A. Arbel

Output stage for current-mode feedback amplifiers, theory and applications

A novel building block is described, termed FCS (floating current source), which may serve as class A output stage for CFAs (current-mode feedback amplifiers). It is capable of driving a grounded load with a bipolar signal, and yields a feedback current equal to the output current over a wide frequency range. Its possible range of application covers MOSFET amplifiers employed in analog signal processing and current-operated control systems. An internal interconnection converts the FCS into a CCII-. Another novel CCII- configuration employs a push-pull folded cascode and may serve as noninverting input stage for a standard amplifier configuration. Finally, a feedback-stabilized CCII- and a CFA are described, both employing the FCS as output stage.

Comparison between the noise performance of current-mode and voltage-mode amplifiers

A comparison is made between the noise performance of current mode- and voltage mode feedback amplifiers, in terms of their noise figureFi andFv, respectively. WhereasFv of a well designed voltage amplifier is dominantly affected by the input stage only, all stages of a current amplifier including the feedback network may potentially contribute toFi. This paper investigates the transferfunction to the output of the various noise sources of current mode feedback amplifiers, whose full recognition is essential in order to design CMFAs of reasonable noise performance. Also discussed is the effect the dynamic range of a postamplifier may have on its noise figure. Although the bulk of this paper deals with postamplifiers, the important subject of low noise current amplification is covered as well.

Current operated nucleonic modules

Abstract Current modules represent a unified approach to nucleonic instrumentation employed in low- and high-energy physics. The final system envisaged should be capable of replacing the presently used combination between readily available standard instruments and “custom made” special purpose circuits by a single line of instruments, whose use will shorten significantly the delay between the conception of a particular experiment and its execution.

Negative feedback revisited

Rules are derived for choosing one among the four kinds of basic amplifiers to be embedded in a particular passive feedback network. The loop transmission is identified and conditions for mismatch are obtained from a unified analysis of feedback circuits. Past experience gained with voltage mode basic amplifiers, the laws of duality and mismatch are applied to choose the building blocks employed in the design of the remaining three kinds of basic amplifiers. This intuitive synthesis is confirmed by formal network transformations employing nullors.

Current-mode feedback amplifier employing a transistorized feedback network

Current mode feedback amplifiers providing a linear transfer function have the unique property, that their feedback network can be designed employing nonlinear devices instead of resistors. The paper describes a current mode amplifier employing such a transistorized feedback network consisting of n identical grounded- and one feedback transistor, yielding a gain of n+1. Control of the corresponding gate voltages makes it possible to cut off j individual transistors, thereby reducing the gain to n-j. Employing 5- and 2 GHz complementary bipolar transistors and electronic gain control, a bandwidth of 375/spl divide/300 MHz has been obtained at a variable gain between 10 and 100.<<ETX>>

Multistage Transistorized Current Modules

The choice of current as signal parameter leads to the design of amplifiers, in which voltage swings are minimized throughout. These amplifiers utilize the full accidental current gain of the transistors by employing a particular class of feedback systems, whose natural frequencies approach asymptotically a well-defined value as the loop-gain at dc approaches infinity. This design method provides complete control by external circuitry over transistors by grouping them in pairs or triples into building blocks which may be suitably interconnected into complete feedback amplifiers comprising six or more transistors in a single loop. Two practical examples are given for commerical current amplifiers( Manufactured by Elron Electronic Industries, Ltd., Haifa, Israel.) incorporating these design principles.

Towards a Perfect CMOS CCII

DC offset and inaccurate voltage follower action betweenterminals Y and X, the relatively highimpedance seen at terminal X and inaccurate currentfollower action between terminals X and Zplague simple circuit realizations of the CCII. This paper summarizesthe methods employed and compares the improvement of the CCIIcharacteristics achieved by a variety of published circuits,with emphasis laid on CMOS realizations. In these comparisons,the conventionally employed but inaccurate current mirror outputstages were replaced by a circuit employing current isolationand thereby providing accurate current follower action betweenterminals X and Z. Furthermore, twonovel realizations of a differential CCII – basedtransconductance amplifier are described, which can also be employedas input stage of an instrumentation voltage amplifier exhibitingan exceptionally high CMRR and excellent frequency response.Finally, a current fan out circuit for cascode CCII –‘swill be introduced, whose transfer function value is electronicallyadjustable.

Innovative current sensitive differential low noise preamplifier in CMOS

A low noise differential preamplifier is described. A novel circuit is introduced, reducing the 2nd stage noise contribution of a conventional current mirror fed from the differential input stage to negligible amounts by employing a differential cascode. A desirable by-product of the modified circuit is its increased PSRR. The output of the charge sensitive voltage amplifier employed as low noise input stage is reconverted into current by a Y/sub T/ amplifier consisting of a CCII-, whose X terminal output impedance is reduced by including it in the feedback loop of the input amplifier, and by augmenting the CCII- by a current follower. The available current gain is 10/sup 4/ at a bandwidth of 80 kHz and a loop transmission of 90 dB.

Snap-off constant fraction timing discriminators

A snap-off constant fraction timing discriminator has been designed, involving no critical adjustments and free of shape limitations. The instrument accepts both short pulses from photo-multipliers (using fast or slow scintillators) and tail pulses from low-noise preamplifiers. Its performance (without rejecting events close to the detector junction) was checked with the aid of fast and slow scintillators and Ge(Li) detectors. Typical time resolutions (measured at a dynamic range of 100:1) were 0.41 nsec FWHM or 0.92 nsec FWTM for Co60 with a pair of fast scintillators, and 3.85 nsec FWHM or 8.10 nsec FWTM for Na22 with a 7% efficiency closed-end Ge(Li) detector in the stop channel.

Pile up rejection by comparison of the shaped pulse with its second derivative

Abstract A pile up rejector is described for use in a high resolution nuclear spectrometer operating at high counting rates. The principle of pile up rejection described is based upon the “pile up inspection signal”, which is obtained by a summation of the shaped pulse and its second derivative. If the summation is performed with the proper constant of proportionality, then the sensitivity to pile up is dependent of the amplitude of the pulse under scrutiny. It is a function only of the delay between that pulse and the one causing pile up, and the latters amplitude. An expression has been derived for ideal Gaussian shaping, defining the amplitude and delay of those pile-up pulses which introduce a certain minimum error in measuring the pulse under scrutiny. The graph thus obtained provides a criterion for the performance comparison of pile up rejectors employing various kinds of shaping. Results are presented, which were obtained with a practical system employing near Gaussian shaping.