SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Planar submillimeter wave circuits are slowly replacing whisker-contacted devices at frequencies above 100 GHz, but in many cases the size constraints dictated by the short wavelengths found at high frequencies have not been adequately addressed.
Recent developments in semiconductor technology have enabled advanced submillimeter wave (300 GHz) transistors and circuits. These new high speed components have required new test methods to be developed for characterizing performance, and to provide data for device modeling to improve designs. Current efforts in progressing high frequency testing have resulted in on-wafer-parameter measurements up to approximately 340 GHz and swept frequency vector network analyzer waveguide measurements to 508 GHz. On-wafer noise figure measurements in the 270-340 GHz band have been demonstrated. In this letter we report on on-wafer power measurements at 330 GHz of a three stage amplifier that resulted in a maximum measured output power of 1.78mW and maximum gain of 7.1 dB. The method utilized demonstrates the extension of traditional power measurement techniques to submillimeter wave frequencies, and is suitable for automated testing without packaging for production screening of submillimeter wave circuits.
Proposed Quartz substrate Upside-down Integrated Device (QUID) relies on UV-curable adhesive to bond semiconductor with quartz. Integrated circuits including planar GaAs Schottky diodes and passive circuit elements (such as bandpass filters) fabricated on quartz substrates. Circuits designed to operate as mixers in waveguide circuit at millimeter and submillimeter wavelengths. Integrated circuits mechanically more robust, larger, and easier to handle than planar Schottky diode chips. Quartz substrate more suitable for waveguide circuits than GaAs substrate.
This oscillator uses a single-emitter 0.3- m InP heterojunction bipolar transistor (HBT) device with maximum frequency of oscillation (fmax) greater than 500 GHz. Due to high conductor and substrate losses at submillimeterwave frequencies, a primary challenge is to efficiently use the intrinsic device gain. This was done by using a suitable transmission-line media and circuit topology. The passive components of the oscillator are realized in a twometal process with benzocyclobutene (BCB) used as the primary transmission line dielectric. The circuit was designed using microstrip transmission lines. The oscillator is implemented in a common-base topology due to its inherent instability, and the design includes an on-chip resonator, outputmatching circuitry, and an injection-locking port, the port being used to demonstrate the injection-locking prin ciple. A free-running frequency of 311.6 GHz has been measured by down-converting the signal. Ad di tionally, injection locking has been successfully demonstrated with up to 17.8 dB of injection-locking gain. The injection-locking reference signal is generated using a 2 20 GHz frequency synthesizer, followed by a doubler, active tripler, a W-band amplifier, and then a passive tripler. Therefore, the source frequency is multiplied 18 times to obtain a signal above 300 GHz that can be used to injection lock the oscillator. Measurement shows that injection locking has improved the phase noise of the oscillator and can be also used for synchronizing a series of oscillators. A signal conductor is implemented near the BCP -InP interface and the topside of the BCB layer is fully metallized as a signal ground. Because the fields are primarily constrained in the lower permittivity BCB region, this type of transmission line is referred to as an inverted microstrip. In addition, both common-emitter and commonbase circuits were investigated to determine optimum topology for oscillator design. The common -base topology required smaller amount of feedback than the common-emitter design, therefore preserving device gain, and was chosen for the oscillator design. The submillimeter-wave region offers several advantages for sensors and communication systems, such as high resolution and all-weather imaging due to the short-wavelength, and improved communication speeds by access to greater frequency bandwidth. This oscillator circuit is a prototype of the first HBT oscillator operating above 300 GHz. Additional development is necessary to increase the output power of the circuit for radar and imaging applications.
A submillimeter-wave monolithic integrated- circuit (S-MMIC) amplifier has been designed and fabricated using an indium phosphide (InP) 35-nm gate-length high electron mobility transistor (HEMT) device, developed at Northrop Grumman Corporation. The HEMT device employs two fingers each 15 micrometers wide. The HEMT wafers are grown by molecular beam epitaxy (MBE) and make use of a pseudomorphic In0.75Ga0.25As channel, a silicon delta-doping layer as the electron supply, an In0.52Al0.48As buffer layer, and an InP substrate. The three-stage design uses coplanar waveguide topology with a very narrow ground-to-ground spacing of 14 micrometers. Quarter-wave matching transmission lines, on-chip metal-insulator-metal shunt capacitors, series thin-film resistors, and matching stubs were used in the design. Series resistors in the shunt branch arm provide the basic circuit stabilization. The S-MMIC amplifier was measured for S-parameters and found to be centered at 320 GHz with 13-15-dB gain from 300-345 GHz. This chip was developed as part of the DARPA Submillimeter Wave Imaging Focal Plane Technology (SWIFT) program (see figure). Submillimeter-wave amplifiers could enable more sensitive receivers for earth science, planetary remote sensing, and astrophysics telescopes, particularly in radio astronomy, both from the ground and in space. A small atmospheric window at 340 GHz exists and could enable ground-based observations. However, the submillimeter-wave regime (above 300 GHz) is best used for space telescopes as Earth s atmosphere attenuates most of the signal through water and oxygen absorption. Future radio telescopes could make use of S-MMIC amplifiers for wideband, low noise, instantaneous frequency coverage, particularly in the case of heterodyne array receivers.
The authors have designed, fabricated, and tested a quasi-optical submillimeter wave receiver with an Nb/AlO(x)/Nb tunnel junction. This design incorporates a hybrid antenna fed by a planar logarithmic spiral structure in order to couple to the radiation field from the telescope. The novel matching circuit requires several layers of photolithographic processing on top of the actual tunneling device. Computer modeling of the device correctly predicted the measured bandwidth and the characteristic frequencies to within 8 percent. A good match has been obtained from 200 to 475 GHz between the antenna and a relatively large area (1 sq mm) tunnel junction with omega(R(n))C about 2-4. Noise measurements at 318 GHz, 395 GHz, 426 GHz, and 492 GHz yielded uncorrected double sideband receiver noise temperatures of 200 K, 230 K, 220 K, and 500 K, respectively. Using the same optics, coupling efficiencies between the receiver and the Caltech Submillimeter Telescope were found to have values approaching those achieved by the best waveguide-horn-based receiver systems.
Fabrication technology for submillimeter-wave monolithic circuits has made tremendous progress in recent years and it is now possible to fabricate sub-micron GaAs Schottky devices on a number of substrate types, such as membranes, frame-less membranes or substrateless circuits. These new technologies allow designers to implement very high frequency circuits, either Schottky mixers or multipliers, in a radically new manner. This paper will address the design, fabrication, and preliminary results of a 1.2 THz planar tripler fabricated on a GaAs frame-less membrane, the concept of which was described previously. The tripler uses a diode pair in an antiparallel configuration similar to designs used at lower frequency. To date, this tripler has produced a peak output power of 80 microW with 0.9% efficiency at room temperature (at 1126 GHz). The measured fix-tuned 3 dB bandwidth is about 3.5%. When cooled, the output power reached a peak of 195 microW at 120 K and 250 microW at 50 K. The ease with which this circuit was implemented along with the superb achieved performance indicates that properly designed planar devices such as this tripler can now usher in a new era of practical very high frequency multipliers.
Connecting several tunneling diodes in series shows promise as a method for increasing the output power of these devices as millimeter-wave oscillators. However, due to the negative differential resistance (NDR) region in the dc I-V curve of a single tunneling diode, a circuit using several devices connected in series, and biased simultaneously in the NDR region, is dc unstable. Because of this instability, an oscillator with several tunneling diodes in series has a demanding excitation condition. Excitation using an externally applied RF signal is one approach to solving this problem. This is experimentally demonstrated using an RF source, both with frequency close to as well as with frequency considerably lower than the oscillation frequency. Excitation by an RF (radio frequency) source with a frequency as low as one sixth of the oscillation frequency was demonstrated in a proof-of-principle experiment at 2 GHz, for an oscillator with two tunnel diodes connected in series. Strong harmonics of the oscillation signal were generated as a result of the highly nonlinear dc I-V curve of the tunnel diode and a large signal oscillator design. Third harmonic output power comparable to that of the fundamental was observed in one oscillator circuit. If submillimeter wave resonant-tunneling diodes (RTD's) are used instead of tunnel diodes, this harmonic output may be useful for generating signals at frequencies well into the terahertz range.
The Terahertz (THz) region of the electromagnetic spectrum (about 300 - 3000 GHz in frequency or about 0.1 - 1 mm free space wavelength) has enormous potential for high-data-rate communications, spectroscopy, astronomy, space research, medicine, biology, surveillance, remote sensing, industrial process control, etc. It has been characterized as the most scientifically rich, yet under-utilized, region of the electromagnetic spectrum. The most critical roadblock to full exploitation of the THz band is lack of coherent radiation sources that are powerful (0.001 - 1.0 W continuous wave), efficient (> 1%), frequency agile (instantaneously tunable over 1% bandwidths or more), reliable, and comparatively inexpensive. To develop vacuum electron device (VED) radiation sources satisfying these requirements, fabrication and packaging approaches must be heavily considered to minimize costs, in addition to the basic interaction physics and circuit design. To minimize size of the prime power supply, beam voltage must be minimized, preferably 10 kV. Solid state sources satisfy the low voltage requirement, but are many orders of magnitude below power, efficiency, and bandwidth requirements. On the other hand, typical fast-wave VED sources in this regime (e.g., gyrotrons, FELs) tend to be large, expensive, high voltage and very high power devices unsuitable for most of the applications cited above. VEDs based on grating or inter-digital (ID) circuits have been researched and developed. However, achieving forward-wave amplifier operation with instantaneous fractional bandwidths > 1% is problematic for these devices with low-energy (< 15 kV) electron beams. Moreover, the interaction impedance is quite low unless the beam-circuit spacing is kept particularly narrow, often leading to significant beam interception. One solution to satisfy the THz source requirements mentioned above is to develop micromachined VEDs, or "micro-VEDs". Among other benefits, micro-machining technologies provide superior high frequency wall conductivity as a result of superior surface smoothness compared with conventional mechanical or electric discharge machining approaches. Micro-VED technologies are already being applied to the development of millimeter-wave klystrons at Stanford Linear Accelerator Center and submillimeter-wave klystrons at the University of Leeds. We are investigating the use of micro-machining technologies to develop THz regime TWTs, with emphasis on folded-waveguide TWTs. The folded-waveguide TWT (FW-TWT) has several features that make it attractive for THz-regime micro-VED applications. It is a relatively simple circuit to design and fabricate, it is amenable to precision pattern replication by micro-machining, and it is has been demonstrated capable of forward-wave amplification with appreciable bandwidth. We are conducting experimental and computational studies of micro-VED FW-TWTs to examine their feasibility for applications at frequencies from 200 - 1000 GHz.
The current generation of submillimeter-wave instruments is relatively mass and power-hungry. The receiver front ends (RFEs) of a submillimeter instrument form the heart of the instrument, and any mass reduction achieved in this subsystem is propagated through the instrument. In the current implementation, the RFE consists of different blocks for the mixer and LO circuits. The motivation for this work is to reduce the mass of the RFE by integrating the mixer and LO circuits in one waveguide block. The mixer and its associated LO chips will all be packaged in a single waveguide package. This will reduce the mass of the RFE and also provide a number of other advantages. By bringing the mixer and LO circuits close together, losses in the waveguide will be reduced. Moreover, the compact nature of the block will allow for better thermal control of the block, which is important in order to reduce gain fluctuations. A single waveguide block with a 600- GHz RFE functionality (based on a subharmonically pumped Schottky diode pair) has been demonstrated. The block is about 3x3x3 cubic centimeters. The block combines the mixer and multiplier chip in a single package. 3D electromagnetic simulations were carried out to design the waveguide circuit around the mixer and multiplier chip. The circuit is optimized to provide maximum output power and maximum bandwidth. An integrated submillimeter front end featuring a 520-600-GHz sub-harmonic mixer and a 260-300-GHz frequency tripler in a single cavity was tested. Both devices used GaAs MMIC membrane planar Schottky diode technology. The sub-harmonic mixer/tripler circuit has been tested using conventional metal-machined blocks. Measurement results on the metal block give best DSB (double sideband) mixer noise temperature of 2,360 K and conversion losses of 7.7 dB at 520 GHz. The LO input power required to pump the integrated tripler/sub-harmonic mixer is between 30 and 50 mW.