Showing papers in "IEEE Transactions on Electron Devices in 1964"

A self-consistent iterative scheme for one-dimensional steady state transistor calculations

Abstract: A self-consistent iterative scheme for the numerical calculation of dc potentials and currents in a one-dimensional transistor model is presented. Boundary conditions are applied only at points representing contacts. Input data are: doping profile, parameters governing excess carrier recombination, parameters describing the dependence of mobility on doping and on electric field, applied emitter and collector voltages, and a trial solution for the electrostatic potential. The major limitation of the present approach results from use of Boltzmann rather than Fermi statistics. Convergence of the iteration scheme is good for low and moderate injection levels.

Characteristics of the metal-Oxide-semiconductor transistors

Abstract: The theory of the MOS transistor in the gradual channel approximation is presented with the assumption of constant surface and bulk charge, and constant surface mobility. From the simple theory, the complete design equations are derived and design curves are calculated. From the analysis, the equivalent circuit parameters of the device are related to the basic properties of the material and geometry of the device. The simple theory is then critically compared with experimental measurements of MOS transistors with circular geometry. The comparison shows good general agreement with the theory of the dc characteristics but discrepancies are found for the differential characteristics such as the transconductance and the gate capacitance. The possible sources of the discrepancies are discussed.

The effects of distributed base potential on emitter-current injection density and effective base resistance for stripe transistor geometries

Abstract: Simple analytic expressions are derived for the variation of base current, emitter current and emitter-base voltage along the emitter junction of a transistor structure due to the flow of dc base current parallel to the emitter junction. The effective width of the emitter region is discussed, and it is shown that as the emitter width is increased, the effective emitter width approaches a constant value. The change in base resistance with emitter current is also calculated. The use of base contacts on each side of an emitter region is discussed and the analysis is shown to be applicable to this problem for equal or unequal division of base current between the two contacts. The results are applied to a planar transistor structure.

Frequency dependence of the reverse-biased capacitance of gold-doped silicon P + N step junctions

Abstract: The experimentally observed frequency dependences of the reverse-biased capacitance of gold-doped silicon step junctions over the frequency range from 10 cps to 30 Mc are found to be in agreement with a simple physical model which takes into account the charge condition and the charging and discharging time constant of the deep-gold acceptor level in the transition region of the junction. Analysis based on the simple physical model provides explicit theoretical formulas for the junction capacitance at low- and high-frequency limits which show that the high-frequency capacitance under reverse bias is approximately proportional to \sqrt{N_{D} - N_{Au}} and is considerably reduced below the low frequency or dc capacitance if the donors are nearly compensated by the gold. The frequency effect is important for deep energy level impurities and becomes negligible if the impurity level is at or near the band edges. The presence of gold, however, has negligible effect on the avalanche breakdown voltage if N_{Au} .

Small-signal, high-frequency theory of field-effect transistors

Abstract: In this theory a field-effect transistor of planar geometry is considered as an active, distributed, nonuniform transmission line. The wave equation of this line is determined and solved by an approximation method. From this solution the y parameters are determined. By comparing the results for y 11 with van der Ziel's expression for the high-frequency gate noise of field-effect transistors, it is shown that the noise temperature of y 11 is of the order of the device temperature. The conductance part g 11 of y 11 varies as ω2over a wide frequency range. The high-frequency cutoff of the field-effect transistor is determined.

Analysis and characterization of P-N junction diode switching

Abstract: A new charge control model of a p-n junction diode is introduced in which the reverse current i R as well as the forward current I F are related to the charge Q stored in the base region by time constants τ R and τ F , respectively. The reverse switching transient is analyzed for normal switching operation where a constant current phase (storage phase) and a decaying current phase exist, and for overdriven switching operation where no constant current phase exists. New switching time equations are derived. The equations are expressed in terms of measurable device parameters τ F , τ R , and C j external circuit variables I F and I R and an external circuit parameter R. The proposed model is applicable to p-n junction diodes of any type. Experimental results using various types of diodes are also reported. It is shown that the experimental results are in very good agreement with the theory.

Construction of a promethium-147 atomic battery

Abstract: The feasibility of constructing an "atomic battery" by combining the β-emitting radioisotope Pm147and a semiconductor containing a p-n junction has been studied both theoretically and experimentally. The properties of cells made by diffusing a p-type impurity into n-type GaAs, p-impurity into n-type silicon (p/n) and n-impurity into p-type silicon (n/p) are compared. The problems associated with the presence of an alpha emitting impurity (Am147) in the Pm147and with the fabrication of a suitable Pm147source are considered. Experiments with prototype batteries composed of Pm147and n/p silicon cells have led to the conclusion that such batteries would have a half life (1.6 years) somewhat shorter than that of Pm147(2.6 years). Such a battery could generate a few milliwatts of power from a volume of about 2 cubic inches including shielding.

The method of estimating delay in switching circuits and the figure of merit of a switching transistor

Abstract: A simple method of estimating delay in a switching network is outlined. The simplicity of formulas thus obtained makes them readily applicable for circuit comparisons or device optimization purposes. It is shown that a term involving the product of base resistance and diffusion capacitance forms a major limitation on high-speed, voltage-driven circuits. This method is applicable to a general class of switching problems.

An investigation into the use of iteration methods for the analysis of axially symmetric and sheet beam electrode shapes with an emitting surface

Abstract: The techniques for analyzing an axially symmetric or two-dimensional electrode system with an emitting surface have been studied with a computer program. The program is very easy to use and performs in one pass the steps which have usually previously been followed on analog and digital devices separately. The program alternately solves Poisson's equation by a relaxation technique, traces trajectories, recomputes the space-charge distribution and again performs a relaxation process. This procedure continues until satisfactory convergence has been achieved. The ease of use of the computer program has allowed the investigation of different methods of speeding up the convergence. The results are illustrated with two specific examples.

Charge-storage effects in silicon dioxide films

Abstract: The dc properties of the system Si:SiO 2 (6000 A), metal have been investigated at 200-400°C. The system shows unidirectional conduction at 400°C. Little or no current is observed when the silicon electrode is positive. There is a region of reversible charge storage and discharge at negative silicon voltages; the center of this region is around -2 volts when the contact metal is gold, but more positive when it is aluminum. The quantities of charge stored range from 4 × 1012to 2 × 1014charges/cm2. At still more negative silicon voltages there is a region of ohmic conduction, the resistivity at 400°C ranging from 1012to 1014Ω-cm. The charge storage, which is far greater than that anticipated from the oxide capacitance, can be shown to occur principally in the oxide under the metal contact, and not over a broad adjacent area. Storage of charge in the oxide causes an n -type shift of the underlying silicon surface. Charges can develop in the oxide as a result of nonelectrical processes, such as high-temperature gas baking treatments. Conversion of the upper 600 A of the oxide to a phosphate glass eliminates both charge storage and ohmic conductivity in the oxide. The results described are interpreted in terms of an electrochemical potential existing across the oxide, and the presence of mobile charged species within the oxide. Charge storage can occur as the result of double-layer formation by the mobile charged species, or of its electrochemical discharge at the electrodes. The unidirectionality of the ohmic conduction is considered to result from electrolytic rectification at the silicon-oxide boundary.

A critique of the theory of p-n-p-n devices

Abstract: A review of various papers dealing with p-n-p-n device characteristics shows essentially different interpretations of 1) the condition for switching and 2) the operating state defined by \alpha_{p}+ \alpha_{N} = 1 . The present paper attempts to show the relationship between these various papers and to clarify the conditions which exist at the switching point and holding point.

Some effects of localized stress on silicon planar transistors

Abstract: When local compressive stress is applied to the emitter surface of a silicon planar transistor, increases in collector and base currents are observed. From the relationship between these currents and the base-to-emitter junction potential, it is inferred that changes in minority-carrier current occur in the stressed material. It is well known that the energy gap of a semiconductor is affected by mechanical stress, and the effects described here are consistent with a decrease in energy gap of a few tenths of an electron volt at compressive stresses of the order of 1011dynes cm-2. It is shown that the influence of stress on the current gain of a transistor depends on the strain distribution in the material. When the stress is applied over a relatively large area using a blunt stressing element, the increased minority-carrier current causes a rise in common-emitter current gain at low currents. In the case of the stress pattern produced by a sharper stressing element, a decrease in current gain is observed over a wide current range. It is proposed that this decrease results from a degradation in emitter efficiency caused by the stress concentration in the emitter region.

Space charge limited and emitter current limited injections in space charge region of semiconductors

Abstract: If, in a space charge region of a semiconductor device, a unipolar drift current is a main component of the total current, an analogy with motion of electrons in vacuum exists. In the space charge limited emission, injection of minority carriers is limited by space charges, as in a punched-through p-n-p diode. On the other hand, in a transistor, minority carriers are injected into the depletion layer between the base and the collector, after diffusion in the neutral base region. This injection is limited by emitter current, and this mode of injection corresponds to a temperature limited emission in a vacuum tube. In this paper, equations of minority carrier impedances will be calculated in both these cases, and negative resistances can be expected at some transit angles. Then a generalized equation, which includes these two modes of injections as special cases, will be obtained. This generalized equation corresponds to the Llewellyn-Peterson equation of vacuum tubes, which include the space charge limited emission and the temperature limited emission as special cases.

Noise transport in the crossed-field diode

Abstract: The present work treats the transport of noise, originating at a thermionic cathode, across a planar crossed-field diode. A multivelocity Monte Carlo analysis is used to simulate thermionic emission by using random numbers to generate electron emission times and velocities. Trajectories for many charges are followed through the diode space by means of a high-speed digital computer. Measurements performed on a planar triode model for the crossed-field diode show that the anode noise current increases rapidly near the critical magnetic field, at which the stream just grazes the anode. The theory agrees well with these experimental findings. The Monte Carlo study predicts the increase of both the normal anode velocity fluctuations and the spread of the velocity distribution with magnetic field. This is caused mainly by the "geometrical" spreading of the initial velocity distribution. The crossed-field potential minimum exhibits no instabilities in the planar model. The noise current smoothing at low frequencies in ordinary streams is predicted for moderate fields but disappears near the critical field.

Properties of ring-plane slow-wave circuits

Abstract: Ring-plane circuits consist of circular rings or slotted pipes supported by one or more radial planes. These circuits propagate slow waves (v_{p} in one or more modes, with velocity and impedance governed primarily by the ring or slot resonances much as on planar ladders. The multimode behavior depends very much on the mutual coupling near the radial planes. Tests of many variations about a prototype configuration are presented, including methods for suppression of unwanted modes. The results are given in terms of ω -- β diagrams, field shapes and impedances. The study presents essentially all parameters necessary for design of circuits for traveling-wave amplifiers and backward-wave oscillators, as well as information useful in other applications, such as linear accelerators. Ring-plane circuits are of large diameter (ka \ge 1) , are quite fast (v_{p}/c large) and have relatively large impedances. The radial support planes provide good heat conduction paths, allowing for high average power capability. These properties make these circuits well suited for use in high-power, high-frequency traveling-wave tubes.

Mixing and detection of coherent light in a bulk photoconductor

Abstract: The use of a bulk photoconductor as a detector for microwave-modulated light or as a mixer to detect the beat frequency between coherent optical sources is studied. A model for the photoconductor as a parametric conductance is presented and used in the analysis of a dc-biased detector and an ac-biased mixer-detector arrangement. The results of experimental measurements on both dc- and ac-biased detectors are presented and found to be in good agreement with the theoretically predicted performance. The bulk photoconductor is compared with the traveling-wave phototube and the PIN photodiode in both signal power and SNR and is found to compare favorably with these devices if the semiconductor parameters can be reasonably optimized and if it is biased with fairly high fields. The desired semiconductor properties are that the lifetime should be on the order of the reciprocal of the angular frequency of modulation, the mobility should be high, and the dark conductance of the sample should be very small.

Low-frequency operation of four-terminal field-effect transistors

Abstract: Interesting and useful behavior is obtained by operating a two-gate field-effect transistor as a four-terminal device. This paper considers important aspects of the low-frequency four-terminal behavior of a transistor with symmetrical geometry. The functional dependence of the drain current upon gate bias voltages and material and structural constants is determined, and four general modes of operation of the device are defined. For each mode the dependence of relevant transconductances on gate biases is calculated. Special cases of interest are analyzed, including that in which the transconductance exhibits a linear variation with gate bias voltage. Experimental results are in good agreement with the theory.

Rise and fall time calculations of junction transistors

Abstract: Expressions for rise and fall time calculations of junction transistors are derived based on the charge controlled model. In solving the differential equation of the charge controlled model, a new approach is taken and exact solutions are obtained. Comparison of the experimental results of the rise and fall time measurements with the theoretical prediction, using the measured transistor parameters, shows very good agreement.

A two-cavity extended interaction klystron yielding 65 per cent efficiency

Abstract: Results of experiments with a two-cavity extended interaction klystron are given. The tube is nominally capable of 1 kw-CW at S band and consists of two five-gap short-slot-coupled cavities coupled to the input and output lines by means of TWT matching sections with adjustable mismatches. The unusual features of the tube are 1) the length of extension of the resonant cavities (five electron ballistic wavelengths) and 2) the incorporation of different and controllable beam-circuit synchronism conditions for the input and output cavities. The results of principal interest are the low velocity spread in the spent beam entering the collector providing high total conversion efficiency (65 per cent) with collector depression and stable gain in excess of 15 db. Some treatment of the relationship of efficiency and oscillation stability is included, as well as comments on the usefulness of such a design for high power millimeter wave amplification.

Correction factors for radial resistivity gradient evaluation of semiconductor slices

Abstract: Accurate determinations of radial resistivity gradients on semiconductor slices, using the four-point probe, can be made only if correction factors for the finite boundary conditions are applied to the measurements. The off-center correction is different from the center correction. Both are necessary for gradient evaluation. These corrections were derived assuming homogeneous material, but are, nevertheless, useful approximations for nonhomogeneous materials. The corrections are given as generalized formulas and are graphed for the 62 and 25-mil probes for the center and 1/8-inch from the edge positions.

The Laddertron—A new millimeter wave power oscillator

Abstract: The Laddertron is a multi-gap klystron using ladders. This paper describes the theory and experimental results of the tube. Two types of Laddertrons are distinguished, namely the O -mode Laddertron and the π-mode Laddertron. The operation of each is analyzed and the equivalent circuit representations determined theoretically, whereby one can compute the resonant frequencies within an error of several per cent. A demountable experimental version of the π-mode Laddertron gave an output power of 10 watts for the 50-Gc band, an electronic tuning range of 300 Mc, and a mechanical tuning range of 2.5 Gc.

Fine structure and electromagnetic radiation in second breakdown

Abstract: Thermal phenomena have been suggested as the initiating mechanisms for second breakdown in transistors. The experiments described below examine certain electrical and optical effects associated with second breakdown in order to determine if thermal effects can be held primarily responsible. In studying the open-base behavior of several types of devices, it was observed that second breakdown was followed by additional discontinuities at higher current levels, all similar to second breakdown in their electrical behavior. Simultaneously with the appearance of a voltage discontinuity, light was emitted by the device. The light output level generally increased monotonically with increasing collector current except when a voltage discontinuity occurred, at which time a discontinuous increase in the light output level also occurred. The spectral distribution of the emitted light was examined and found to be similar to the distribution from a single junction biased into avalanche. The results of the electrical measurements strongly suggest the existence of a number of interacting sites capable of sustaining second breakdown. The optical measurements indicate that the characteristics of a transistor in second breakdown are not purely thermal in nature, but that a plasma effect may be associated with a device exhibiting an anomalous breakdown.

Design calculations for MOS field effect transistors

Abstract: Design equations for the Metal-Oxide-Semiconductor Field Effect Transistor are developed. Approximate solutions for static characteristics, transconductance, and frequency cutoff are presented for the case of a very high resistivity substrate. Specific sets of static characteristics from computer calculations are presented graphically to illustrate the effects of oxide thickness and various substrate resistivities.

Verification and use of Herrmann's optical theory of thermal velocity effects in electron beams in the low perveance regime

Abstract: A comparison between Herrmann's optical theory and a pinhole probe study of a highly converged, perveance 0.015 × 10-6A/V3/2, electron beam in a confining magnetic field is presented. This comparison shows the theory to be obeyed with a good degree of accuracy. The application of the theory to the problem of optimizing the transmission of such thermal beams through long drift tubes is shown to lead to criteria which differ radically from those prescribed by the laminar theory.

The traveling-wave phototube part I: Theoretical analysis

Abstract: The traveling-wave phototube (TWP) is a broad-band photodetector which is useful for demodulation of light that has been amplitude-modulated at a microwave frequency, and for optical heterodyning of coherent light signals with microwave difference frequencies. In either application the light beam(s) results in the emission of a current-modulated electron beam from a photocathode; interaction between this beam and a slow-wave circuit such as a helix is employed for broad-band detection of the microwave modulation. Part I of this paper presents a detailed analysis of the TWP beam-circuit interaction process, including the effects of loss, space charge and asynchronism. Computer results are given for both current modulation and velocity modulation excitations, since the latter can be important under certain operating conditions. Conclusions regarding the effects of beam current, circuit length and circuit loss are drawn which are important in the design, understanding and use of TWP's. The results of this analysis are also generally applicable to other cases where a slow-wave circuit is excited by an electron beam. In Part II of this paper, a detailed experimental analysis of the TWP will show good agreement with the predictions of this theoretical treatment.

Nonlinear theory for laser diodes

Abstract: This paper outlines a nonlinear theory for the output power vs excitation current of an injection diode laser. The dependence upon electric field amplitude of effective negative conductivity in the region of inverted population derived and used to determine the level of steady oscillation in the quasilinear approximation (nonlinear conduction current small compared with displacement current). This level is simply related to the output power. A field dependent negative conduction current is derived from the negative conductivity which indicates that, in the quasilinear approximation, the presence of two modes with the same polarization will be an unstable situation; one mode will grow at the expense of the other.