Common-base Current Gain Research Articles

Importance of electron scattering with coupled plasmon-optical phonon modes in GaAs planar-doped barrier transistors

The GaAs planar-doped barrier (PDB) transistor is an MBE-grown structure which employs two unipolar homo junction barriers. One barrier, the emitter, injects energetic electrons into a thin n-type base region where these electrons are intended to experience negligible energy relaxation and thereby surmount the second (collector) barrier. Maximum common-base current gain or α values of 0.75 have been obtained at 77 K in experimental devices with base widths of 870 Å. Microwave measurements from 2 to 18 GHz on these devices imply a unity common-emitter current gain frequency f <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</inf> of ∼ 40 GHz. The observed α values in other devices are unfortunately lower than those predicted by recent Monte Carlo simulations, and an unexpectedly strong dependence of α on ambient electron density in the base is noted. These are attributed to the previously overlooked electron scattering with coupled plasmon-optical phonon modes, and to electron-electron scattering. These scattering mechanisms must be included in any accurate model of energetic electron transport in GaAs in regions where the concentration of cooler ambient electrons is above 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">17</sup> cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> .

Radiation Damage to Integrated Injection Logic Cells

The effects of neutron and total dose gamma irradiations on the electrical characteristics of an integrated injection logic (I2L) cell and an I2L multiple inverter circuit were investigated. These units were designed and fabricated to obtain circuit development information and did not have radiation hardness as a goal. The following parameters of the test structures were measured as a function of total dose and neutron fluence: the dc common-base current gain of the lateral pnp transistor; the dc common-emitter current gain of the vertical npn transistor; the forward current-voltage characteristics of the injector-substrate junction, and the propagation delay versus power dissipation per gate for the multiple inverter circuit. The limitations of the present test structures in a radiation environment and possible hardening techniques are discussed.

Gamma radiation effects on integrated injection logic cells

Integrated injection-logic (I <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> L) cells were tested to determine their characteristics after exposure to a total dose gamma-radiation environment. These particular devices were not designed or fabricated with radiation hardness as a goal. The common-base current gain of the lateral p-n-p transistor, the common-emitter current gain of the vertical n-p-n transistor and the forward current-voltage characteristics of the injector-substrate junction were measured over the current range of 100 nA to 300 µA as a function of dose. In addition, the propagation delay time versus power dissipation per gate at various dose levels was determined from frequency of oscillation measurements of a multiple inverter circuit.

The Surface Oxide Transistor (SOT)

A new type of surface barrier transistor has been investigated in which two junctions are employed each consisting of a thin conducting oxide or insulator between a metal and a semiconductor. The transistor behaviour of a structure composed of two such MIS contacts and a third base contact has been studied as a function of oxide thickness and contact separation distance. Two distinct classes of transistor action are observed each depending on a different mechanism of current transfer through the insulator of the individual formed diodes. The distinguishing feature between these two classes was found to be common base current gain, h FB . Devices characterized by h FB > 1 were noted to involve current conduction through the oxide via tunneling processes while those with h FB < 1 generally involved conduction in Schottky barrier-like diodes.

Appraisal of semiconductor-metal-semiconductor transistor

This paper considers the basic limitations of a semiconductor-metal-semiconductor transistor (SMST) and evaluates its performance in practical microwave circuits. The low frequency common-base current gain α 0 for SiAuGe and SiAgGe transistors is about 0.3. Discussion is presented suggesting that it is unlikely that materials combinations exist capable of substantially higher α 0. Thus the transistor shows current loss and can attain power gain only by means of impedance gain. In coaxial-line and waveguide structures both high and low impedance levels require very small mechanical dimensions and tolerances. With present technology, attainable impedance ratios appear to be limited to about 100. The technological problems associated with the fabrication of the SMST are discussed. The greatest obstacle is the deposition of device-quality semiconductor material on thin metal films. For a SiAuGe structure with a collector epitaxial layer 1 μ thick, base-metal thickness of 100 Å, emitter stripe width of 1 μ, and at a current density of 3000 A/cm 2, the calculated values at 10 GHz are: 26 db for the unilateral gain, 29 db for the maximum stable gain with a resistance ratio of 10 4 and about 10 db for the transducer gain with a stripe length of 1000 μ and a resistance ratio of 100, where no parasitic losses were included in the above figures and optimum adjustments were assumed for reactive components of generator and load immitances. For an α 0 of 0.3 the minimum noise figure in an SMST is 7.9 db.

Stable and unstable operation of complementary composite transistors

It has been found that the common-base current gain of a complementary composite transistor may have values far exceeding unity. An indication is given of the conditions necessary to achieve these high values. It has also been found that the input impedance of such composites may have a negative-resistance part. An oscillator is described which uses this effect.

On the Neutron Bombardment Reduction of Transistor Current Gain

Detailed measurements of the fast-neutron and gamma-ray bombardment behavior of germanium-alloy-transistor current-gain have been obtained concurrent with exposure. These data indicate that previously reported analyses, which lead to a linear dependence of common-base current-gain on fast neutron exposure, yield a good approximation for the npn device, but are not of general validity for the pnp germanium transistor. The extent of departure from the linear approximation depends on the width and conductivity of the base-region and can be appreciable in many cases of practical interest. For the pnp germanium transistor it is necessary to take additional account of both changes during bombardment of the minority-carrier recombination rate at bombardment introduced and initially present recombination centers and changes in the width of the collector junction depletion layer. Observed bombardment curves are in good agreement with an analysis which includes these effects. From initial slopes of the current-gain bombardment curves, values of the product of fast-neutron-exposure times minority-carrier-lifetime at bombardment introduced recombination centers are 9.7×107 for 2.7 ohm-cm p-type and 14.2×107, 6.0×107, and 1.3×107, for 3.6 ohm-cm, 1.2 ohm-cm, and 0.2 ohm-cm n-type germanium, respectively, in units of sec-neutrons/cm2.

Measurement of equivalent circuit parameters of transistors at v.h.f.

A technique is presented for measuring the current gain of transistors in magnitude and phase under near-short-circuit conditions at frequencies up to some 180Mc/s. The basic method of measurement involves the determination of three relative voltage magnitudes, from which the magnitude and phase of current gain can be deduced. An important feature of the technique is its simplicity: no specialized equipment is required, signal amplification and voltage magnitude assessment being carried out by means of a standard communication receiver together with a suitable calibrated attenuator and/or indicator.Using the technique of current-gain measurement in conjunction with other measurements, the parameters of the complete high-frequency equivalent circuit of the transistor are determined for diffusion and drift transistors of essentially one-dimensional geometry. In the case of the drift transistor, an important step in this process of measurement is the determination of the base field parameter, m, defined as the ratio of the ‘built-in’ base-region p.d. to the thermal p.d., kT/e. It is shown that m may be deduced from the phase shift of the terminal common-base short-circuit current gain at its 3 dB cut-off frequency, the emitter depletion-layer capacitance having also been measured. Results of equivalent circuit determination are presented for a diffusion transistor and several types of drift transistor, the dependence of the terminal and intrinsic current gains on frequency being shown as polar plots. The properties of the intrinsic current gain of base transport, as deduced from measurements, are compared with the theoretical behaviour assuming for the latter a constant electric field strength throughout the base region for the drift transistors. It is concluded that the commonly used constant-field theory is adequate for analysis purposes in most practical applications of the drift transistors studied, although certain discrepancies are observed between theory and measurement.

Approximations to α for diffusion transistors

The paper reviews, from the viewpoint of design engineers, various approximations to the common-base intrinsic current gain of junction transistors. The basis of this comparison is the familiar expression for α derived directly from the semiconductor diffusion and continuity equations.A 2-pole approximation is considered in detail. It is shown that both the sinusoidal and transient response predicted using this expression is in good agreement with that predicted using the exact theory. For most practical transistors the parameters of this approximation remain reasonably constant.In the response to a step function of emitter current, the 2-pole equation closely describes the initial time delay arising from the transit time of minority carriers across the base region. The collector response to unit charge inserted into the base in the form of σ function also gives good agreement with the results obtained from the exact theory.The various approximations are compared by graphical presentation of the results.

The current gains of diffusion and drift types of junction transistors

Theoretical aspects of the frequency dependence of current gain are reviewed. A semi-empirical expression for the common-base internal current gain αd is given, which closely approximates to the theoretical expression and is useful for analytical work.Measurements of the complex common-base external current gain α, from which αd is derived, are presented for frequencies up to 210 Mc/s. A comparison of the complex loci of αd with theoretical loci shows good agreement in most cases. A simple method of estimating the relative drift potential across the base of drift transistors is discussed, and two methods of determining the emitter depletion-layer capacitance are described.

Junction Transistor Short-Circuit Current Gain and Phase Determination

This paper presents a physical analysis which establishes that the phase shift associated with the common-base current gain of a junction transistor is of the type defined by Bode as "minimum." The phase of the common-base current gain is therefore uniquely determined by its magnitude characteristic. Using the minimum phase properties of networks and the empirically observed behavior of transistor current gain, it is shown that the complete common-base and common-emitter current-gain magnitude and phase frequency characteristics can be determined from three amplitude measurements, namely the low frequency α, α <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> , the magnitude of the common-emitter current gain at a single frequency in the common-emitter cutoff region, and the common-base cutoff frequency ƒα. The paper develops the equations for determining the common-emitter and common-base short-circuit current gains from these three magnitude measurements.