1. Field of the Invention
This invention relates to phase shifting and steering of high frequency electromagnetic signals.
Description of the Related Art
Electromagnetic signals are commonly guided from a radiating element to a destination via a coaxial cable, metal waveguide, or microstrip transmission line. As the frequency of the signal increases, these devices must have smaller cross-sections to transmit the signals. For example, a metal waveguide that is 58.420 cm wide and 29.210 high at its inside dimensions, transmits signals in the range of 0.32 to 0.49 GHz. A metal waveguide that is 0.711 cm wide and 0.356 cm high at its inside dimensions, transmits signals in the range of 26.40 to 40.00 GHz. [Dorf, The Electrical Engineering Handbook, Second Edition, Section 37.2, Page 946 (1997)]. As the signal frequencies continue to increase, a point is reached where use of these devices becomes impractical. They become too small and expensive, require precision machining to produce, and their insertion loss can become too great.
Frequencies exceeding approximately 100 GHz (referred to as millimeter waves) can be transmitted as a free-space beam. The signal from a radiating element is directed to a lens that focuses the signal into a millimeter wave beam having a diameter up to several centimeters. This form of transmission is referred to as “quasi-optic” when the lens diameter divided by the signal wavelength is in the range of approximately 1-10. In the optic regime, the lens diameter divided by the frequency wavelength is normally much greater than 10. [IEEE Press, Paul f. Goldsmith, Quasi-optic Systems, Chapter 1, Gaussian Beam Propagation and Applications (1999)]
One method of amplifying these high frequency beams is to combine the power output of many small amplifiers in a quasi-optic amplifier array. The amplifiers of the array are oriented in space such that the array can amplify a Gaussian beam of energy rather than amplifying a signal guided by a transmission line. However, commercial use of these “open” systems is not practical because they are fragile and can be contaminated by the surrounding environment. Also, there is no simple, durable and reliable mechanism for beam phase shifting or steering.
Conventional rectangular waveguides cannot be used. In addition to their size and insertion loss disadvantages they do not provide an optimal signal to drive an amplifier array. Because the sidewalls of a metal waveguide are conductive, they present a short circuit to the beam's E field and it cannot exist near the conductive sidewall. The power densities of the beam's E and H fields drop off closer to the sidewalls, with the power density of the beam varying from a maximum at the middle of the waveguide to zero at the sidewalls.
For an amplifier array to operate efficiently, each individual amplifier in the array must be driven by the same power level. When amplifying the type of signal provided by a conventional metal waveguide, the amplifiers at the center of the array will be overdriven before the edge amplifiers can be adequately driven. In addition, individual amplifiers in the array will see different source and load impedances depending upon their locations in the array. The array's edge amplifiers become ineffective, significantly reducing the array's potential output power.
A high impedance surface will appear as an open circuit and the E field will accordingly not experience the drop-off associated with a conductive surface. A photonic surface structure has been developed which exhibits a high impedance to a resonant frequency and a small bandwidth around that frequency [D. Sievenpiper,High Impedance Electromagnetic Surfaces, (1999) PhD Thesis, University of California, Los Angeles]. The surface structure comprises patches of conductive material mounted in a sheet of dielectric material, with conductive vias through the dielectric material from the patches to a continuous conductive layer on the opposite side of the dielectric material. This surface presents a high impedance to the resonant frequency and the gaps between the patches prevent surface current flow in any direction.
A second impedance structure has been developed that is particularly applicable to the sidewalls and/or top and bottom walls of metal rectangular waveguides. [M. Kim et al., A Rectangular TEM Waveguide with Photonic Crystal Walls for Excitation of Quasi-Optic Amplifiers, (1999) IEEE MTT-S, Archived on CDROM]. Either two or four of the waveguide's walls can have this structure, depending upon the polarizations of the signal being transmitted. The structure comprises parallel conductive strips on a substrate of dielectric material. It also includes conductive vias through the sheet to a conductive layer on the substrate's surface opposite the strips. At the resonant frequency, this structure presents as series of high impedance resonant L-C circuits.
When used on a rectangular waveguide's sidewalls, the structure provides a high impedance boundary condition for the resonant frequency's E field component for a vertically polarized signal, the E field being transverse to the conductive strips. The high impedance prevents the E field from dropping off near the waveguide's sidewalls, maintaining an E field of uniform density across the waveguide's cross-section. Current can flow down the waveguide's conductive top and bottom walls to support the signal's H field with uniform density. Accordingly, the signal maintains near uniform power density across the waveguide aperture.
When the high impedance structure is used on all four of the waveguide's walls, the waveguide can transmit independent cross-polarized signals with near-uniform power density. The structure on the waveguide's sidewalls presents a high impedance to the E field of the vertically polarized signal, while the structure on the waveguide's top and bottom walls presents a high impedance to the horizontally polarized signal. The structure also allows conduction through the strips to support the signal's H field component of both polarizations. Thus, a cross-polarized signal of uniform density can be transmitted.
Waveguides employing these high impedance structures are also able to transmit signals close to the resonant frequency that would otherwise be cut-off because of the waveguide's dimensions if all of the waveguide's walls were conductive. At the resonant frequency, the waveguide essentially has no cut-off frequency and can support uniform density signals when its width is reduced well below the width for which the frequency being transmitted would be cut-off in a metal waveguide.
The present invention provides a new rectangular waveguide that can shift the phase of the signal passing through it. The new waveguide has an impedance wall structure on at least two opposing walls that present a capacitive impedance to the E field of the signal passing through the waveguide. The capacitive impedance increases the signal's propagation constant and shifts its phase.
In one embodiment, the invention utilizes the impedance structures on two or all four of its walls. Instead of transmitting a signal at the wall structure's resonant frequency, the waveguide passes a signal with a frequency well above the structure's resonant frequency. This results in the structure presenting a capacitive impedance to the transverse E field of the waveguide's signal, instead of a very high impedance. The propagation constant of the signal increases and the waveguide becomes a “slow wave” structure, shifting the phase of the signal. The preferred impedance structure is the parallel conductive strip described above.
In another embodiment, the phase shifting waveguide again has an impedance structure on two or all four of its walls, with the impedance structure being voltage controlled to resonate at different frequencies. The range of resonant frequencies is below the signal frequency being passed by the waveguide, and changes in the structure's resonant frequency result in different shifts in the phase of the signal being passed. The preferred impedance structure has parallel conductive strips. To change the resonant frequency, the impedance structures include varactor diodes along the gaps between the structure's conductive strips. A change in the voltage applied to the varactor diode changes both the capacitance across the gap and the resonant frequency of the structure.
Another embodiment of the new waveguide includes both a phase shifter and an amplifier array to amplify the phase shifted signal. For a vertically polarized signal, a multi-region impedance structure is initially provided on the waveguide's sidewalls. The first region is a conductive strip impedance structure that is resonant to the beam frequency at the front of the waveguide. Progressing further down the waveguide, the gap between the conductive strips narrows, reducing the structure's resonant frequency. Next the signal enters the phase shift region where the gap between the strips maintain a constant width. Between the gaps is a varactor structure that varies the capacitance across the gaps in response to voltage changes. As described above, this change in capacitance shifts the beam's phase. The signal then enters the second transition region where the gaps widen so that the structure resonates at the signal frequency. The signal then enters the amplifier region, which has a strip structure on all four walls that resonates at the signal frequency. This section provides a near uniform signal to the amplifier, and the amplified signal emits from the waveguide.
The new waveguides can be used in a new millimeter beam module that is placed in a millimeter beams path to shift the beam's phase and/or steer the beam, as well as amplify the beam. The module includes a plurality of new waveguides adapted to receive at least part of the electromagnetic beam. The waveguides are adjacent to one another, with their longitudinal axes aligned with the propagation of the beam. In one embodiment, each waveguide can be set to cause the same phase shift in its portion of the beam, shifting the phase in the entire beam uniformly. Each waveguide can also cause a different phase shift to steer the beam, and can also include-a amplifier array to amplify the beam.
To reduce beam degradation from reflection off the front edge of the module the waveguides in the module include a front end launching region in the form of a patch impedance structure that is resonant at the beam frequency. This makes the front edges of the waveguides invisible to the entering wavefront, allowing only the TEM mode of the signal to enter the waveguide and preventing signal reflection.
These and other further features and advantages of the invention will be apparent to those skilled in the art from the following detailed description, taken together with the accompanying drawings, in which:
Waveguide Phase Shifter
With the impedance structures 12 on its sidewalls, the waveguide 10 is particularly applicable to passing vertically polarized signals that have an E field transverse to the strips 18. As shown in
The new waveguide is not designed to transmit signals with a frequency that causes the structure 12 to resonate. Instead, it functions as a phase shifter by passing signal well above the structures' resonant frequency. It relies on the unique relationship between the propagation constant of a particular frequency signal in a waveguide, and the frequency at which the impedance structures resonate. In
In this example, the two curves intersect at 44 GHz (point 40 in the graph). Thus, forming the waveguide with a resonant frequency of 44 GHz will allow the waveguide to transmit a 44 GHz signal as if propagating in free space. Changes in the impedance structure's resonant frequency changes the signal's propagation constant. Due to the near-vertical slope of curve 32 at lower frequencies and its near-horizontal slope at higher frequencies, increasing the structure's resonant frequency results in only small changes in the signals propagation constant, while reducing the resonant frequency causes a significant change in the beam's propagation constant.
Accordingly, to shift the phase of the signal passing through the waveguide 10, the resonant frequency of the structure 12 is lower than the frequency of the signal passing through the waveguide. The structure presents a capacitive impedance to the signal's E field, increasing the signals propagation constant and shifting its resonant frequency. For example, if waveguide 10 is passing a 44 GHz signal and has a structure 12 on its sidewalls 14, 16 that is designed to resonate at 35 GHz, the 44 GHz signal passing through the waveguide will experience a phase shift.
Numerous materials can be used to construct the impedance structure 12. The dielectric substrate 20 can be made of many dielectric materials including, but not limited to, plastics, poly-vinyl carbonate (PVC), ceramics, or high resistance semiconductor material such as Gallium Arsenide (GaAs), all of which are commercially available. Highly conductive material should be used for the conductive strips 18, conductive layer 24 and vias 22.
One embodiment of the structure 12 that resonates in response to a 35 GHz signal, comprises a dielectric substrate 20 of gallium arsenide (GaAs) that is 10 mils thick. The conductive strips 18 can be 1-6 microns thick with the preferred strips being 2 microns thick. The conductive strips 18 are 16 mils wide with a 1.5 mil gap etched between adjacent strips. The conductive layer 24 on the opposite side of the dielectric substrate 20 can also be 1-6 microns thick. Both the conductive layer 24 and the conductive strips 18 are preferably gold. The dimensions of the structure can change depending on the resonant signal frequency and the materials used. Accordingly, the above example is included for illustration purposes only and should not be construed as a limitation to this invention.
The structure 12 is manufactured by first vaporizing a layer of conductive material on one side of the dielectric material using any one of various known methods such as vaporization plating. Parallel lines of the newly deposited conductive material are etched away using any number of etching processes, such as acid etching or ion mill etching. The etched lines (gaps) are of the same width and equidistant apart, resulting in parallel conductive strips 18 on the dielectric material 20, the strips 18 having uniform width and a uniform gap between adjacent strips.
Holes are created through the dielectric material at uniform intervals. The holes can be created by various methods, such as conventional wet or dry etching. The holes are then filled or covered with the conductive material and outer surface of the dielectric material is covered with the conductive layer 24, both preferably accomplished using sputtered vaporization plating. The holes do not need to be completely filled, but their walls must be covered with the conductive material. The completed holes provide conductive vias 22 between the conductive layer 24 and the conductive strips 18.
Waveguide with Variable Phase Shifting
A second embodiment of the new waveguide phase shifter 40 according to the present invention is shown in
In operation, a voltage is applied to each conducting voltage strip 67. The diodes across the gaps on either side of the strip 48a are connected through the N+ layer 60. The ground for the voltage is provided strips 48 and the vias 55, to the conducting layer 56. The insulating layer 66 insulates the voltage strip 67 from the underlying via cap 65 to prevent strip from shorting to the via 55. A high voltage applied the voltage strips 67 reduces the capacitance of each diode 58 and reduces the capacitance across the gaps. The structure then resonates at a higher frequency. As the voltage is reduced, the capacitance across the gaps increases, decreasing the frequency at which the structure resonates. Increasing the voltage to a particular level can provide the desired shift in the beam's phase.
In fabricating the diodes 58, N+ layers 60 of a semiconductor material such as GaAs, are etched into mesas before the strips 48 are formed. The layer 60 runs along the gaps between the strips and will be partially below the strips 48 on each side of the gaps. The diodes 58 are then formed on the N+ layer 60, with both the N+ layer 60 and the diodes terminating short of the vias 54 and 55 and separated therefrom by intervening portions of the dielectric material. When the strips 48, insulating layer 66, coupling strip 68 and voltage strip 67 are formed, they extend over a diode 58 on each lateral side.
As shown in
In another embodiment of the new waveguide (not shown), all four walls of the waveguide 40 can have the impedance structure. The waveguide can then be used to shift the phase of either a vertically or horizontally polarized signal, or both. For a vertically polarized signal the impedance structures on the waveguides sidewalls 43, 45 shift the signal's phase. For horizontally polarized signals the structures on the waveguide's top and bottom walls 44, 46 shift the signal's phase.
Waveguide with Phase Shifter and Amplifier Array
The signal entering the waveguide encounters a first transition region 90 which is shown in more detail in
As shown by the graph in
The transition region is manufactured in a manner similar to the previous embodiments, except for etching the initially deposited conductive material to provide conductive strips with a narrowing gap between adjacent strips.
Referring back to
The beam then passes through a second transition region 104. This region is similar to the first transition region, but the gaps between the strips increase in the beam's direction. The frequency at which this structure resonates thus increases until at the end of the region it resonates at the beam frequency. At this location the beam has the desired phase shift and because the impedance structure is resonating, it also has uniform E and H fields.
The signal then enters the amplifier region 106. An array amplifier chip 108 is positioned within this section to amplify the signal from the second transition section 104. The amplifier region 106 has impedance structures mounted on all four waveguide walls to support both horizontal and vertical polarizations (cross polarized). A signal reaching the array amplifier chip 108 will have uniform E and H fields, and thus, equally drives each of chip's amplifiers. Array amplifier chips 108 are generally transmission devices rather than reflection devices, with the input signal entering one side and the amplified signal transmitted out the opposite side. This reduces spurious oscillations that can occur because of feedback or reflection of the amplified signal toward the source.
Array amplifiers chips also change the polarity of the signal 90° as it passes through as is amplified, further reducing spurious oscillations. However, a portion of the input signal carries through the array amplifier with the original input polarization. In addition, a portion of the output signal reflects back to the waveguide area before the amplifier. Thus, in amplifier section 106 both polarizations will exist.
The strip feature of the wall structures allows the amplifier section 106 to support a signal with both vertical and horizontal polarizations. The wall structure presents a high impedance to the transverse E field of both polarizations, maintaining the E field density across the waveguide for both. The strips allow current to flow down the waveguide in both polarizations, maintaining a uniform H field density across the waveguide for both. Thus, the cross polarized signal will have uniform density across the waveguide.
Matching grid polarizers 110 and 112 are mounted on each side of and parallel to the array amplifier chip 108. The polarizers appear transparent to one signal polarization, while reflecting a signal with an orthogonal polarization. For example, the output grid polarizer 112 allows a signal with an output polarization to pass, while reflecting any signal with an input polarization. The input polarizer 110 allows a signal with an input polarization to pass, while reflecting any signal with an output polarization. The distance of the polarizers from the amplifier can be adjusted, allowing the polarizers to function as input and output tuners for the amplifier, with the polarizers providing the maximum benefit at a specific distance from the amplifier.
Phase Shifting and Beam Steering Module
As shown in
The module 114 can be comprised of any of the above described waveguides. If waveguide 10 from
Using waveguide 40 from
If the waveguide 70 from
A portion of the incoming beam can reflect off the front edges of the waveguides 113, degrading the signal. To reduce this reflection, each of the waveguides can be provided with a launching region 120, beginning at the entrance to the waveguide 113 and continuing for a short distance down its length.
The launching regions resonate at the frequency of the beam entering the waveguides in the module. The vias which extend through the substrate present an inductive reactance (L), while the gaps between the patches present an approximately equal capacitive reactance (C) at the waveguides resonant frequency. The launching regions thus present parallel resonant high impedance L-C circuits to the beams E field component. The L-C circuits present an open-circuit to the E-field, allowing it to remain uniform across the waveguide. The low impedance on the top and bottom waveguide walls, which do not have impedance structures, allows current to flow and maintain a uniform H field.
The gaps between the patches 122 block surface current flow in all directions, preventing surface waves in the high impedance structures. This blocks TM and TE modes from entering the waveguide 112, admitting allowing TEM modes. Blocking the TM and TE modes reduces the front edge reflection with the front edge of the waveguide appearing nearly transparent to the beam at the resonant frequency.
The launching regions can be manufactured in a manner similar to the strip impedance structure. However, instead of etching the initially deposited conductive layer into strips, it is etched to form conductive patches.
The module can be used in various millimeter wave applications.
To steer the beam, the waveguides 113 shift the phase of their respective beam portions by different amounts, as described above. Each of the waveguides 113 can also have amplifier arrays to amplify the beam 147.
Although the present invention has been described in considerable detail with reference to certain preferred configurations thereof, other versions are possible. For example, the phase shifting and steering module can have different impedance structures and the module can be used in other applications. Therefore, the spirit and the scope of the appended claims should not be limited to their preferred versions described herein or to the embodiments in the above detailed description.
Number | Date | Country | |
---|---|---|---|
Parent | 10365031 | Feb 2003 | US |
Child | 11392274 | Mar 2006 | US |
Parent | 09676142 | Sep 2000 | US |
Child | 10365031 | Feb 2003 | US |