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.
The progress of the Jet Propulsion Laboratory in developing gallium arsenide junction field-effect transistors (GaAs JFETs) for application in infrared readout electronics operating below 10 Kelvin is discussed.
This paper presents ceramic substrates and thick-film metallization based packaging technologies in development for 500C silicon carbide (SiC) electronics and sensors. Prototype high temperature ceramic chip-level packages and printed circuit boards (PCBs) based on ceramic substrates of aluminum oxide (Al2O3) and aluminum nitride (AlN) have been designed and fabricated. These ceramic substrate-based chip-level packages with gold (Au) thick-film metallization have been electrically characterized at temperatures up to 550C. The 96 alumina packaging system composed of chip-level packages and PCBs has been successfully tested with high temperature SiC discrete transistor devices at 500C for over 10,000 hours. In addition to tests in a laboratory environment, a SiC junction field-effect-transistor (JFET) with a packaging system composed of a 96 alumina chip-level package and an alumina printed circuit board was tested on low earth orbit for eighteen months via a NASA International Space Station experiment. In addition to packaging systems for electronics, a spark-plug type sensor package based on this high temperature interconnection system for high temperature SiC capacitive pressure sensors was also developed and tested. In order to further significantly improve the performance of packaging system for higher packaging density, higher operation frequency, power rating, and even higher temperatures, some fundamental material challenges must be addressed. This presentation will discuss previous development and some of the challenges in material science (technology) to improve high temperature dielectrics for packaging applications.
Theoretical analysis and some experiments have shown that the silicon-on-insulator (SOI) 4-gate transistors known as G4-FETs can be used as building blocks of four-quadrant analog voltage multiplier circuits. Whereas a typical prior analog voltage multiplier contains between six and 10 transistors, it is possible to construct a superior voltage multiplier using only four G4-FETs. A G4-FET is a combination of a junction field-effect transistor (JFET) and a metal oxide/semiconductor field-effect transistor (MOSFET). It can be regarded as a single transistor having four gates, which are parts of a structure that affords high functionality by enabling the utilization of independently biased multiple inputs. The structure of a G4-FET of the type of interest here (see Figure 1) is that of a partially-depleted SOI MOSFET with two independent body contacts, one on each side of the channel. The drain current comprises of majority charge carriers flowing from one body contact to the other that is, what would otherwise be the side body contacts of the SOI MOSFET are used here as the end contacts [the drain (D) and the source (S)] of the G4-FET. What would otherwise be the source and drain of the SOI MOSFET serve, in the G4-FET, as two junction-based extra gates (JG1 and JG2), which are used to squeeze the channel via reverse-biased junctions as in a JFET. The G4-FET also includes a polysilicon top gate (G1), which plays the same role as does the gate in an accumulation-mode MOSFET. The substrate emulates a fourth MOS gate (G2). By making proper choices of G4-FET device parameters in conjunction with bias voltages and currents, one can design a circuit in which two input gate voltages (Vin1,Vin2) control the conduction characteristics of G4-FETs such that the output voltage (Vout) closely approximates a value proportional to the product of the input voltages. Figure 2 depicts two such analog multiplier circuits. In each circuit, there is the following: The input and output voltages are differential, The multiplier core consists of four G4- FETs (M1 through M4) biased by a constant current sink (Ibias), and The G4-FETs in two pairs are loaded by two identical resistors (RL), which convert a differential output current to a differential output voltage. The difference between the two circuits stems from their input and bias configurations. In each case, provided that the input voltages remain within their design ranges as determined by considerations of bias, saturation, and cutoff, then the output voltage is nominally given by Vout = kVin1Vin2, where k is a constant gain factor that depends on the design parameters and is different for the two circuits. In experimental versions of these circuits constructed using discrete G4- FETs and resistors, multiplication of voltages in all four quadrants (that is, in all four combinations of input polarities) was demonstrated, and deviations of the output voltages from linear dependence on the input voltages were found to amount to no more than a few percent. It is anticipated that in fully integrated versions of these circuits, the deviations from linearity will be made considerably smaller through better matching of devices.
The fabrication and testing of the first semiconductor transistors and small-scale integrated circuits (ICs) to achieve up to 3000 h of stable electrical operation at 500 C in air ambient is reported. These devices are based on an epitaxial 6H-SiC junction field-effect transistor process that successfully integrated high temperature ohmic contacts, dielectric passivation, and ceramic packaging. Important device and circuit parameters exhibited less than 10% of change over the course of the 500 C operational testing. These results establish a new technology foundation for realizing durable 500 C ICs for combustion-engine sensing and control, deep-well drilling, and other harsh-environment applications.
A solid state amplifier has an output transistor stage and an input transistor stage interconnected as a Darlington circuit. An interstage transistor interconnecting the collectors of the two stages of the Darlington circuit is connected with the first stage transistor in a compound configuration. When the load current furnished by the amplifier is less than a predetermined value, the output stage operates as a simple saturated switch with its base drive current flowing through the emitter-base junction of the first stage which produces no collector current because the base-emitter junction of the interstage transistor is reverse-biased. When the load current exceeds said predetermined value, the interstage transistor begins to conduct allowing the input and output stages to operate as a Darlington amplifier and diverting a significant portion of the base drive current of the output transistor back into the load.
It is possible to synthesize a voltage-controlled negative-differential-resistance (NDR) device or circuit by use of a pair of complementary G4FETs (four-gate field-effect transistors). [For more information about G4FETs, please see the immediately preceding article]. As shown in Figure 1, the present voltage-controlled NDR device or circuit is an updated version of a prior NDR device or circuit, known as a lambda diode, that contains a pair of complementary junction field-effect transistors (JFETs). (The lambda diode is so named because its current-versus- voltage plot bears some resemblance to an upper-case lambda.) The present version can be derived from the prior version by substituting G4FETs for the JFETs and connecting both JFET gates of each G4FET together. The front gate terminals of the G4FETs constitute additional terminals (that is, terminals not available in the older JFET version) to which one can apply control voltages VN and VP. Circuits in which NDR devices have been used include (1) Schmitt triggers and (2) oscillators containing inductance/ capacitance (LC) resonant circuits. Figure 2 depicts such circuits containing G4FET NDR devices like that of Figure 1. In the Schmitt trigger shown here, the G4FET NDR is loaded with an ordinary inversion-mode, p-channel, metal oxide/semiconductor field-effect transistor (inversion-mode PMOSFET), the VN terminal of the G4FET NDR device is used as an input terminal, and the input terminals of the PMOSFET and the G4FET NDR device are connected. VP can be used as an extra control voltage (that is, a control voltage not available in a typical prior Schmitt trigger) for adjusting the pinch-off voltage of the p-channel G4FET and thereby adjusting the trigger-voltage window. In the oscillator, a G4FET NDR device is loaded with a conventional LC tank circuit. As in other LC NDR oscillators, oscillation occurs because the NDR counteracts the resistance in the tank circuit. The advantage of this G4FET-NDR LC oscillator over a conventional LC NDR oscillator is that one can apply a time-varying signal to one of the extra control input terminals (VN or VP) to modulate the conductance of the NDR device and thereby amplitude-modulate the output signal.
The use of solid state power amplifiers (SSPA) in satellites has been quite prevalent in several frequency bands. This trend is evidenced by the use of SSPAs at Hughes in the UHF band (Leasat/Syncom IV), S band (GOES), C band (Telstar), and SHF band. The junction temperature of the transistor is the driving requirement which determines the lifetime of the transistor, SSPA, and the payload. This temperature is determined by the transistor characteristics, use of the device, and mounting temperature of the SSPA. The temperature of the spacecraft in the area of the SSPA can be controlled by active or passive means. The various factors and interrelationships used to calculate and control the temperatures of SSPAs are described. The thermal design and calculation of junction temperatures are exemplified with the Geostationary Operational Environmental Satellite spacecraft.
To provide theoretical support for investigating different ways to obtain high open-circuit voltages in p-n junction silicon solar cells, an analytical treatment of heavily doped transparent-emitter devices is presented that includes the effects of bandgap narrowing, Fermi-Dirac statistics, a doping concentration gradient, and a finite surface recombination velocity at the emitter surface. Topics covered include: (1) experimental determination of bandgap narrowing in the emitter of silicon p-n junction devices; (2) heavily doped transparent regions in junction solar cells, diodes, and transistors; (3) high-low-emitter solar cell; (4) determination of lifetimes and recombination currents in p-n junction solar cells; (5) MOS and oxide-charged-induced BSF solar cells; and (6) design of high efficiency solar cells for space and terrestrial applications.
Flowgraph models describing relationships between thermal and electrical parameters of devices and associated circuits
Noise specifications for junction field-effect transistors are presented in different ways depending on the particular semiconductor manufacturer. Arithmetic involved in converting these specifications to equivalent RMS noise in microvolts developed at the preamplifier input terminal is reviewed. These methods were useful for estimating the noise performance of high input impedance preamplifiers used with E-field antennas operating in the range of 1 KHz to 10 MHz. Both the JFET MPF-102 transistor and the COS/MOS CA3600 transistor array provided amplification for VLF receivers where the internally generated noise was well below the atmospheric noise level. The CA3600 transistor array provided better performance because of the more symmetrical complementary MOS transistor transfer characteristics than a single N-type biased JFET transistor. The CMOS amplifier resulted in self-compensating gain characteristics over a very wide temperature range from -55 to +125 C.
A method for analyzing the radiation sensitivity of an integrated circuit is described to determine the components. The application of a narrow radiation beam to portions of the circuit is considered. The circuit is operated under normal bias conditions during the application of radiation in a dosage that is likely to cause malfunction of at least some transistors, while the circuit is monitored for failure of the irradiated transistor. When a radiation sensitive transistor is found, then the radiation beam is further narrowed and, using a fresh integrated circuit, a very narrow beam is applied to different parts of the transistor, such as its junctions, to locate the points of greatest sensitivity.
A high frequency oscillator circuit is provided using a low cost junction type field effect transistor (T sub 1) with a tuned circuit connected to its gate. The frequency of operation is determined by the tuned circuit and the capacitance reflected from the source to the gate. The transistor is matched to the frequency of operation so that this frequency falls within the roll-off portion of the transistor's transconductance verses frequency curve, preferably somewhat above the 3 db point in frequency. Phase shift necessary to sustain oscillation occurs due to the operation of the transistor in the roll-off portion of the curve and the addition of a phase shifting network (R sub 1, C sub 1) at the source.
Report discusses aspects of effort to develop low-noise, low-gate-leakage gallium arsenide-based junction field-effect transistors (JFETs) for operation at temperature of about 4 K as readout amplifiers and multiplexing devices for infrared-imaging devices. Transistors needed to replace silicon transistors, relatively noisy at 4 K. Report briefly discusses basic physical principles of JFETs and describes continuing process of optimization of designs of GaAs JFETs for cryogenic operation.
Gallium arsenide junction field-effect transistors (GaAs JFETs) can be made immune to carrier freeze-out, making such transistors useful for the readout of detectorr arrays that operate at 4K.
Prototype logic gates made of n-channel junction field-effect transistors (JFETs) and epitaxial resistors have been demonstrated, with a view toward eventual implementation of digital logic devices and systems in silicon carbide (SiC) integrated circuits (ICs). This development is intended to exploit the inherent ability of SiC electronic devices to function at temperatures from 300 to somewhat above 500 C and withstand large doses of ionizing radiation. SiC-based digital logic devices and systems could enable operation of sensors and robots in nuclear reactors, in jet engines, near hydrothermal vents, and in other environments that are so hot or radioactive as to cause conventional silicon electronic devices to fail. At present, current needs for digital processing at high temperatures exceed SiC integrated circuit production capabilities, which do not allow for highly integrated circuits. Only single to small number component production of depletion mode n-channel JFETs and epitaxial resistors on a single substrate is possible. As a consequence, the fine matching of components is impossible, resulting in rather large direct-current parameter distributions within a group of transistors typically spanning multiples of 5 to 10. Add to this the lack of p-channel devices to complement the n-channel FETs, the lack of precise dropping diodes, and the lack of enhancement mode devices at these elevated temperatures and the use of conventional direct coupled and buffered direct coupled logic gate design techniques is impossible. The presented logic gate design is tolerant of device parameter distributions and is not hampered by the lack of complementary devices or dropping diodes. In addition to n-channel JFETs, these gates include level-shifting and load resistors (see figure). Instead of relying on precise matching of parameters among individual JFETS, these designs rely on choosing the values of these resistors and of supply potentials so as to make the circuits perform the desired functions throughout the ranges over which the parameters of the JFETs are distributed. The supply rails V(sub dd) and V(sub ss) and the resistors R are chosen as functions of the distribution of direct-current operating parameters of the group of transistors used.
The figure schematically depicts an oscillator circuit for driving a piezoelectric transducer to excite vibrations in a mechanical structure. The circuit was designed and built to satisfy application-specific requirements to drive a selected one of 16 such transducers at a regulated amplitude and frequency chosen to optimize the amount of work performed by the transducer and to compensate for both (1) temporal variations of the resonance frequency and damping time of each transducer and (2) initially unknown differences among the resonance frequencies and damping times of different transducers. In other words, the circuit is designed to adjust itself to optimize the performance of whichever transducer is selected at any given time. The basic design concept may be adaptable to other applications that involve the use of piezoelectric transducers in ultrasonic cleaners and other apparatuses in which high-frequency mechanical drives are utilized. This circuit includes three resistor-capacitor networks that, together with the selected piezoelectric transducer, constitute a band-pass filter having a peak response at a frequency of about 2 kHz, which is approximately the resonance frequency of the piezoelectric transducers. Gain for generating oscillations is provided by a power hybrid operational amplifier (U1). A junction field-effect transistor (Q1) in combination with a resistor (R4) is used as a voltage-variable resistor to control the magnitude of the oscillation. The voltage-variable resistor is part of a feedback control loop: Part of the output of the oscillator is rectified and filtered for use as a slow negative feedback to the gate of Q1 to keep the output amplitude constant. The response of this control loop is much slower than 2 kHz and, therefore, does not introduce significant distortion of the oscillator output, which is a fairly clean sine wave. The positive AC feedback needed to sustain oscillations is derived from sampling the current through the piezoelectric transducer. This positive AC feedback, in combination with the slow feedback to the voltage-variable resistors, causes the overall loop gain to be just large enough to keep the oscillator running. The positive feedback loop includes two 16-channel multiplexers, which are not shown in the figure. One multiplexer is used to select the desired piezoelectric transducer. The other multiplexer, which is provided for use in the event that there are significant differences among the damping times of the 16 piezoelectric transducers, facilitates changing the value of one of the resistors in the positive-feedback loop to accommodate the damping time of the selected transducer.
Arkansas Power Electronics International (APEI), Inc., is developing a high-efficiency, radiation-hardened 3.8-kW SiC power supply for the PPU of Hall effect thrusters. This project specifically targets the design of a PPU for the high-voltage Hall accelerator (HiVHAC) thruster, with target specifications of 80- to 160-V input, 200- to 700-V/5A output, efficiency greater than 96 percent, and peak power density in excess of 2.5 kW/kg. The PPU under development uses SiC junction field-effect transistor power switches, components that APEI, Inc., has irradiated under total ionizing dose conditions to greater than 3 MRad with little to no change in device performance.
Gallium arsenide junction field-effect transistors (GaAS JFETS) are promising for deep cryogenic (<10k) readout electonics applications.