Search

1. The 60 Hertz and impulse electrical strength of oil impregnated polyamide paper is as good as that of high quality electrical grade kraft paper and better than that of rope-kraft paper used as conductor insulated. 2. The 60 Hertz dissipation factor of polyamide paper in oil increases moderately with increasing temperature. 3. Based on burst strength and tensile strength retention, the thermal stability of the polyamide paper is substantially superior to that of cellulosic paper when aged in oil or askarel at 150°C and 175°C. 4. Polyamide paper showed little or no degradation after extended life tests in distribution transformers. 5. Polyamide paper with its superior mechanical properties can better withstand manufacturing operations in production than can cellulosic papers. 6. Based on life test results the polyamide paper will permit the design of transformers with higher operating temperatures and greater overload capacities than can be obtained using cellulosic paper.

Insulation paper is primarily used in electric motors as slot cells, coil separators and phase insulation in integral horsepower equipment. The slot cell being defined as the basic ground insulation that provides mechanical and dielectric barriers from the magnet wire to the stator core. This definition emphasizes that the insulation paper forming the slot cell is mainly ground insulation.

Improved functional properties and process simplification now can be attained through the use of solventless, resin-filled Mica Mat tapes used for high voltage ground insulation. Previously, all B-staged resin-filled tapes composed of either flake mica or mica paper have been manufactured with resin/solvent solutions because the solvents were necessary to reduce the viscosity of the resin for impregnation and to soften the tapes for application. However, when solvents are used, removal of them is difficult, costly, and complete removal is often virtually impossible.

This paper is about the applications of an epoxy impregnated paper insulation to ultra-high voltage systems. An epoxy impregnated paper insulation is a resin-rich, void-free structure, and therefore has extremely superior voltage withstand characteristics. It also has outstanding resistances to thermal cycle and mechanical vibration. Consequently, such insulation system can be applied easily to condenser bushings. From the standpoint of the production facilities, techniques, and economical considerations, in items with an overall length of more than three meters, “pre-fab” type condenser bushings, consisting of several condenser cores, are desirable. In cases when two unit condenser cores are assembled leaving a simple conical shaped gap between them, it is extremely effective to adopt a “potential expanding” type gap designed so that the equipotential surface in the gap will be in the vertical position to the cone surfaces as far as possible. The “pre-fab” type condenser bushing of the 500 kV class was trially manufactured using this arrangement; it displayed all the desired properties. This paper also describes the other examples of applications of epoxy impregnated paper insulation: CT bushings and pulse transformers in the magnetic firing system for high voltage D.C. converters using thyristors, and draw-out type condenser bushing for disconnecting devices used in the compact type substation.

This paper is concerned only with the effect of the dissipation factor of the liquid upon the overall dissipation factor of impregnated paper insulation.

The voltage endurance of liquid impregnated paper insulation is important in both test and service for high voltage equipment such as transformers, capacitors, and cable. It has been recognized for a long time that if electrical discharges occur, they may limit the life. Especially with cable, corona studies have been carried out for many years using both dielectric loss measurements (i. e., Schering Bridge) and several types of amplified band pass techniques for more direct measurement of corona pulses.

The use of capacitors in electric networks is becoming more and more common. The capacitor remains under the voltage of the network, and in addition it is subjected to various over-voltages appearing in the network. The latter will originate from lightning strokes and more frequently from switching operations. Due to the incompleteness of data on the capacitor dielectric strength under switching surges, the problem became of interest. The lightning and switching surges may cause a weakening of the dielectric strength of many types of equipment. Power capacitors, with their impregnated paper insulation, belong to this category of equipment.

The effect of switching surges on the dielectric strength of Askarel impregnated capacitor paper is less than for standard impulses and approximately the same for A.C. From the practical point of view the small variation in breakdown strength resulting from a change in the wave front duration from 35/110 µ sec to 600/2000 µ sec can be neglected in the design of capacitors.

This session is devoted to the insulation probieros in the manufacture and use of transformers. These probieros resuit from the progress in delermining the life of equipment from many viewpoints. New matcrials and ways of using them, processes, quality control, and corona effects, are illustrated as problems in the papers being presented.

Measurements of temperature histories of epoxy/glass prepreg layers during multilayer printed wiring board lamination have been recently obtained. In the present paper, it is shown that the temperature histories may be predicted by an analysis of the transient heat flow process within the MLB. The analysis requires a knowledge of the press platen temperature history during laminations and the thermal resistance of the padding material. Theoretically predicted results show close agreement with the experimental data for both 5 layer and 15 layer MLB's using the "hot press" as well as the "cold press" cycles. The present method may be used in the design of MLB lay-ups, in particular the determination of the necessary padding thickness to achieve heating rates which give adequate timing "windows" for cure pressure applications. The present theory has been applied to study the effect of kraft paper padding on the prepreg temperatures. Results are shown for both 15 and 5 layer laminations employing the hot press as well as cold press cycles. The number of kraft paper sheets used is shown to have a significant bearing on the temperatures.

The thought has been expressed that the manufacturer and user of insulating materials, who may use electrical resistance measured in accordance with ASTM D-257(1) as one means of evaluation and control, may not be familiar with the d-c tests that these materials are subjected to after assembly into insulation systems in apparatus. One of the purposes of this paper is to review d-c tests as they are used in the field to evaluate the condition of the insulation in power apparatus. It is largely in power equipment such as rotating machinery, switch-gear, cable and transformers that such field tests can be economically justified (2).

Mineral oil and askarel have been in general use in transformers, capacitors, cables, and switchgear for many decades. With the exception of polybutenes, no other dielectric liquids have gained acceptance for use in high voltage power apparatus, primarily because oil and askarel are so inexpensive and possess such excellent electrical properties in combination with kraft paper.

Machine taped coils insulated with a Resin Rich B Stage glass backed mica paper tape have now been in production for a number of years. This process allows all the main wall insulation to be applied and consolidated in one application to all types of coils without any vacuum treatment. The system can be used at Class F temperatures. No voltage restriction has been made and the production processes have been developed for use up to 30 kV.

Many applications in the electrical industry require the use of materials that must sustain a mechanical load as well as provide electrical insulation. Applications such as these are electromechanical in nature, and include the armature bar support system in the stator, and the slot liner, blocking and retaining ring insulation in the rotor. In order to fill both the electrical and mechanical requirements of these applications, compromises have traditionally been necessitated in one of the material properties. The development of new materials has reduced the need for these compromises, as it has provided materials with increased mechanical strength without sacrificing insulating properties.

Miniature electrical connectors manufactured in conformance to Specification MIL-C-26500 (USAF) are designed as two separate units which when mated provide for multiple electrical circuits having the ability to perform satisfactorily during extreme environmental condition. The outstanding features of this development are the physical size and performance achieved under extreme environmental conditions. The paramount development responsible for accomplishing these features is the composite dielectric material. The results obtained from this development contract plus supporting manufacture experience have shown that composite dielectric construction of resilient silicone and hard epoxy dielectric possess the necessary electrical and mechanical properties to meet the requirements of MIL-C-26500 (USAF). In addition, the resilient silicone developed by Amphenol incorporates sufficient mechanical strength to permit mold construction employing acute discontinuities in combination with .025 inch molded sections. Resilient silicone and hard epoxy are chemically bonded to form integral dielectric inserts which provide for environmental sealing at both the mating surfaces of the connector assembly plus the wire termination scaling areas. Correspondingly the contact retention device is housed internally in the integral and/or composite inserts. Bonding of the intersurfaces which form the interfaces between the three dielectric insert components eliminating air voids thereby protecting the connectors from degrading environmental conditions.

The stabilization of chlorinated capacitor impregnants by chemical additives has been practiced by a number of capacitor manufacturers for several years. Anthraquinone has been the most widely used stabilizer, and has been added to chlorinated diphenyl and chlorinated naphthalene to the extent of 0.5 to 1.0% by weight. The purpose of stabilization is to increase the d.c. life and to avoid undesirable changes in insulation resistance and power factor of capacitors in d.c. service. The benefit is illustrated by the 10- to 300- fold increase in life observed in accelerated life tests. The improvement noted in a particular case depends upon the nature and concentration of stabilizer and the conditions of test.

As indicated previously, use of the term "Class F" has become increasingly prevalent. Many users of electrical insulation are intensely interested and are inquiring about the availability of such materials. They want to design equipment in the 155°C range, but must have materials which can retain their properties sufficiently and adequately to insulate their motors, transformers, and other electrical equipment.

The subject of this session relates to Electrical Insulation Systems and it is to such systems or more particularly to one such system that I wish to direct attention.

Summary form only given, as follows. This paper is divided into two sections. Section I describes the results of approximately six million transistor hours of operation on various types of transistors from four different manufacturers. The data was taken on two different transistor systems containing digital type circuitry and on a set of transistor flip-flop circuits operated under semi-controlled conditions. The types of transistor failures that have occurred will be listed and the general aging trend will be discussed. This section will be accompanied with slides showing some of the circuitry. Based on the above life data along with the experience gained in operating transistor systems, Section II will cite certain conclusions leading to the design of reliable switching-type circuits. In particular a design procedure is demonstrated in the form of a clamped-level multivibrator. This section will show how initial discrepancies in transistor parameters are taken into account along with long term drifts in 'B' and 'Ico.' Aside from the fundamental design practices involved, the multivibrator circuit is chosen because it is relatively difficult to obtain the functional equivalent of a vacuum tube multivibrator due to transistor idiosyncrasies. These idiosyncrasies include low input impedance and hence slow circuit response because of the large coupling capacitors necessary for even small delays. The paper will be concluded with comments regarding clamped vs. unclamped logic and large system operation.

Summary form only given, as follows. The complementary pair family of circuits can yield great economy in digital computer design, as has been shown in the preceding paper. One member of this family is attractive, because of its simplicity, in those situations where the ultimate in speed and load handling capacity is not required. If it were possible to allow saturation in the transistors, a very simple bistable element would result. In general, ·this is not permissible, for the degradation of performance is too severe. By application of a new principle, which has been called by the author that of "Controlled Saturation," one of the transistors may be allowed to saturate without undue performance deterioration. Then the other transistor may be prevented from saturating at a cost of only a single resistor. The paper first describes the process of Controlled Saturation, and then shows the synthesis of a practical trigger circuit employing it.

This paper will discuss a monolithic operational amplifier which utilizes an improved feedforward technique to compress the feedforward pole-zero pair and yield an experimentally-observed total unity gain acquisition time (slewing plus settling) of 600 ns to 0.01% with a 10-V input step, without compromise of the dc performance parameters or the use of complex compatible processing.

Studies of the processes involved in conduction and breakdown in vacuum again proved to be an active field in 1974 with 140 papers published. Approximately one-third of these papers were presented at the VI International Symposium on Discharges and Electrical Insulation in Vacuum (1). Prebreakdown and breakdown phenomena received much attention at this symposium, and there were several significant reviews (2).

The continuing evolution of new components is resulting in corresponding advances in radar systems. The paper commences with a brief review of some of the basic radar parameters which have been influenced in the recent past by improvements in component technology and an indication is given of the ranges of these parameters which are now available to the designer. The principal part of the paper is a survey of the recent advances made in radar sub-systems and systems and in the understanding of the radar characteristics of targets and clutter. The review of sub-systems covers electronic beam steering, pulse compression, signal and data processors, constant false alarm rate circuits and displays. System principles discussed include within-pulse scanning, low-angle tracking techniques and synthetic aperture radar.

The paper will discuss a low-power, high-performance static RAM using a compact inverted transistor flip-flop memory cell. Small chip size and conventional processing have been found to yield an economical device suited for mainframe storage applications.

The use of gold in integrated circuit packages has become prohibitive because of cost. Of the current alternatives available, silver plating offers a favorable economic advantage to the MOS-LSI manufacturer. This paper discusses the economics of various gold substitute systems. In addition, the paper presents data which demonstrates the reliability of molded packages using silver plated lead frames. Reliability data on Life, Temperature Cycling, 85°C/85% R.H. and Solderability Testing is presented. Silver plating is an economical and reliable system for use in molded packages for MOS-LSI devices.

In previous work1,2 we showed that paper impregnated with liquids of high permittivity, such as nitrobenzene, exhibited strong non-linear effects. When, however, the paper was replaced by polypropylene foil, the nonlinear effects were much smaller. We have done numerous experiments with different kinds of insulating solids and liquids, using either industrial capacitor foils, or cells of small capacitance with well-defined geometry (liquid layer between two parallel solid sheets). Depending on the nature of the materials, the insulator/liquid association exhibits, to a variable extent, both a “barrier” — or GARTON — effect, and an “absorption” effect (ions are trapped into the solid where they can move very slowly only).

The use of N-channel MOS technology has made it possible to develop chips with speed advantages relative to PMOS because of the higher mobility of electrons. This paper will describe a chip employing 5000 components, discuss a 90% reduction in execution time compared to PMOS designs, and also compare circuit features with existing P-channel approaches.

A highly accelerated ac life test for high voltage dc capacitors has been developed which is useful for solid dielectric capacitors such as mica paper and high dielectric constant ceramics with low ac losses. The test utilizes the fact that corona intensity on ac is several orders of magnitude higher than on dc. It is shown that the mechanism of failure involves electrochemical erosion for both the mica paper and the ceramic dielectric material. The life times follow a power law with an exponent of 26 for mnica paper and 31 to 100 for the ceramic.

Voice communication receivers frequently encounter interference as a result of radar systems operating in the same general vicinity. Although the resulting interference may not completely prevent communications, it does have several undesirable effects. The most notable effects are operator fatigue and loss in communication range. To insure that these effects are given proper consideration during system planning and design, it is necessary to have a capability for analyzing pulse EMI effects in voice systems. This paper presents mathematical expressions that may be used for this type of analysis. These mathematical expressions define the relationship between desired signal and pulse interference conditions at the receiver input and audio performance at the receiver output. The expressions relate audio performance to significant input signal parameters such as desired signal level and frequency, interfering signal level and frequency, pulse width, pulse repetition rate, pulse shape, and receiver bandwidth. Comparisons are made between calculated and measured performance over a wide range of values for the input signal parameters including both on-tune and offtune interference conditions. The results presented in this paper provide an insight into the mechanism of pulse-to-voice degradation and a technique for calculating voice system performance in the presence of pulse interference. The relationships are simple, fundamental in principle, and require a.minimum amount of readily available information for their implementation.

In 1894 the first oil-impregnated paper-insulated lead-sheathed cable was used for high-voltage power transmission in the United States. For approximately fifty years no serious challenger appeared on the scene, and even at present, more than seventy years later, paper-insulated cable is still very much in use for all voltages from 15 kv up, and in one form or another is the only insulation used commercially in the 230–345 kv range. However, in the range from 5 kv (or lower) up to 138 kv, a 30-year old material, polyethylene, has seen increasing service. This paper will outline the requirements for a high-voltage cable insulation, and will investigate the extent to which polyethylene meets these requirements.

In developing a mature LSI and MSI product capability, the quality of the primary artwork or masks has emerged as one of the chief determinants of yield and resulting cost. This paper will review the state-of-the-art of masking technology, with discussions of various methods. The limitations and advantages of each will be described, with an indication of the level of excellence in practice in the industry.

This paper will describe a semiconductor display with inherent memory whose central element is a monolithic integrated circuit that performs the logic function in additon to the optical ouput on the display surface. Matrix address is permitted.

In September of 1970 the 4th International Symposium on Discharge and Electrical Insulation in Vacuum was held at Waterloo University in Ontario, Canada. While the meeting was somewhat less well attended than previous ones in the series this was not reflected in the many interesting papers which were presented. The listing and discussion of these make up a large part of the material presented in this chapter. Other conferences of importance occurred in London, Munich, and Chicago. The review papers presented at Waterloo involved a wide range of topics (1–6), as can be seen from the titles. Among the important points raised by reviewers was the fact that it has now become very clear that the interpretation of prebreakdown currents using the Fowler-Nordheim (F.N.) equation can be very misleading. This is because of the likely multiplicity of emission sites coupled with the variety of adsorbate states possible on each emitter. The general effect is to lead to a lack of information concerning the emitting area. In order to explain the later stages of the field emission induced breakdown due to the explosion of a cathode emitter it has now been found necessary to consider the emission of electrons from the cathode due to the field generated by the presence of a plasma cloud (explosion created) above the surface; this enhanced emission causes intense anode heating and vapor production. The latter theory is one of many different voltage collapse processes now under review, and is as yet one of the few to have been subjected to as careful an experimental test as has the hypothesis of Davies and Biondi (3, 28, 29). The process proposed by these authors involves the detachment of a high temperature anode region and its subsequent heating in the electron beam as it crosses the gap.

This paper is concerned with a class of oscillators having otuy two intrinsic windings on the core. lt is capable of true square-wave output at 700 kc.; the upper frequency limit being determined by the present state of the art of transistor development. On the basic two-winding oscillator described, the collector winding of the OFF transistor itself supplies the necessary regenerative voltage to the base of the ON transistor. The parallel RC's in the base circuits speed up the turn off action by application of inverse current to the ON base. Direct electrical crossconnection from the ON base to the OFF collector means that at switchover the in ver se transient voltage, due to flyback, will momentarily place the OFF transistor into inverse saturated conduction, thereby presupplying it with holes so that it can take on its load in the forward direction extremely rapidly when the ON transistor shuts off. With this circuit, selfstarting is inherent due to the direct connections from base to negative supply. The transistors used are SB-lOO's (low saturation resistance), and the switching flux of the core is 0.04 microwebers.

This paper will describe a 1-dimensional transistor analysis program featuring Fermi statistics, saturated velocity and transient effects that predicts switching delays (loaded, packaged CSEF circuits) Of 50 ps(Ge) and 150 ps(SI) for near-ultimate impurity profiles, and at current densities of 105A/cm2.

In Part I of this paper, the polarization characteristics of antennas using "geographical" analogies are reviewed. From the longitude polarization of a dipole and the latitude polarization of a loop, a vector description is given to the radiation from these basic elements. Combinations of dipoles, loops, or dipoles and loops can then be readily analyzed from simple complex vector formulas describing the amplitude, phase, and polarization of the far fields. In particular, turnstile-type combinations providing nominal circular polarization are discussed. The vector technique is also applied to aperture antennas. It is shown that the obliquity factor of aperture antennas ( 1 + \cos \theta ) is a scalar constituent of an obliquity vector characterizing the radiated field, whether associated with linear or circular polarization. The vector antenna pattern for a linearly polarized horn antenna is derived and applied to the problem of the square horn fed by orthogonal TE modes in quadrature. Strong left circularly polarized lobes are radiated by a nominally right circular horn. The reduction of cross-polarization by providing a square horn with polarized teeth is described. The paper contains "geographical" drawings and sketches of 3 D contour plots of right and left circular component antenna patterns. In Part II of this paper, "Punch-Through" problems associated with sidelobe suppression systems are discussed. The effectiveness of side-lobe suppression systems at any given angle of incidence can be determined by the measurement of punch-through probability for all possible polarizations of the input signal. For systems where only random linear polarizations need be considered, a rotating linearly polarized test signal is used, and the punch-through probability can be measured directly as a percentage of a complete rotation that produces an acceptable response. For side-lobe suppression systems that must consider response to completely random polarized signals, a rotating linearly polarized source can again be used as the test signal. In this case, however, ellipticity measurements are obtained on each of the antennas separately, and this data can then be used to complete the punch-through probability by either analytic or plotting techniques. Typical results obtained for an S-band circularly polarized side-lobe suppression system are presented which support the conclusion that in a multiantenna side-lobe suppression system, careful selection of antenna types with regard to polarization tracking characteristics must be used in order to minimize the amount of unwanted signal reception or punch-through.

This paper consists of two parts. Part I, by Giguere and Jamison, discusses transistor circuits capable of regenerating 6.25-m?sec pulses at a 160-mc bit rate. Part II, by Noll, discusses techniques for multiplexing 16 digital signals with a 10-mc clock rate into a single signal with a 160-mc clock rate. Two methods of performing the regeneration function are presented. One method consists of dc level restoration for recognition of the signal and constant current coincident circuitry for the reconstruction of the pulses. The second method consists of operating on changes in the signal for pulse recognition and the use of a bistable circuit for pulse reconstruction. Timing in the second case is obtained by a constant current coincidence gate. Parallel-to-serial multiplexing techniques have been developed to combine sixteen parallel 10-mc clock-rate signals into a 160-mc clock-rate pulse train. The sixteen synchronous signals are applied to sixteen AND gates along with a 10-mc narrow gate pulse. The space separations of the resulting regenerated and timed AND gate output pulses is converted to time separation with only a small amount of signal loss. This is done by injecting the pulses at sixteen equally separated points on a broad-band delay line. Methods have been developed to reduce spurious responses resulting from multiple reflections on the delay line. The current mode transistor AND gates are suitable for AND/OR functions for individual 4-m?sec logic. The multiplexer may also be used as a piece of test equipment to generate repetitive 16-bit binary.

As the voltages and MVA ratings of transformers increase, the weight of even single-phase transformers become so large that shipment as a unit is difficult and field assembly may be necessary. Proposed 1000 MVA single phase, 765/345 kV autotransformers using conventional designs weigh between 200 and 300 tons. This problem will get much worse at 1000 and 1500 kV where the units may have single phase MVA ratings much greater than 1000 MVA. Because of the numerous insulating barriers and the extreme care needed in drying and oil filling, conventional oil-paper units may be difficult to assemble on site. In addition, dielectric losses in oil-paper insulation can be a problem at UHV levels.1

Summary form only given, as follows. Due to the dependence of transistor small-signal parameters on the d-c operating point (emitter current Ie and collector voltage Ve), the gain of a transistor amplifier stage decreases with either decreasing emitter current or decreasing collector voltage. Thus, we have two principles of gain control: Ie-control and Ve-control. The variation of the small-signal parameters and of the gain with Ie, and Ve, is discussed in the first part of this paper on the basis of experimental and analytical results. The appreciable control power necessary to vary I, or V, (of possibly several stages) can be obtained most advantageously by operating the controlled stages as a d-c amplifier of the control signal. AGC systems using Ie, and Ve, control have been developed; they are described and compared. Problems of distortion, detuning, and bandwidth variation are discussed.

This paper deals with one phase of a general study of the microbiological deterioration of rubber and similar insulations. In the earlier papers 2 types of microbial attack were shown: 1) direct eating of natural rubber insulation, 2) production of micropores thru walls of both natural and synthetic rubber insulation, or neoprene jackets.

This paper describes the design advantages and limitations of a hybrid digital-to-analog converter that is capable of a 12-bit resolution with an accuracy that is better than one-half the least significant bit. Through advances made in the research and applications of tantalum - nitride and nickel - chromium resistors, analog conversion systems can be designed to be compatible in size, weight, and power to monolithic digital and linear circuits. The system described here, also indicates why the hybrid circuit design is necessary for obtaining the accuracy required and where the monolithic design joins to optimize the characteristics of both fabrication methods. The basic elements that make up a digital-to-analog converter are the ladder network and the semiconductor switches. In order to produce output voltage steps that are accurate to +0.0122%, the tolerances and matching characteristics of tantalum - nitride resistors are required. It is also extremely vital to have an ideal switch having very low resistance to connect the reference voltage to the ladder. This requirement is best satisfied through the use of low offset chopper transistors and can presently be optimized only in hybrid form. Resolution and accuracy -- two of the most important requirements in digital-to-analog conversion, can be made presently available, only when the advantages of monolithic and hybrid technologies are joined under the conditions described in this paper.

This paper will discuss a miniaturized solid-state transformerless dc-to-dc converter using hybrid techniques that converts a 0.4-V nuclear cell output to 1.35 V, making possible an equivalent mercury cell rated in mA-years rather than mA-hours.

A special-purpose radar model has been developed to provide interim capability in the evaluation of Army radar performance. This irterim model is designed to be used pending development of the radar scoring facility and will be replaced by the empirical prediction model developed from the radar scoring facility. The radar model has three basic sections: (1) cross section and signal simulation, (2) receiver function simulation, and (3) a threshold detection and evaluation section. This paper describes the model and the results of tests which have been performed with the model. The paper also discusses capabilities and limitations of the present form of the model, and proposed model improvements.

This paper describes an instrument for recording depth of oceanographic and other underwater towed gear. A pressure-rated, streamlined housing contains the pressure sensing element and recording mechanism. The housing is rated for depths as deep as 1,000 feet and towing speeds as high as 40 knots. Basically, the depth sensing element works on the principle of displacing a precision-made, spring-loaded piston. The piston is displaced an amount proportional to the water pressure. A self-contained battery-powered drive mechanism makes a permanent record of the towing operation depth history. In this paper, specific details are given for the principle of operation, design of the streamlined housing, design for a specific depth range, drag predictions, expected error, calibration techniques, and maintenance procedures.

This paper will discuss 29-gate ECL integrated circuits with a 400-ps loaded delay, Proposed current sources for ECL's and a transistor base reduced by unique self-aligning technique contribute to the low-power, high-speed LSI.

This paper will discuss a linear IC designed and fabricated to perform the functions of sync separation, noise inverson, and AGC amplification. The IC uses P-channel MOS units as active transistors, diodes, and resistors.

Electrical and thermal interactions in silicon integrated circuits provide a means for realizing large effective time constants and, correspondingly, low cutoff frequencies in filters. This paper describes a method of analyzing the behavior of a general configuration of lumped or distributed heaters and sensors in an integrated circuit. The method yields results that are amenable to computer evaluation, and that also indicate how heaters and sensors should be shaped to obtain a particular type of response. A practical synthesis procedure is given for filter applications, and experimental results are presented that show close agreement with calculated values for examples of both analysis and synthesis of filters.

A RECENTLY-DESCRIBED single-dimension self-scanning photodiode array has been seminal to a growing number of modified structures for a variety of applications. The MOS shift register, which controls array scanning, has in particular demonstrated considerable versatility and applicability for different array designs. This paper will discuss the tradeoffs in shift register and photodiode structure which set the limits on array performance.

This paper will describe a composite power transistor, using a new technique for reducing lead inductance, that is capable of 17-W saturated and 11-W linear output, dc-500 MHz, with a gain-bandwidth product of 2.5 GHz.

This paper will describe a new differential amplifier, the dual of conventional amplifiers, whose output is proportional to the differential-mode input current, while common-mode currents are rejected. High common-mode rejection can be maintained over wide bandwidths.