Newsletter 5.5

Antenna Magus Version 5.5 released!

Version 5.5 sees a number of extensions to Antenna Magus features as well as the addition of new transitions and antennas. Though there are too many new additions to discuss in this newsletter, we will briefly highlight some of the feature extensions and new designs that have been made available. More information on these can be found on the Antenna Magus website and in the full release notes.

New features

Besides the improvement and updating of many export models to support the latest capabilities available in the simulation tools supported by Antenna Magus, a number of small feature extensions have been added.

The updated Chart Tracer

The Chart Tracer tool is used to help Antenna Magus Users trace plots from measurements or papers and convert them into digitized traces. In addition to tracing the data, relevant type-specific properties of the traced data are calculated. These derived values can be used to quickly populate value-based specifications for design in the Specification library.

In the Version 5.5 update of Antenna Magus, the Chart Tracer tool has been converted to use a ribbon menu interface. An intelligent locking mechanism has been added to allow the overlay and tracing of charts with different aspect ratios and axis ranges and types. A number of new derived quantity types have also been added.

The Chart Tracing tool with a ribbon menu interface and new secondary side-lobe derived quantities shown

Find Mode

In Find Mode, a button has been added to the active keywords making it possible to find templates that do not match that specific keyword. In the image on the left, for example, a search for Medium gain antennas that are not Horns, Patches or Arrays is now possible.

The usage of the Not button in Find Mode to exclude templates that do not have as specific keyword

New Antennas

In addition to the following new antennas, a diverse set of new transitions have also been added. The new designs are:


  • Quad-ridged wide-band conical horn antenna
  • 2 x 2 array of wire helix antennas
  • Cylindrical Luneberg lens (with pyramidal horn feed)
  • Spherical Luneberg lens (with pyramidal horn feed)
  • Planar elliptical dipole antenna


  • Microstrip to waveguide transition
  • Microstrip quadrature hybrid coupler
  • Microstrip ‘rat race’ coupler
  • Microstrip radial stub band-stop filter
  • Coax to circular waveguide transition
  • Coax to microstrip line transition
  • Coax to quad-ridged waveguide transition
  • Coax to coplanar waveguide (CPW) transition

2x2 array of wire helix antennas
Image of the 2x2 array of wire helix antennas.

Sequentially rotating elements to form an array was originally investigated for increasing the main-beam axial ratio of patch antennas [Kraft].Using this technique with an axial mode helix antenna, the gain and main-beam polarisation is improved w.r.t a single element.

This axial-mode helix array consists of four identical helices placed in a 2-by-2 grid, rotated sequentially about their axis, and fed sequentially out of phase.

Typical polarisation specific gain pattern at the centre frequency

A comparison of the gain versus angle and the axial ratio versus frequency on broadside between a single axial-mode helix and a sequentially rotated array of axial-mode helices illustrates the performance improvement.

Typical normalised circularly polarised gain versus angle pattern at the centre frequency of a sequentially rotated array compared to a single element
Typical axial ratio on boresight versus normalised frequency of a sequentially rotated array compared to a single element

U. R. Kraft, "Main-beam polarization properties of four-element sequential-rotation arrays with arbitrary radiators", IEEE Transactions on Antennas and Propagation, vol. 44, no. 4, pp. 515-522, 1996.

Quad-ridged conical horn antenna
Image of the Quad-ridged conical horn antenna.

The quad-ridged horn or quadruple-ridge flared horn (QRFH) has two important characteristics: (1) a nearly constant beamwidth over a 5:1 frequency band for nominal 10 dB beamwidths ranging between 50 Ohm and 140 Ohm (2) the input impedance can be designed between 50 Ohm and 100 Ohm. The horn is also popular in weight and cost-sensitive applications where multi-polarisation capable antennas are used, as only one single-ended low-noise amplifier (LNA) is required per polarisation.

The QRFH may be divided into two distinct sections - the coaxial-to-quad-ridged-waveguide transition and the flared ridge section. The basic working of the transition is to convert the input coaxial TEM mode to a TE11 aperture field distribution at the flare-transition intersection (commonly called the throat of the horn). Only the dominant TE11 mode should be present and all unwanted higher order modes should be suppressed. The purpose of the flare of the QRFH is to generate the higher-order modes required to achieve a good distribution in the horn aperture from the dominant TE11 mode incident at the throat.

The operational bandwidth of the QRFH is limited by the point at which the contribution of the secondary mode (TE12) excited by the coaxial feed in the transition becomes considerable enough to impact on the aperture distribution - and therefore the radiation pattern of the horn.

The performance shown below is for a 100 Ohm input resistance and a nominal 10 dB beamwidth of 75 degrees. The nominal beamwidth is taken as the average between the 10 dB beamwidths in the two orthogonal planes. The radiation performance for two different polarisations are plotted: dual-linear and right-hand circularly polarised. The radiation pattern is more symmetrical when excited for circular polarisation.

Total linear gain 3D patterns at (a) 0.5 x f0, (b) f0, and (c) 1.5 x f0
Total circular gain 3D patterns at (a) 0.5 x f0, (b) f0, and (c) 1.5 x f0

The graph below shows the relationship between the flare length and the flare diameter on the nominal beamwidths. For this illustration the flare diameter was set to a wavelength at the minimum frequency (0.5 x frequency centre). The results show that as the flare length becomes longer (ratio becomes larger), the nominal beamwidth at the centre frequency drops and the change in beamwidth response is more gradual versus frequency.

Nominal 10 dB beamwidth for different flare-length-to-flare-diameter ratios
Cylindrical Luneberg lens antenna
Image of the Cylindrical Luneberg lens antenna.

The conventional spherical Luneburg lens can be redesigned using a coordinate transformation to change the size and shape for integration into planar microwave applications - resulting in a cylindrical lens shape.

The cylindrical Luneburg does not achieve gains higher than 22 dBi as easily as the spherical Luneburg lens and also has reduced performance i.t.o. side lobe level, front-to-back ratio and return loss. By moving the feed parallel to the lens, the beam can be steered. Steering the beam may, however, cause a drop in the gain and a further increase in side lobe level.

Typical radiation pattern at the centre frequency
Normalised radiation pattern cuts at the centre frequency
Spherical Luneberg lens antenna
Image of the Spherical Luneberg lens antenna.

The Luneburg lens - a type of gradient-index (GRIN) lens, proposed by R. K. Luneburg in 1944 - is a spherical lens that focuses a plane wave to a point on the lens surface.

Luneburg lenses exhibit broadband behaviour and have many useful applications in antenna applications as they can form multiple beams in arbitrary directions due to their symmetry. The disadvantages of the spherical Luneburg lenses include lens weight and size as well as the manufacturing complexity.

For an ideal Luneburg lens, the relative permittivity must decrease continuously from the centre of the lens to 1 on the surface. In practice Luneburg lenses are constructed using a series of concentric dielectric shells whose permittivity is discretely varied. The implementation in Antenna Magus consists of 6 concentric shells and the lens is fed using a waveguide-fed pyramidal horn.

The lens can produce a high gain, symmetrical pencil beam with low side lobes and good cross-polarisation.

The figures below show patterns for a 20 dBi design.

Typical radiation pattern at the centre frequency
Normalised radiation pattern cuts at the centre frequency
Planar elliptical dipole
Image of the Planar elliptical dipole.

The elliptical dipole antenna may find application for use in the FCC (or ETSI) defined ultra-wideband (UWB) radio band of 3.1 - 10.6 GHz. While it may not be the optimal structure to serve as a Gaussian impulse shaping filter required for UWB radio, it is a relatively simple, widely used antenna which finds application in a number of devices.

The elliptical dipole consists of 2 elliptical discs excited at their closest point. Achieving a practically realizable implementation of this excitation without impacting on the radiation pattern requires some care. Since the antenna pattern displays nulls in the axial direction, a microstrip-to-balanced stripline (also known as parallel strips line) feed is used, as shown below.

Practical implementation with microstrip-to-balanced stripline feed

The antenna achieves a good match (S11 < -10 dB) over a > 6:1 bandwidth for a 100 Ohm input impedance. In this band, the radiation pattern changes from an omnidirectional pattern at low frequencies, to a more erratic multipath shape at higher frequencies.

While various ratios of ellipse width-to-length are possible, a 1:1 ratio is also common.

Typical reflection coefficient of an elliptical dipole
Radiation pattern of the elliptical dipole at various frequencies in the operating band
Coax to Quad-ridged waveguide transition
Image of the Coax to Quad-ridged waveguide transition.

The quad-ridged circular waveguide transition consists of a circular waveguide with 4 ridges positioned in a cross shape. The two orthogonal coaxial feeds have the advantage of being able to support a fundamental propagation mode polarised in either of two orthogonal directions.

This transition converts the coaxial TEM input mode to the fundamental TE11 mode at the throat/transition output. The throat is typically connected to a quad-ridged flare to create a quad-ridged flared horn (QRFH) antenna. The flare acts as a good matching device between free-space and the fundamental TE11 mode present at the throat interface.

The transition is wideband with an operational bandwidth of 3:1. This bandwidth is limit by the cut-off frequency of the TE11 and the frequency where the secondary mode, TE12, starts to propagate. In order for the aperture distribution of the QRFH to be correct, the TE12 may not be present at the throat-flare interface.

Typical S-parameter response versus frequency
Coax circular waveguide transition
Image of the Coax circular waveguide transition.

Although coaxial to rectangular waveguide transitions are far more common than coaxial to circular waveguide transitions, circular waveguides are easier to manufacture, implement and is often used to feed conical horns. In radar applications where rotary joints are used a circular cross section is necessary.

The circular waveguide can be matched to a coaxial line for a range of impedances by the appropriate choice of probe height and inset from the waveguide's short-circuited end.

The circular waveguide's dominant mode is the TE11 mode with a maximum practical bandwidth of 12.4 % due to cutoff frequency of the TM01 as the coaxial feed does also excite the TM01 mode.

The figures below show the reflection and transmission coefficients for a 50 Ohm design.

Typical reflection coefficient versus normalised frequency.
Typical transmission coefficients versus normalised frequency.
Coax to CPW transition
Image of the Coax to CPW transition.

The conventional method of connecting a coaxial cable to CPW is co-planar, i.e. end fed. Problems arise with this connection method when a large dimensional discrepancy between the coaxial inner conductor and the inner conducting strip of the CPW exists.

The coax to CPW transition addition to Antenna Magus has a different topology to the above mentioned. The coaxial connector is mounted on the underside of a substrate with the centre conductor extending through the substrate and is attached in a perpendicular manner to the central conducting strip of the CPW. To overcome the dimensional difference between the centre conductors of the CPW and the coax, a small tab like structure is used to form a linear taper between the two. The outer conductor of the coax is modified, and only a section of it extends through the substrate and is attached to the groundplane.

Modified coaxial connector and tab structure (substrate is removed for clarity)

The perpendicular connection allows the high-frequency electric field of the coax to directly couple, in a uniform manner, on the same plane with the high-frequency electric field in the gap between the CPW centre conductor and the groundplane. This modification in physical geometry allows minimal mismatch and ultra-wideband (where S11 < -10 dB and S21 > -1 dB) with the upper limit dependent on the exact physical parameters and geometry of the coaxial connector used. Antenna Magus has coaxial connector two design options: a practical coax, for designs below 20 GHz, and a frequency scaled connector.

Typical S-parameter response versus frequency
Coax to microstrip transition
Typical S-parameter response versus frequency

The transition between a coaxial transmission line and a microstrip line is a vital part in the interconnection of microwave circuits. It is one of the simplest microstrip transitions and because both media supports TEM mode it is generally broadband. Two configurations are widely used. The first is the end-launch or edge-mounted configuration where the axis of the coaxial line is aligned with the end of the microstrip line. The second is the vertical-mounted configuration where the coaxial line is perpendicular to the substrate of the microstrip line. The latter is the one implemented in Antenna Magus

The coaxial-to-microstrip line has an ultra-wideband performance (where S11 < -10 dB and S21 > -1 dB) with the upper limit dependent on the exact physical parameters and geometry of the coaxial connector used. Magus, however, does not model a specific coaxial connector.

Typical reflection and transmission coefficient behaviour versus frequency for a 50 Ohm port impedance and epsr = 2 and 2.2 for the coaxial cable and substrate respectively.
Quad hybrid microstrip coupler
Image of the Quad hybrid microstrip coupler.

The quadrature hybrid may be implemented in microstrip or stripline, and is also known as a branch-line coupler. It is a four-port network which produces a 90 degree phase shift between the two output ports, hence the name "90 degree hybrid junction".

With reference to the port numbering in the image below - when fed at port 1, two equal signals are received at ports 2 and 3 with a 90 degree phase shift between them; while good isolation is achieved between the input and output ports (1 and 4). Unequal power division may be achieved by using different impedances at the output ports.

The quadrature hybrid with its relevant line lengths and impedances

Since the quadrature hybrid has a high degree of symmetry, any port may be used as the input port. The output ports will always be on the opposite side of the junction, and the isolated port will be the remaining port on the same side as the input port.

Typical reflection coefficient behavior versus frequency
Typical phase behavior versus frequency
Microstrip radial stub
Image of the Microstrip radial stub.

Microstrip circuits, such as low-pass filters and mixers, often require stubs with a low impedance level and an accurate localisation of a zero-point impedance [Giannini and Sorrentini et al.], [Giannini and Ruggieri et al.]. Shunt-connected quarter-wavelength rectangular microstrip lines can be used for this purpose, but suffer from narrow impedance bandwidths. The bandwidth can be increased by increasing the rectangular stub width, but this can lead to the generation of higher order modes and also result in a badly defined insertion point on the microstrip line. Radial stubs provide low impedance levels with a well-defined insertion point over a wide frequency range [Sorrentino and Roselli], [Atwater].

Typical transmission coefficient versus frequency

R. Sorrentino and L. Roselli, "A new simple and accurate formula for microstrip radial stub", IEEE Microwave and Guided Wave Letters, vol. 2, no. 12, pp. 480 - 482, 1992.

F. Giannini, R. Sorrentino, and J. Vrba, "Planar Circuit Analysis of Microstrip Radial Stub", MTT-S International Microwave Symposium Digest, vol. 32, no. 12, pp. 1652 - 1655, 1984.

F. Giannini, M. Ruggieri, and J. Vrba, "Shunt-Connected Microstrip Radial Stubs (Short Paper)", IEEE Transactions on Microwave Theory and Techniques, vol. 34, no. 3, pp. 363 - 366, 1986.

H. A. Atwater, "Microstrip reactive circuit elements", IEEE Transactions on microwave theory and techniques, vol. 31, no. 6. pp. 488 - 491, 1983.

Ratrace microstrip coupler
Image of the Ratrace microstrip coupler.

The rat-race may be considered as the microstrip equivalent of the waveguide Magic-T already designed by Antenna Magus. It is a four-port network which produces a 180 degree phase shift between the two output ports, hence the name "180 degree hybrid junction".

With reference to the port numbers shown below - when fed at port 1 (also known as the Sum port), two equal and in-phase signals are received at ports 2 and 3; while when fed at port 4 (also known as the Difference port), two equal but out-of-phase signals are received at ports 2 and 3. Good isolation is achieved between the two input ports (1 and 4).

The rat-race ring hybrid with its relevant line lengths and impedances
Typical reflection coefficient behaviour versus frequency with an input applied at port 1
Typical phase behaviour versus frequency with an input applied at port 1
Typical reflection coefficient behaviour versus frequency with an input applied at port 4
Typical phase behaviour versus frequency with an input applied at port 4
Microstrip to waveguide transition
Image of the Microstrip to waveguide transition.

The transition between a microstrip line and a rectangular waveguide is accomplished by inserting a probe, connected to the microstrip line on the same substrate, into a slot in the broad wall of a rectangular waveguide. These transitions are useful for the device and circuit characterization of millimeter-wave MICs and MMICs. Some systems may also require transitions between microstrip and waveguide to combine IC’s and waveguide components.

Operation of the transition is similar to that of a coaxial to waveguide transition where the extension of the probe couples strongly to the TE10 waveguide mode by acting as a short capacitive monopole.

Typical reflection coefficients versus frequency
Typical transmission coefficient versus frequency