the following way: At least a first end of the fingers described above is moved from a first position to a second position in a first lateral direction x (perpendicular to the axial direction z) . As examples, the whole fingers can be rotated or translated to locally interact with the cavity of the waveguide. Thus, the local properties of the cavity are changed, e.g., its capacitance. Due to these changes, an adjustable phase shift is imposed to the electromagnetic wave depending on the position of the fingers. By arranging at least two of these fingers at an axial distance d (with respect to their first ends) the interaction length between the fingers and the cavity and/or the electromagnetic wave is increased and a larger phase shift can be introduced to the electromagnetic wave. Furthermore, qualitatively new features and/ or design parameters arise from this "distributed interaction" which is "lar- ger than small" compared to the wavelength. Specifically, the device cannot be modeled by a single shunt capacitor, but the distributed interaction needs to be taken into account. Thus, a low loss phase shifter device with larger imposable phase shifts can be realized which can also tolerate higher powers than a planar line system.

Advantageously, the fingers are rigid and are thus not or at least not substantially bent (i.e., with a bending radius larger than 25 mm in the second position) during their movement from the first to the second position. Thus, a more repeatable positioning of the fingers is achieved compared to, e.g., bending cantilever switches [4,5] . This leads to a better repeatability and control over the imposed phase shift to the electromagnetic wave.

In an advantageous embodiment, the actuation mechanism that moves the first end of the fingers (or the entire fingers) from their first position to their second position comprises a MEMS device on a substrate, e.g., in the form of a MEMS chip. Thus, an efficient manufacturing process can be implemented, which reduces per-unit costs.

Advantageously, the distance d between the first ends of the fingers in the axial direction z is larger than small compared to the wavelength of the electromagnetic wave, i.e., larger than λ/10, in particular larger than λ/6. Thus, an increased interaction zone (distributed interaction) between the conductive fingers and the cavity of the waveguide and/ or the electromagnetic wave is provided, which leads to a larger achievable phase shift.

In another advantageous embodiment, the actuation mechanism is adapted to rotate the (first end of the) first finger in a first rotation movement about a first rotation axis Al . The same movement applies to the (first end of the) second finger in a second rotation movement about a second rotation axis A2. Advantageously, the fingers can be arranged in a "parallel" configuration (i.e., their first ends pointing all into the same direction along z and their rotation vectors ω^, C02 , . . . pointing all into the same direction, e.g., along +y or -y) ; alternatively, an "anti-parallel" configuration of the fingers is possible, too (i.e., the first ends pointing into opposing directions, i.e., along +z and -z and the rotation vectors col, ω2 , ... pointing into contrary directions, e.g., along +y and -y) . Thus, a large variety of conductive-finger-configurations is possible and a plurality of different actuation mechanisms becomes applicable.

Advantageously, the conductive fingers overlap in the z-direction. This leads to an advantageous "leaping effect" of the electromagnetic wave from one conductive finger to (a) neighboring finger(s) . Specifically, the electromagnetic wave leaps or jumps from one conductive finger to (a) neighboring finger (s) over the gap(s) that separate(s) the finger(s) . This helps to realize said "distributed" interaction zone between the conductive fingers and the cavity and/ or the electromagnetic wave and thus, larger phase shifts can be achieved.

In another advantageous embodiment, at least three fingers are arranged in the phase shifter device, which are advantageously interdigitated in the y- direction. The term interdigitated means - in a parallel finger configuration - that each one of said second finger (s) is/ are arranged at the axial distance d with respect to each one of said first finger(s), thus leading to two "hair comb" like structures. In an anti-parallel finger configuration, "interdigitated" means that the first end(s) of each of said second finger (s) point (s), e.g., toward +z, while the first end(s) of each of said first finger (s) point (s) towards -z, thus leading to a "folded hands" like finger structure. Thus, a compact finger arrangement in the phase shifter device can be achieved and a "distributed" interaction zone between the conductive fingers and the cavity and/ or the electromagnetic wave becomes mechanically easier to realize. Thus, a low-loss phase shifter device with larger achievable phase shifts can be realized.

Advantageously, a ridge that extends into the cavity (ridge waveguide) is arranged opposite to the finger arrangement and the (first ends of the) fingers (in their second position) extend into the cavity towards this ridge. Thus, a higher capacity and a better -adjustability of the local electric capacity of the waveguide is achieved and an increased phase shift is possible. The ridge advantageously has a width (along the y-direction) that approximately (i.e., within ±25%) corresponds to the overall width or total extension of the fingers in the y- direction (i.e., the total width of all conductive fingers plus the gaps between the individual conductive fingers) . This arrangement minimizes losses and enables a good achievable phase shift. In another advantageous embodiment, the fingers are rotated by the actuation mechanism as described above. In addition, in their first position, the fingers are arranged in a wall section of the waveguide. When moved towards their second position, the fingers are rotated from the wall section (first position) into the cavity of the waveguide towards said ridge which is arranged opposite the wall section (second position) . The distance between the ridge and the wall section can vary along the z-axis according to a predefined elevation profile or ridge structure. Thus, on the one hand, a seamless integration of the fingers in the waveguide is enabled and on the other hand, a large phase shift can be achieved by a suitably defined elevation profile of the opposing surface as a function of z. As an example, a well designed elevation profile closely approximates the fingers in their second position at three positions along z via protruding studs (at the first ends of the fingers and in the center or intersection plane of the fingers) . Furthermore, notches of specific depth can be arranged between the protruding studs to realize a specific electrical length of a transmission line between the capacitances formed by the protruding studs.

Advantageously, the waveguide can be a coaxial or an evacuated or gas- or air-filled rectangular or cylindrical metallic waveguide. Thus, low-loss conduction of the electromagnetic wave and higher tolerable powers are enabled.

Advantageously, the phase shifter device furthermore comprises an actuation control unit which drives the actuation mechanism to provide accurate positioning of the conductive fingers in the first and/ or second positions. Closed-loop control is also possible, for example by measuring the capacitance (at low frequency) between the respective finger and the waveguide part opposing the finger. Furthermore, the actuation control unit ensures that the conductive fingers do not concurrently contact two opposing conductive side wall section of the waveguide and thus short-circuit the waveguide. This would lead to a switching off of the electromagnetic wave as in [3,4] .

Advantageously, the dimensions of the waveguide and other elements in the phase shifter device are suitable for wavelengths λ of the electromagnetic wave in the range of λ=1 mm to λ=10 mm, in particular in the range of λ=2 mm to λ=5 mm. Low loss conduction of a high-power electromagnetic wave is provided by a waveguide of proper dimensions.

Furthermore, phase shifter device as described can be used in a frequency tunable filter, a tunable resonator, a signal modulator, a communication system, an imaging system, a sensing system, or a phased- array antenna. Furthermore, a method for operating a phase shifter as described above is disclosed, the method comprising the movement of at least the first ends of the conductive fingers from their first positions to their second positions (which are different in x) such that a local interaction between the waveguide and the fingers is changed. This can be, e.g., a capacitance in the waveguide. The movement is achieved by means of the actuation mechanism.

Brief Description of the Drawings

The invention will be better understood and objectives other than those set forth above will become apparent when consideration is given to the following detailed description of the invention. This description makes reference to the annexed drawings, wherein:

Fig. 1 (background art) shows a rectangular metal waveguide,

Fig. 2 (background art) shows a ridge waveguide, Fig. 3 (background art) shows a schematic circuit model of a phase shifter comprising a variable shunt capacitor,

Fig. 4 (background art) shows a schematic principle of a phase shifter comprising a variably loaded 90° hybrid,

Fig. 5 shows a schematic circuit model of a phase shifter according to a first embodiment of the invention,

Fig. 6 shows a ME S-based actuation mechanism with a conductive finger,

Fig. 7 shows a top view of a second embodiment of the invention comprising two conductive fingers in an anti-parallel configuration,

Fig. 8 shows a perspective view of a third embodiment of the invention comprising three conductive fingers, Fig. 9 shows a top view of the third embodiment of the invention,

Fig. 10 shows an elevation profile of a ridge of the third embodiment of the invention,

Fig. 11 shows a perspective view of a fourth embodiment of the invention comprising a structured ridge with the fingers in a second position,

Fig. 12 shows a perspective view of the fourth embodiment of the invention with the^: fingers in a first position,

Fig. 13 shows an elevation profile of the ridge of the fourth embodiment of the invention,

Fig. 14 shows a top view of a fifth embodiment of the invention comprising two conductive fingers in a parallel configuration,

Fig. 15 shows a diagram providing a simulated performance of the fourth embodiment of the invention.

Modes for Carrying Out the Invention

The disclosed invention relates to a device to efficiently realize (that is, with low dissipative loss) a large variable phase shift in a millimeter wavelength electromagnetic wave. In an advantageous embodiment, a phase shifter device is disclosed which comprises an axially extended waveguide with a cavity for conducting the electromagnetic wave, a MEMS-actuation mechanism, and a plurality of conductive fingers which - due to their mechanical movement by means of the MEMS-actuator - adjust a distributed electrical capacitance in the waveguide and thus cause the adjustable phase shift to the electromagnetic wave. Specifically, at least two conductive fingers are arranged at an axial distance in a side wall of the waveguide. By rotating opposing ends of these fingers into the cavity of the waveguide towards the opposing surface, which can have a predefined elevation profile, a large adjustable phase shift is achieved due to the distributed interaction between the fingers and the electromagnetic wave. Furthermore, by integrating the MEMS device directly into an air-filled metallic waveguide, a low-loss, high-power handling system is realized.

Description of the Figures:

Metallic waveguides:

A transmission line with low losses that is widely used for electromagnetic waves with wavelengths λ in the millimeter range is the rectangular metal waveguide 2 as shown in Fig. 1. Electrically conductive wall sections 22 (e.g., comprising a gold coating) confine a cavity 20 in which the electromagnetic wave is conducted. Usually the waveguides 2 are used in single- mode operation, i.e., a single fundamental mode of the electromagnetic wave is efficiently conducted. The waveguide's cross section is usually between 1/2 and one wavelength λ in width (i.e., along a second lateral direction y) and 1/4 to 1/2 wavelength λ in height (i.e., along a first lateral direction x) . The cross sections are standardized; e.g., a waveguide for the frequency range of operation of v=75-110 GHz is called WR10; it is 2.54 mm in width and 1.27 mm in height (these are the inside dimensions of the cavity 20 for air or vacuum dielectric) . The ends of the waveguide 2 may be hermetically sealed, thus reducing the risk of contamination of the cavity and devices arranged therein (e.g., MEMS devices 3 and/ or conductive fingers 4, not shown) . An electric field intensity of the electromagnetic wave is denoted by the arrows and the dotted line in Fig. 1.

Fig. 2 shows a ridge waveguide. The height of the rectangular metal waveguide as shown in Fig. 1 for a given frequency can be reduced by introducing a ridge 21 ("ridge waveguide"; or two ridges on opposite sides — "dual-ridge waveguide", not shown) . The electric field (dotted line in Fig. 2) is concentrated below the ridge (whereas for the rectangular waveguide 2 in Fig. 1, it has sinusoidal distribution over the width) . The concentration is the stronger, the smaller the gap below the ridge. However, for very deep ridge cross sections, loss increases (because of higher current density in the metal wall sections 22) and sensitivity to manufacturing tolerances increases. As an example, dimensions for a ridge waveguide 2 for frequencies around 100 GHz are: Width 1.45 mm, height 1 mm, width of ridge w_r 0.5 mm, ridge depth 0.5 mm (therefore, air gap below ridge 0.5 mm) . Ridge waveguides are advantageous with respect to the current invention, as a smaller distance between the fingers 4 and the ridge 21 is possible (see, e.g., Fig. 8) and therefore a higher capacitance can be reached. This is because capacitance between two conductors scales with C~l/d wherein C denotes the capacitance and d denotes the distance between the conductors.

Principles of phase shifters

Variable transmission phase shift can be realized by a variable shunt capacitor 70 in a transmission line 69 (such as a waveguide 2) with an input 68 and an output 71 (Fig. 3). The electromagnetic wave at the output 71 has a shifted phase (depending on the length of the transmission line 69 and on the shunt capacitor 70) with respect to the electromagnetic wave at the input 68. Input reflection and insertion loss will vary with the varying capacitance of the shunt capacitor 70. The variation of transmission phase which can be obtained from a single shunt capacitor 70 is measured for a maximum insertion loss. For the circuit of Fig. 3, this variation of transmission phase is rather limited, i.e., on the order of 17° (for a fixed insertion loss, e.g., maximum insertion loss <0.4 dB) . In other words, when measuring obtainable phase shifts, it is always assumed that the phase shifter 1 remains "almost" impedance matched with respect to the environment, i.e., input reflection remains small, e.g., below -15 dB, and transmitted power remains large, e.g., above -0.4 dB.

It is understood that the range of phase variation can be increased by concatenating phase shifting elements (as the one described above) , or by using the above-mentioned phase-shifting elements in reflection mode. In the latter mode of operation, a so-called 90- degrees -hybrid (or directional coupler) is connected to two identical elements which can vary the phase of their input reflection coefficients. Fig. 4 (taken from [6]) shows the principle of operation of such a device. A transmission-phase-shifter can in principle be converted to a reflection-phase-shifter with roughly twice the phase shifting range by simply (that is, with appropriate phase) short-circuiting the output port of the device.

First embodiment

Fig. 5 schematically shows an approximate circuit model of a first embodiment of the invention. Three shunt capacitors 70 (two are supposed to be located at the first ends 40 of the fingers 4a and 4b and one in the middle, see Fig. 8) are separated by two short segments of (impedance-mismatched) transmission line 69. The circuit model allows for fast parameter sweeps and performance optimization of the phase shifter 1. As a result, it turns out that a specific setup with three shunt capacitors (n=2.74) and two short sections of impedance- mismatched transmission line (electrical length of 77.75° each, impedance 25 Ω) results in a large variation of transmission phase, i.e., on the order of 160° for a maximum insertion loss of 0.4 dB. The following figures disclose different technical realizations with a MEMS device 3 integrated into a waveguide 2 of the schematic three-capacitor-two-lines phase shifting topology of Fig. 5.

MEMS actuators

As it is shown in Fig. 6, a typical MEMS actuator 3 is based on an electrostatic comb actuator 33 (i.e., stator combs 31 and movable combs 32 in an inter- digitated setup) leading to a rotational movement (arrows) of a mechanical part such as a finger 4. The MEMS actuator 3 furthermore comprises a substrate 34. Typically, MEMS actuators 3 are manufactured in a multi-step etching and metallization process from the substrate 34. In the scenario of the phase shifter 1 according to the invention, the moved finger 4 is highly conductive (e.g., gold coated) and placed inside a waveguide 2, thereby affecting the electromagnetic fields of the electromagnetic wave in a specific way. The electrostatic comb actuator (s) 33 of the MEMS actuator 3 is/ are placed as much as possible outside the waveguide 2 in order to avoid disturbance of the electromagnetic wave and unwanted couplings. The finger 4 comprises a first end 40 which is rotated from a first position 41 - which is preferably located in a wall section 22 of the waveguide 2 - to a second position 42 in the cavity 20 of the waveguide 2. This rotation movement is described by a rotation vector coi or a>2 around a rotation axis Al or A2.

Second embodiment

Fig. 7 shows a top view of a phase shifter 1 according to a second embodiment of the invention. Electromagnetic waves with a wavelength λ in the millimeter range are conducted along z in a cavity 20 of a waveguide 2. The waveguide 2 furthermore comprises a ridge 21 (of width w_r) and metallic (i.e., highly conductive) wall sections 22. Besides the waveguide 2, the phase shifter 1 comprises a MEMS actuator 3 (actuation mechanism 3, e.g., of the type from Fig. 6) comprising electrostatic comb actuators 33 which rotate first ends 40 of fingers 4a and 4b from the wall section 22 (first position 41) into the cavity 20 (second position 42) of the waveguide 2. Thereby a capacitance in the waveguide 2 is changed. Thus, a MEMS-adjustable variable shunt capacitor 70 is realized in the ridge metal waveguide 2, the fingers 4a and 4b primarily varying the capacitance in the gap between the (top) ridge 21 and the (bottom) wall section 22 of the waveguide 2. The dimension (in particular along z) of the actuated element (i.e., the anti-parallel fingers 4a and 4b) realizes a distributed device: The largest capacitances occur at the first ends 40 of the actuated fingers 4a and 4b as they come closest to the ridge 21 when in their second position. A different, smaller capacitance occurs at the center (along z) between the first ends 40 of fingers 4a and 4b. These first ends 40 are separated by a distance d along z which is "larger than small", (e.g., larger than λ/10 or λ/6) compared to the wavelength λ of the electromagnetic wave hv at the frequency of operation v. As an example, the distance d can be in the range of 1.2 mm and the wavelength λ at 100 GHz is 3 mm. The distance of the rotation axes Al and A2 is 1.4 mm. As a result, the device cannot be modeled by a single variable shunt capacitor 70 as in Fig. 3 but realizes the three-capacitor-two-lines phase shifting topology of Fig. 5. Accordingly, a larger variable phase shift is realized by the phase shifter 1. The total extension Wf along y of the fingers 4a, 4b (widths of the fingers and spacing between fingers) approximately (i.e., within ±25%) corresponds to the ridge width w_r . When rotating the first finger 4a from its first position 41 to its second position 42 in a first rotation movement, the corresponding rotation vector coi points along -y. When rotating the second finger 4b from its first position 41 to its second position 42 in a second rotation movement, the corresponding rotation vector a>2 points along +y. This is due to the anti-parallel configuration of the fingers 4a and 4b.

Third embodiment

Figs. 8, 9, and 10 show a third embodiment of the invention. Fig. 8 shows a perspective view of the phase shifter 1, whereas Fig. 9 shows a top view of the finger configuration, and Fig. 10 shows a ridge contour (elevation profile) with geometrical details. The phase shifter 1 comprises a ridge waveguide 2 with a ridge 21 and a cavity 20 and extends along the z -direction. Furthermore, the phase shifter 1 comprises three conductive fingers 4a, 4b, and 4c in an anti-parallel, interdigi- tated configuration. Note: For clarity, only two fingers 4a and 4b are shown in Fig. 8 as the center part of the figure is cut along the symmetry plane S. As for operation, the third embodiment works similar to the second embodiment: By rotating the fingers 4a and 4b from their first positions 41 in the wall section 22 to their second positions 42 (around rotation axes Al and A2), the first ends 40 come closer to the ridge 21 (or, as shown in Fig. 8, a second ridge 21a on the ridge 21) and the capacitances between the conductive fingers 4a and 4b and the second ridge 21a are thus increased. Thus, a phase shift is introduced to the electromagnetic wave hv according to Fig. 5. Again, the arrangement of the fingers 4a and 4b leads to a distributed interaction with the distance d being "larger than small" compared to the wavelength λ of the electromagnetic wave h . Here, the fingers 4a and 4c are rotated by the same MEMS actuator 3 around Al with a rotation vector pointing along -y. The finger 4b is rotated by a different MEMS actuator 3 around A2 with a rotation vector a>2 pointing along +y. ₍

An actuation control unit 35 drives the MEMS actuators 3 such that an accurate positioning of the fin- gers 4a, 4b, and 4c in their first 41 and/ or second positions 42 is achieved and that the conductive fingers do not concurrently touch two opposing conductive side wall section 22 of the waveguide 2 (i.e., the bottom wall section 22 and the second ridge 21a) and thus short-circuit the waveguide 2.

The performance of the device is analyzed by a 3D electromagnetic simulator based on finite elements. Such a simulation takes geometrical and material parameters into account and solves for the Maxwell equations over frequency. For ease of comparison, the geometrical structure has been simplified (less geometrical details) and the dissipative loss (conversion of a part of the signal energy to heat) is neglected. With these assumptions, the simulation of a device according to the third embodiment gives 38° variable phase shift at a frequency of 99 GHz. This is much more than the circuit model value of a single shunt capacitor 70 (Fig. 3), indicating the advantage of the distributed interaction. The following parameters were used: Second position of fingers 4a and 4b: 7° mechanical deflection angle from wall section 22, first position: 0° mechanical deflection angle from wall section 22, ridge depth (along x) : 0.5 mm, waveguide height (along x) : 1 mm, w_r (along y) : 0.5 mm, waveguide width (along y) : 1.45 mm, depth (along x) of second ridge 21a (from top of waveguide) : 0.8 mm, length (along z) of second ridge 21a: 1.45 mm, width (along y) of fingers 4a, 4b and 4c : 0.6 mm, length (along z) of fingers 4a and 4c : 1.4 mm, distance along z between Al and A2 : 1.4 mm.

Fourth embodiment

A fourth embodiment of the phase shifter 1 according to the invention is shown in Figs. 11, 12, and 13. The phase shifter 1 according to the fourth embodiment is very similar to the third embodiment described above. However, in an attempt to further increase the achievable phase shift, an elevation profile (i.e., ridge structure of the ridge 21 along the x-direction) is introduced to approximate the optimum configuration from the analytical model of the first embodiment (Fig. 5) . A variable capacitance ratio ( "n" in Fig. 5) is realized by protruding studs 21b, 21c, and 2 Id on the ridge 21. The length of the two short transmission lines 69 in Fig. 5

(electrical lengths 77.75°) connecting the three shunt capacitors ; 70 is realized by varying the depth along the x-direction of two notches 21e and 21f in the ridge 21. As a result, a structured ridge with an elevation profile

(along x as a function of z) is obtained. Fig. 13 gives geometrical details. This adds very small complexity and/ or cost to the device. A 3D simulation of the electromagnetic fields in a device with structured ridge shows a large increase of phase shift. With a 7.5° mechanical deflection angle of the fingers 4a, 4b, and 4c and the geometrical parameters from Fig. 13, a field-simulation of the phase shifter 1 according to the fourth embodiment of the invention yields a 111° variable phase shift (see below) . All other parameters (i.e., dimensions of the MEMS fingers and the waveguide dimensions) are identical to the third embodiment described above.

Fig. 15 shows the simulated performance of a phase shifter 1 according to the fourth embodiment

(curves for finger deflection angles 0°-7.5°) . Over the entire -0.4 dB bandwidth, the phase tuning range varies from 73° at v=85.6 GHz to 111° at v=94 GHz. Over the frequency bandwidth from 85.6 GHz to 94 GHz and over the MEMS tuning range from zero deflection to 7.5 degrees deflection, the input reflection is always smaller than -10.5 dB. From the variation of the shape of the transmission magnitude curve over the MEMS tuning range, it is apparent that the device acts like a tunable, two- resonator bandpass filter with variable transmission phase . Fifth embodiment

Fig. 14 shows a top view of a fifth embodiment of the invention comprising two conductive fingers in a parallel configuration. Besides the finger configuration, the fifth embodiment is very similar to the second embodiment (Fig. 7 ) described above. Arranging the fingers 4a and 4b in such a parallel configuration and co-rotating the fingers 4a and 4b when moving them from their first 41 to their second position 42 enables a variety of different MEMS-actuators 3.

Definitions :

The term "millimeter-wave" refers to electromagnetic waves hv with a wavelength λ in the millimeter range (in vacuum), in particular with a wavelength λ in the range of λ=1 mm to λ=10 mm or frequencies v in the range of v=30 Ghz to v=300 Ghz (as calculated from c=vX) .

The term "conductive finger" refers to an element, e.g, an elongated block of material, which at least in parts (e.g., in a layer) possesses electrical conductivity similar to that of a metal (e.g., σ=4.52 · 10⁷S/m) . This conductivity of the conductive fingers is advantageously featured over the entire length (dimension along the z-direction) of the conductive finger to enable low-loss conduction of the electromagnetic wave .

"MEMS" stands for "Micro-Electro-Mechanical System", where as small mechanical movement is realized by electrical actuation, typically by electrostatically, electro-magnetical , piezo-electrical , or electrically- induced thermal actuation elements. MEMS devices are often integrated on a substrate, thus forming a Micro- Electro-Mechanical Systems chip or MEMS-chip.

The terms "parallel"/ "anti-parallel" denote that two vectors or directions have the same bearing (i.e., they are defined by the same two separate points in space) but point in the same/ opposite directions.

An "intersection plane of fingers" refers to an anti-parallel finger configuration of at least two fingers. The term "intersection plane" then denotes a plane which is perpendicular to the axial direction z and where the top surface (along the first lateral direction x) of the first finger intercepts the top surface (along the first lateral direction x) of the second finger in an xz-proj ection, i.e., a projection along the second lateral direction y. Such an intersection plane is denoted by reference number 25 in Fig. 11.

Notes :

Planar lines show inherently higher loss than waveguides, because of their small cross section (causing high current densities) and due to potentially lossy dielectric substrates.

The power-handling capacity of a waveguide is much higher than that of a planar line.

If the application requires waveguide connections (as it can be the case, for example, in antenna- array structures) , a conventional MEMS needs waveguide- to-planar transitions, which add significant loss and mechanical complexity at millimeter wave frequencies.

Most MEMS devices need hermetic packaging for reliable operation. This not only contributes to the costs of the devices, but also the in/ out leads of a packaged MEMS add significant loss. A MEMS device integrated into a waveguide needs hermetic waveguide windows, which can be made to have very low loss. Furthermore, applications requiring several MEMS chips (e.g., feeder trees for phased array antennas) can avoid hermetic packaging of the individual MEMS chips. Instead, hermetic enclosures can be placed at the input and output of the system (e.g., at waveguide feed and aperture of the array antenna) , thereby greatly reducing the number of hermetic feedthroughs .

Compared to the 3D simulations, a practical device shows slightly increased transmission loss (due to dissipative loss, that is, signal partly converted to heat) and slightly reduced phase shift performance (due to added mechanical complexity) compared to the 3D electromagnetic field simulations. However, the benefits of the proposed design features (e.g., distributed interaction between fingers and electromagnetic wave and/ or cavity of the waveguide, structured ridge with an elevation profile) remain significant.

Similar characteristics are achievable with numbers of crossed fingers different from two or three. However, with an increasing number of fingers, more area is "wasted" for gaps separating the fingers. This leads to deteriorated electrical properties and increased losses .

Improved phase shift characteristics can be achieved by introducing more complexity in the ridge structure, for example, using more protruding studs and more notches separating them. This adds, however, complexity to the manufacturing process due to smaller features, and an increase of sensitivity to tolerances. Then, instead of the three-capacitor-two-lines phase shifting topology of Fig. 5, also other topologies including more elements are possible, e.g., a four- capacitor-three-lines topology, a five-capacitor-four- lines topology and so on.

Instead of the shown rotation motion of the fingers in which the first end of the fingers is rotated from its first position to its second position, also a piston motion in which the whole finger is moved from its first position to its second position is possible. In this case, however, an electrical connection between the finger and the surrounding wall section needs to be established .

It is understood that several or all positions between the said first and second positions of the fingers can be set in all embodiments by the actuation mechanism, allowing for a continuously variable setting of the phase shift.

References

[1] C. Siegel, V. Ziegler, U. Prechtel, B. Schonlinner, H. Schumacher, „A Ka-band RF-MEMS phase shifter approach based on a novel dual-state microstrip line," Proc . European Microwave Conference, Munich 2007, pp. 1221-1224.

[2] M. Daneshmand, R.R. Mansour, N. Sarkar, "RF MEMS waveguide switch," IEEE Int Microwave Symp, 2004, pp. 589-592.

[3] M. Daneshmand, R.R. Mansour, N. Sarkar, "RF MEMS waveguide switch," IEEE Trans Microwave Theory Tech, vol. 52, no. 12, Dec. 2004, pp. 2651-2657.

[4] M. Daneshmand, R.R. Mansour, "Multi-Port RF MEMS Waveguide Switch," IEEE Int. Microwave Symp., 2005, pp. 935-938.

[5] M. Daneshmand, R.R. Mansour, "Multiport MEMS-based waveguide and coaxial switches," IEEE Trans Microwave Theory Tech., vol. 53, no. 12, 2005, pp. 3531-3537.

[6] J. Hesselbarth, R. Vahldieck, "Concepts for millimeter-wave waveguide-MEMS phase shifters," Microwave Optical Technology Letters, Wiley, vol. 52, no. 5, May 2010, pp. 1028-1030.

[7] J. Hesselbarth, R. Vahldieck, „Tunable Millimeter- Wave Device Based on Waveguide-Packaged MEMS," Proc. Int. Conf . Device Packaging, pp. TP26, Scottsdale/AZ, March 2009.

[8] B. Lacroix, A. Pothier, A. Crunteanu, P. Blondy, "Phase shifter design based on fast RF MEMS switched ca- pacitors" , Proc . European Microwave Conference, Amsterdam 2008, pp. 1505-1508.

[9] Y. Li, S. Kiihne, D. Psychogiou, J. Hesselbarth, C. Hierold, "Large deflection actuator for variable-ratio RF MEMS power divider application," 21st Micromechanics and Microsystems Europe Workshop, Enschede, September 2010.

References numbers:

1 phase shifter device

2 waveguide

20 cavity

21 ridge

21a second ridge

21b, 21c, 2 Id protruding studs

21e, 21f notches

w_r width (of ridge) along y

22 wall section (of waveguide 2)

25 intersection plane

3 actuation mechanism, such as MEMS device

31 stator combs

32 movable combs

33 electrostatic comb actuator

34 substrate

35 control unit

Al first rotation axis

A2 second rotation axis

coi first rotation vector

a>2 second rotation vector

4, 4a, 4b finger

40 first end (of finger)

41 first position (of first end of finger)

42 second position (of first end of finger)

Wf total extension (of fingers) along y

d (axial) distance (of first ends of fingers) 68 input (of transmission line 69)

69 transmission line

70 shunt capacitor

71 output (of transmission line 69) z axial direction

X first lateral direction

y second lateral direction λ wavelength

hv electromagnetic wave

Claims

1. Phase shifter device (1) for an electromagnetic wave with a wavelength λ in the millimeter range comprising a waveguide (2), an actuation mechanism (3), and at least a first and a second conductive finger (4, 4a, 4b) ,

wherein said waveguide (2) comprises a cavity (20), is adapted to conduct said electromagnetic wave, and extends along an axial direction (z),

wherein said fingers (4, 4a, 4b) are adapted to locally interact with said waveguide (2) depending on their position (41, 42) ,

wherein said actuation mechanism (3) is adapted to reversibly move at least a first end (40) of each of said fingers (4, 4a, 4b) from a first position (41) to a second position (42), and

wherein said first position (41) is different from said second position (42) in a first lateral direction (x) which is perpendicular to said axial direction (z) ,

characterized in that said first end (40) of said first finger (4a) is arranged at a distance (d) in said axial direction (z) to said first end (40) of said second finger (4b) .

2. Phase shifter device (1) of claim 1 wherein said fingers (4, 4a, 4b) are rigid and wherein said movement of said fingers from said first position

(41) to said second position (42) by said actuation mechanism (3) does not comprise a bending of said fingers

(4, 4a, 4b) .

3. Phase shifter device (1) of any of the preceding claims wherein said actuation mechanism (3) comprises a MEMS device (3) on a substrate (34)..

4. Phase shifter device (1) of any of the preceding claims wherein said distance (d) in said axial direction (z) of said first ends (40) of said fingers (4a, 4b) is larger than λ/10, in particular larger than λ/6.

5. Phase shifter device (1) of any of the preceding claims wherein said actuation mechanism (3) is adapted to rotate said first finger (4a) in a first rotation movement about a first rotation axis (Al) and wherein said actuation mechanism (3) is furthermore adapted to rotate said second finger (4b) in a second rotation movement about a second rotation axis (A2), and wherein said rotation axes (Al, A2) are parallel to a second lateral direction (y) which is perpendicular to said axial direction (z) and to said first lateral direction (x) .

6. Phase shifter device (1) of claim 5 wherein said first finger (4a) is arranged in a parallel configuration with respect to said second finger (4b) and wherein a first rotation vector (ω ) of said first rotation movement is parallel to a second rotation vector (002) of said second rotation movement.

7. Phase shifter device (1) of claim 5 wherein said first finger (4a) is arranged in an anti- parallel configuration with respect to said second finger (4b) and wherein a first rotation vector (ω]_) of said first rotation movement is anti-parallel to a second rotation vector (Q2) of said second rotation movement.

8. Phase shifter device (1) of any of the preceding claims wherein said first finger (4a) is arranged to at least partly overlap said second finger (4b) in said axial direction (z) .

9. Phase shifter device (1) of any of the preceding claims wherein a number of said fingers (4, 4a, 4b) is at least three.

10. Phase shifter device (1) of claim 9 wherein said fingers (4a, 4b) are interdigitated in a second lateral direction (y) which is perpendicular to said axial direction (z) and to said first lateral direction (x) .

11. Phase shifter device (1) of any of the preceding claims wherein said waveguide (2) comprises a ridge (21) extending into said cavity (20) and wherein at least said first ends (40) of said fingers (4, 4a, 4b) are adapted to be extendable into said cavity (20) towards said ridge (21) by means of said actuation mechanism (3) .

12. Phase shifter device (1) of claim 11 wherein a width (w_r) of said ridge (21) in a second lateral direction (y) is equal within ±25% to a total extension (wf) of said fingers (4a, 4b) in said second lateral direction (y) , wherein said second lateral direction (y) is perpendicular to said axial direction (z) and to said first lateral direction (x) .

13. Phase shifter device (1) of claim 11 wherein said fingers (4a, 4b) are arranged in a wall section (22) of said waveguide (2), wherein said ridge (21) is arranged opposite said wall section (22) with at least said first ends (40) of said fingers (4a, 4b) being extendable into said cavity (22) towards said ridge (21) by means of said actuation mechanism (3) and wherein at a location of said fingers (4a, 4b) a distance along said first lateral direction (x) between said ridge (21) and said wall section (22) varies as a function of said axial direction ( z) .

14. Phase shifter device (1) of claim 13 wherein said ridge (21) further comprises a protruding stud (21b, 21c, 21d) and a notch (21e, 21f ) .

15. Phase shifter device (1) of claims 8 and

14,

wherein said fingers (4a, 4b, 4c) have within ±25% the same length in said axial direction (z) and are arranged in an anti-parallel configuration,

wherein a first protruding stud (21b) is arranged on said ridge (21) opposite said first end (40) of said first finger (4a) ,

wherein a second protruding stud (21c) is arranged on said ridge (21) opposite an intersection plane (25) in said axial direction (z) of said first finger (4a) and said second finger (4b) ,

wherein a third protruding stud (21d) is arranged on said ridge (21) opposite said first end (40) of said second finger (4b) ,

wherein a first notch (21e) is arranged on said ridge (21) between said first protruding stud (21b) and said second protruding stud (21c) in said axial direction ( z) ,

wherein a second notch (2 If) is arranged on said ridge (21) between said second protruding stud (21c) and said third protruding stud (21d) in said axial direction (z) .

16. Phase shifter device (1) of any of the preceding claims wherein said waveguide (2) is a coaxial waveguide or an evacuated or air- or gas-filled rectangular or cylindrical metal waveguide.

17. Phase shifter device (1) of any of the preceding claims further comprising an actuation control unit (35) which is adapted to drive said actuation mechanism (3) and to provide positioning of said conductive fingers (4, 4a, 4b) in said first (41) and/ or said second positions (42) ,. wherein in said first (41) and said second positions (42) said conductive fingers (4, 4a, 4b) are positioned such that they do not short-circuit two opposing conductive side wall section (22) of said waveguide (2) .

18. Phase shifter device (1) of any of the preceding claims wherein said wavelength λ in vacuum is in the range of λ=1 mm to λ=10 mm, in particular in the range of λ=2 mm to λ=5 mm.

19. A use of a phase shifter device (1) of any of the preceding claims in one of the group of a frequency tunable filter, a tunable resonator, a signal modulator, a communication system, an imaging system, a sensing system, and a phased-array antenna.

20. A method for operating the phase shifter device (1) of any of the claims 1 to 18 comprising the steps of

conducting said electromagnetic wave in said waveguide (2) ,

moving at least said first ends (40) of said conductive fingers (4, 4a, 4b) from said first positions (41) to said second positions (42) by means of said actuation mechanism (3) such that a local interaction between said waveguide (2) and said fingers (4, 4a, 4b) is changed.