Showing papers on "Return loss published in 1988"

Optimum design of waveguide E-plane stub-loaded phase shifters

Abstract: Broadband low-insertion-loss E-plane stub-loaded rectangular waveguide phase shifters are designed with the method of field expansion into normalized eigenmodes, which includes higher-order mode interaction between the step discontinuities. Computer-optimized three-stub prototypes of 90 degrees differential phase shift with reference to an empty waveguide of appropriate length, designed for R140-band (12.4-18 GHz) and R320-band (26.5-40 GHz) waveguides, achieve typically +or-0.5 degrees phase shift deviation within about 20% bandwidth. For two-stub designs, the corresponding values are about +2.5 degrees /-1 degrees and 17%. Both designs achieve minimum return loss of 30 dB. The theory is verified by measurements of a compact R120-band (10-15 GHz) waveguide phase shifter design example milled from a solid block, showing measured insertion loss of about 0.1 dB and about +2.5 degrees /-0.5 degrees phase error between 10.7 and 12.7 GHz. >

A novel type of waveguide polarizer with large cross-polar bandwidth

Abstract: A wideband quarter-wave polarizer is presented having a rectangular cross-section, where all four walls are loaded with a dielectric or artificial dielectric. A much larger bandwidth compared to existing polarizers can be obtained without increasing the insertion loss. A polarizer so constructed was fabricated and measured. The polarizer had a differential phase shift within 90 degrees +or-0.7 degrees corresponding to 44 dB isolation, insertion loss below 0.06 dB, and return loss below -24 dB (VSWR >

The design of a 6-port active circulator

Abstract: An approach is presented for realizing circulators in the form of GaAs monolithic microwave integrated circuits (MMICs) without using any ferrite material and external magnets. The configuration is known as an active circulator and is suitable for system applications where a fully integrable, lightweight, and small-size subsystem is required. The circuit is capable of providing 0 dB loss, and return loss and isolation of better than 16 dB and 23 dB, respectively, over an octave or higher bandwidth. >

Measurements of the Magnex DC characteristics at microwave frequencies

Abstract: This paper deals with the measurements of the equivalent impedance, insertion loss (IL), and return loss (RL) of a conductive composite material called “Magnex DC”, performed at microwave frequencies. It is found that the equivalent impedance of this composite material decreases as the frequency increases in the X-band (8–12.4 GHZ) and the IL of a 1.4 mm thick specimen is greater than 9 dB over the whole band. The utilization of this material in electromagnetic shielding is considered. Furthermore, results of the measurements performed on a tapered specimen are also reported.

DC -50 GHz MMIC variable attenuator with a 30 dB dynamic range

Abstract: A description is given of a monolithic microwave integrated circuit (MMIC) FET variable attenuator demonstrating a 30-dB dynamic range of attenuation over a DC -50-GHz frequency band with a minimum insertion loss of 1.8 dB at 26.5 GHz and 2.6 dB at 40 GHz. The associated input/output return loss was always greater than 10 dB at any attenuation setting. An operational amplifier feedback scheme implemented around a DC reference circuit enabled automatic input/output impedance matching and linear attenuation with a single control voltage without any noticeable degradation of radiofrequency (RF) performance. The maximum attenuation increases with frequency from 32 dB at DC to 42 dB at 50 GHz. >

A High Isolation DC to 18 GHz Packaged MMIC SPDT Switch

Abstract: An high isolation MMIC GaAs SPDT switch was designed using a travelling wave configuration to achieve a very broadband performance from DC to 18 GHz. The SPDT switch die measurements give less than 2.5 dB insertion loss at 18 GHz and a return loss greater than 12 dB over the full bandwidth in both the ON and OFF states. The MMIC chip was packaged in a new 20 GHz package. The packaged switch has insertion loss of better than 3.1 dB an isolation higher than 34 dB at 18 GHz and a good match from DC to 18 GHz. The switch can handle an input power of 0.6 Watt from 1 GHz to 18 GHz.

A manufacturable, 26 GHz GaAs MMIC technology

Abstract: A 26 GHz MMIC technology has been developed to manufacture high performance integrated circuits for microwave instrumentation. The fabrication process features an MBE active layer, a 0.4 mu m gate patterned by DUV contact lithography, Ta/sub 2/N thin-film resistors, Si/sub 3/N/sub 4/ MIM capacitors and dry-etched backside vias. The completed MESFET has a nominal f/sub t/ of 23 GHz and a f/sub max/ of 55 GHz. Two standard products are currently being manufactured. The first is a 2-26.5 GHz travelling wave amplifier (TWA) which routinely achieves 8.7 dB+or-0.5 dB gain across the band with 21 dBm saturated output power and a noise figure of better than 8 dB at 18 GHz. RF yield to a gain spec of 7 dB at 26.5 GHz and 10 dB input and output return loss is 40%. The second product is a DC-50 GHz variable attenuator which achieves 72% yield to a 26.5 GHz spec of 2.4 dB minimum insertion loss, 27 dB maximum attenuation and greater than 8 dB input and output return loss. >

Design of Microstrip Exponentially Tapered Lines to Match Helical Antennas to Standard Coaxial Transmission Lines

Abstract: Based on the solution of Riccati equation a design of a microstrip exponentially tapered transmission line for matching two real impedances is worked out. An iterative procedure is used to yield the phase constant leading to an accurate design. The design is used to match a helix to a coaxial cable. A comparison of theoretical and experimental results shows an excellent agreement. It is found that small adjustments in the return loss and flattening of the frequency response is possible with a metallic tuning foil.

Wideband and low return loss coplanar strip feed using intermediate microstrip

Abstract: A wideband and low return loss coplanar strip feed has been achieved using an intermediate microstrip line between coax and coplanar strip lines. The measured and calculated return loss are below 13 dB up to 18 GHz. An extended calculation using TDR data shows the return loss below 10 dB up to 30 GHz. This coplanar strip feed is very practical for feeding wideband electro-optic devices which use coplanar strip as the electrodes.

On the design of an active circulator

Abstract: Design of a fully monolithically integrable active circulator without any externally added circuit components is described in this paper. Two different configurations for active circulators using power dividers and Lange couplers are investigated. The circuit using Lange couplers indicates superior performance and is capable of providing O-dB loss and return loss and isolation better than 16 dB and 23 dB, respectively, over an octave or higher bandwidth.

Design and performance of diode laser and photodetector microwave packages

Abstract: Fiber-coupled diode laser and photodetector microwave packages with performance extending to 20 GHz have been developed for optical tranmission of broadband analog and digital signals. These units provide all the features expected in field-deployable hermetic modules. The laser package1 incorporates a thermoelectric heat pump, thermistor, and photodetector for laser stabilization and is operable to an ambient temperature of 60°C. Microstrip coupling from the rf port to the laser chip minimizes parasitic elements and allows either direct connection to the laser or 50-Ω impedance matching over the entire band with a minimum return loss of 10 dB. The laser is coupled to a single-mode tapered lensed fiber fly-lead with solder-fixed fiber alignment. The photodetector package utilizes a substrate-illuminated mesa-geometry InGaAs PIN photodiode which terminates the rf port of the package (Fig. 1).

Highly stable low-insertion- and high-return-loss PC optical fiber connectors

Abstract: Highly stable physical-contact single-mode optical fiber connectors with low insertion-loss of 0.07 dB and high return-loss of 32.5 dB have been realized, by optimum design of ferrule end structure.

Confocal laser package optimized for minimum reflections and maximum coupling

Abstract: The coupling of reflections to semiconductor lasers can have a catastrophic impact on lightwave system performance at high bit rates. We explore the means by which an adequate return loss can be achieved in a laser package with discrete optics without excessive penalties due to aberrations.

Analysis of microstrip line to waveguide end launchers

Abstract: The expression for impedance is derived using the self-reaction concept with the assumption of a sinusoidal current distribution existing in the conductor loop. Three different cases of end launchers are computed for their corresponding input impedances. Comparison between calculated and measured input return loss of an end launcher shows good agreement between theory and experiment at Ka-band frequencies. >

Ridge‐shaped narrow‐wall directional coupler. Tight coupling and high‐power handling capability

Abstract: A new ridge-shaped narrow-wall directional coupler is proposed which can realize a tight coupling and a high-power handling capability. In the proposed coupler, the right-angled corners of the coupling slot of the conventional ridge-shaped directional coupler are rounded off so that a tight coupling characteristic can be obtained in a broad frequency range. The new structure preserves the features of the conventional ridge-shaped coupler, that is: (1) flexibility of the spatial arrangement of the two coupling waveguides, and (2) the possibility as a variable power divider. The boundary element method has been used for the electromagnetic analysis. The obtained performances are such that the coupling is 1.5±0.5 dB, the return loss is more than 20 dB, and the isolation is more than 20 dB in the frequency range of 27 to 33 GHz. The measured coupling characteristics agree well with the computed results.

Characteristics of fibre-coupled optical isolators at 1.5μm

Abstract: Compact, fibre-coupled Faraday optical isolators have been assembled. Isolators with both conventional and polarisation maintaining monomode fibre tails have been fully characterised in insertion loss, return loss and isolation ratio over the wavelength range 1480-1560 nm. >

Two-wire/four-wire conversion circuit

Abstract: PURPOSE:To increase the quantity of mismatching attenuation on the side of two-wire line by setting the impedance on the side of a four wire to a pure resistance and providing an equilibrium connection network on an infiltrating cancellation line part. CONSTITUTION:For matching the load impedance 12 of the two-wire analogue line in an analogue trunk 11, the impedance 13 on the side of the four-wire is set to the pure resistance, and the equilibrium connection network 14 which prevents a reception signal from a four-wire digital channel 10 form infiltrating into the channel 10 is provided in the infiltrating cancellation line part 15 with the least influence of the impedance on the pure resistance. Thus, the equilibrium connection network 14 does not influence the quantity of mismatching attenuation on the side of the two-wire, and the input impedance 13 on a return loss on the side of the four-wire. Since a two-wire/four-wire conversion circuit can be designed without the consideration of influence on both, the quantity of mismatching attenuation on the side of the two-wire can freely be increased.

Microwave behavior of electroless microstrip lines on SI GaAs substrates

Abstract: A technique of fabrication and evaluation of microstrip lines onto undoped semi-insulating (SI) GaAs substrates is described The fabrication technique is based on the electroless plating of nickel flashed with gold The evaluation includes the measurements of transmission and reflection properties of microstrip lines The measured data for insertion and return loss are encouraging

Polarization adaptor

Abstract: The design and theory of a passive microwave polarization adaptor (PA) for very small aperture terminal (VSAT) use are described. The PA changes two orthogonally polarized signals into two copolarized signals. The PA consists of two filters, a twisted waveguide, an orthomode transducer, a transition adaptor, and a 90-degree waveguide bend. A prototype has been designed with an insertion loss of -0.5 dB and return loss of -18 dB. Test results for the device are given. >

Design and Performance of Broadband Circulator

Abstract: Broadband miniaturized stripline junction circulator (operating in a below resonance mode) using lithium ferrite is discussed. The design criteria and the advantages of this device over the other type namely EGW circulator are outlined. The device operates over a bandwidth of 8 to 18 GHz with 15 dB (min) isolation, insertion loss of than 0.6 dB and return loss 15 dB minimum.

A Universal Test Fixture For Characterizing MM-Wave Solid State Devices Using A Novel Deembedding Procedure

Abstract: The design and evaluation of a novel fixturing technique for characterizing mm-wave solid state devices, including monolithic microwave integrated circuits, is presented. The technique utilizes a cosine tapered ridge guide fixture and a one-tier deembedding procedure to provide accurate and repeatable device level data. Advanced features include "nondestructive" testing, full waveguide bandwidth operation, a high degree of universal ism, and rapid yet repeatable chip-level characteriza­ tion. Furthermore, only one set of calibration standards is required regardless of device geometry. INTRODUCTION Accurate characterization of solid state devices at millimeter wavelengths is hindered by limited metrological techniques due to unpredictable parasitics, diminutive physical dimensions, and inconvenient interface methods (1). Typically, devices are characterized by S-parameters using an automatic network analyzer (ANA). To be accessible to the ANA and amenable to characterization, the device must be mounted in a fixture. This fixture introduces substantial insertion and return loss effects, both in magnitude and phase, often masking the true device characteristics. Hence, it would be desirable to devise a mathematical scheme to reposi­ tion the reference (measurement) plane adjacent to the device terminals. This procedure is known as deembedding and requires chip level microstrip stan­ dards to shift the reference plane from the fixture- to-ANA interface to the device area.Drawbacks of existing fixtures include non- repeatability, fixturing and calibration complexity, and limited bandwidth. Most techniques require cum­ bersome device mounting arrangements and are labor intensive. The fixture presented here utilizes a cosine tapered ridge guide transition to match the high waveguide impedance to the 50 Q microstrip. The taper length and profile have been analytically determined to provide maximum bandwidth and minimum insertion loss. Discrete devices and integrated circuits are mounted on a customized carrier which is inserted between the bottom of the ridge and the base of the waveguide. The electric field is con­ centrated by the ridge and launched onto the carrier by direct pressure contact. Control voltages are applied to the carrier using a self-aligned bias module with integral spring loaded contacts. Deem­ bedding is accomplished using a modified "through, short, delay" (TSD) calibration algorithm.DESIGNSince the fixture itself is mechanically rigid, the required versatility was built into the carrier.A 0.020 in. alumina substrate was chosen for ease of manufacturability, good dielectric properties, and relatively good thermal and mechanical characteris­ tics. The thickness was constrained to a narrow range due to certain electrical features and a con­ cern about the generation of higher order modes. DC blocks are required since the carrier microstrip is in contact with the fixture body. A coupled line section 0.032 in. long with 0.0029 in. wide fingers and 0.002 in. gaps was found to be optimum (2,3). Bias can be fed directly to the chip input and out­ put microstrip lines through two high RF impedance bias filters. The chip is mounted in a laser machined well in the center of the carrier and is supported by a gold plated kovar subcarrier.The ridge waveguide transition can be described by dividing the design into three calculations, first outlined in (4). First, the ridge waveguide cutoff wavelength must be obtained. At cutoff, propagation may be considered to be restricted to a wave traveling transversely across the guide with no longitudinal component. The cutoff wavelength is derived by solving for the input impedance of a simple equivalent network (5). Next, the ridge thickness is determined by evaluating the infinite frequency characteristic impedance for the TE10 mode as given in (6). Finally, the profile of the physi­ cal taper must be determined. A sinusoidal varia­ tion is assumed since this geometry outperforms most other candidates in terms of optimizing return loss in the shortest electrical length. Based on the above calculations, a taper length of 1.366 in. and a ridge thickness of 0.0695 in. was chosen.Due to its practical implementation in micro- strip, a TSD calibration routine was chosen for deembedding. The technique, first proposed by Franzen and Special (7) used a one-tier calibration approach and assumes an eight term error model. Based on measurements of the three standards, all 16 terms (magnitude and phase) of the error net­ works can be determined. The topography of the model consists of three cascaded two port networks; hence, the S-parameters of the error networks and device are converted to T-parameters for mathemati­ cal convenience. The T-matrix of the model is equal to the product of the individual T matrices; there­ fore, the matrix of the deembedded device can be extracted and reconverted to S-parameters (8). Sub­ sequent to the initial calibration, the ancillary reference plane is mathematically shifted to the device boundaries through the translation exp(-y&). Gamma (y), the complex propagation constant, is determined from the "through" and "delay" measure­ ments. The physical length of the device, Q,, must be known. It is this feature which permits the use of only one set of calibration standards.