Translinear Principle Research Articles

Winner-takes-all associative memory: a Hamming distance vector quantizer

We present a design methodology for mapping neuraly inspired algorithms for vector quantization, into VLSI hardware. We describe the building blocks used: memory cells, current conveyors, and translinear circuits. We use the basic building blocks to design an associative processor for bit-pattern classification; a high-density memory based neuromorphic processor. Operating in parallel, the single chip system determines the closest match, based on the Hamming distance, between an input bit pattern and multiple stored bit templates; ties are broken arbitrarily. Energy efficient processing is achieved through a precision-on-demand architecture. Scalable storage and processing is achieved through a compact six transistor static RAM cell/ALU circuit. The single chip system is programmable for template sets of up to 124 bits per template and can store up to 116 templates (total storage capacity of 14 Kbits). An additional 604 bits of auxiliary storage is used for pipelining and fault tolerance re-configuration capability. A fully functional 6.8 mm by 6.9 mm chip has been fabricated in a standard single–poly, double–metal 2.0µmn–well CMOS process.

Translinear circuit for phase detection

Based on Gilbert's translinear principle, the multiplication and filtering within an analogue phase detector can be combined into a single nonlinear filtering operation. This is illustrated by the design of a simple translinear phase detector.

Computer aided analysis and design of MOS translinear circuits operating in strong inversion

Recently, it was proposed to generalize the well-known translinear circuit principle in such a way that it also applies to MOS transistors operated in strong inversion. In this paper, the MOS translinear (MTL) principle will be briefly reviewed. A graphical analysis method for MTL-circuits is presented, which can also be applied to bipolar translinear circuits. This graphical method was implemented in a computer program, which is now used as an interactive design tool to implement nonlinear signal processing functions by MTL circuits.

Editorial commentary on papers for special issue on translinear circuits

Editorial commentary on papers for special issue on translinear circuits

Variable-gamma circuit for colour television based on the MOS voltage-translinear principle

In this paper a variable-gamma system is discussed for application in high-quality colour television receivers. A variable-gamma system can significantly improve the quality of the video image by enhancing the contrast in darker parts of the image without degrading the lighter areas of the screen. The adjustment of gamma can replace the traditional contrast and intensity adjustments. The main part of the variable-gamma system is a nonlinear function block, which is implemented by an MOS translinear (MTL) circuit. This circuit was realized in a standard 2.5 μm CMOS process. The circuit operates at a single 5V supply voltage. The measured small-signal-3dB bandwidth is 20MHz. The circuit response is insensitive to temperature, which is an important property of translinear circuits.

Translinear circuits in subthreshold MOS

In this paper we provide an overview of translinear circuit design using MOS transistors operating in subthreshold region. We contrast the bipolar and MOS subthreshold characteristics and extend the translinear principle to the subthreshold MOS ohmic region through a drain/source current decomposition. A front/back-gate current decomposition is adopted; this facilitates the analysis of translinear loops, including multiple input floating gate MOS transistors. Circuit examples drawn from working systems designed and fabricated in standard digital CMOS oriented process are used as vehicles to illustrate key design considerations, systematic analysis procedures, and limitations imposed by the structure and physics of MOS transistors. Finally, we present the design of an analog VLSI “translinear system” with over 590,000 transistors in subthreshold CMOS. This performs phototransduction, amplification, edge enhancement and local gain control at the pixel level.

Translinear circuits: An historical overview

The concept of a translinear circuit is now widely appreciated and applied. This paper traces the history of the concept, delineates the original meaning of the term, explains the basic principles of this class of circuit and discusses recent, more general, interpretations of the term. It is recommended that the word translinear, used without further qualification, should be reserved exclusively for those cells invoking exponential device behaviour, which applies to all bipolar transistors, including heterojunction types, and to MOS transistors operated in the subthreshold, or weak inversion, domain, but does not apply to normal MOS operation in strong inversion. It is proposed that the theory and practice related to such ‘quadratic’ operation of MOS devices, where the transconductance is presumed to be linear with gate-source voltage, should be termed voltage-translinear, or VTL.

Dynamic translinear RMS-DC converter

Translinear, log-domain or exponential state-space filters constitute only a subclass of dynamic translinear circuits. The dynamic translinear principle can also be used to implement nonlinear dynamics. This is illustrated by the design of an RMS-DC converter.

Application of the back gate in MOS weak inversion translinear circuits

Though the MOS transistor is a four-terminal device, it is most often used as a three-terminal device. Therefore, a large number of possible MOS circuits are overlooked. In this brief, the four-terminal point of view is elaborated with respect to MOS weak inversion translinear circuits, a class of circuits naturally very suitable for low-voltage and low-power applications. Some new circuits are described which sometimes are more suitable for low-voltage applications than bipolar translinear networks performing the same function. It is also shown that, using the back gate, translinear networks can be derived which cannot be realized with bipolar transistors. These network topologies increase the possibilities offered by translinear technology. As an example, measurement results of a low input-voltage current mirror and a sin(x)-circuit are shown.

Current-mode subthreshold MOS implementation of the Herault-Jutten autoadaptive network

The authors explore translinear circuits in subthreshold MOS technology and current-mode design techniques for the implementation of neuromorphic analog network processing. The architecture, also known as the Herault-Jutten network, performs an independent component analysis and is essentially a continuous-time recursive linear adaptive filter. Analog I/O interface, weight coefficients, and adaptation blocks are all integrated on the chip. A small network with six neurons and 30 synapses was fabricated in a 2- mu m double-polysilicon, double-metal n-well CMOS process. Circuit designs at the transistor level yield area-efficient implementations for neurons, synapses, and the adaptation blocks. The authors discuss the design methodology and constraints as well as test results from the fabricated chips. >

Monolithic integrated magnetic sensor realizing omnidirectional measurement

A monolithic integrated magnetic sensor realizing omnidirectional measurement is presented. This monolithic integrated sensor consists of three Hall cells which detect magnetic components in x, y, and z directions and the bipolar circuit which converts the output signal of Hall cells VX, VY, and VZ into its absolute value (=(V2X + V2Y + V2Z)1/2). For the magnetic detector, a new “lateral Hall cell” is proposed to detect the magnetic components parallel to the chip surface (x, y directions). The three-dimensional (3-D) magnetic detection has been realized by combining two lateral Hall cells and a conventional epitaxial (lateral) Hall cell (for z direction). The signal conversion circuit uses a Gilbert multiplier for square calculation and a translinear circuit for a root circuit, with a total device count of 158 elements. The test device realized an omnidirectional magnetic measurement with a maximum error of ±4 percent, sensitivity of 14 V/T, and a temperature coefficient of 2100 ppm/°C in the magnetic flux range of −250 to +250 mT.

Current-mode subthreshold MOS circuits for analog VLSI neural systems

An overview of the current-mode approach for designing analog VLSI neural systems in subthreshold CMOS technology is presented. Emphasis is given to design techniques at the device level using the current-controlled current conveyor and the translinear principle. Circuits for associative memory and silicon retina systems are used as examples. The design methodology and how it relates to actual biological microcircuits are discussed.

An analog sinusoidal frequency-to-voltage converter

An analog sinusoidal frequency-to-voltage (F/V) converter, based on the use of nonlinear analog circuits, is introduced. The realization is composed of a differentiator, an integrator, and a translinear divider and square-rooter circuit. The proposed F/V converter can accurately and linearly convert a sinusoidal signal frequency into an output voltage, with fast response and low error, over more than two decades of frequency range. The method can also be implemented in monolithic integrated form. >

Generalized translinear circuit principle

It is shown that a generalized interpretation of the translinear (TL) principle leads quite naturally to an extension to MOS circuits. It is shown that two distinct classes of TL circuits exist, one suitable for bipolar and the other for MOS implementation. The MOS-translinear (MTL) circuit principle is derived and an initial classification of simple MTL circuits is proposed. Some examples are given of MTL circuits synthesizing nonlinear functions. >

Monolithic silicon magnetic compass

Abstract In this report, an integrated magnetic sensor which offers the direction of the applied magnetic field is presented. This integrated sensor contains a two-dimensional magnetic detector which detects the in-plane magnetic vectors, Bx and By, and a signal conversion circuit which calculates the direction θ (=tan−1(By/Bx)) using the translinear circuit technology. This integrated magnetic sensor has not only the direction output but the output for the intensity (the absolute value) of the applied magnetic field. The test device was fabricated using the standard bipolar analog IC technology and its characteristics were examined. As a result, the direction of the applied magnetic field was obtained with maximum error of ± 2% full-scale deflection.

High performance current processing allpass filters

A general configuration for a first order current processing allpass filter is described. It requires an active element having unity current gain in conjunction with two passive components only (i.e., one resistance and one capacitance). Two implementations using the following translinear circuits as active elements: (a) a negative impedance converter with current inversion (I-NIC) and (b) a current controlled current source (CCCS) have been inspected and experimentally compared. Because of the high performances obtained, they appear to be suitable for wideband applications.

A wide-band technique for vector summation and RMS-DC conversion

A circuit configuration is presented for wideband analog computation of both the vector-sum function and the RMS-DC conversion function. Design of the circuit is based on a systematic approach to translinear circuit synthesis. Errors due to finite transistor current gain are compensated by means of a novel technique. Measured performance results based on an experimental monolithic implementation in a standard bipolar process are presented. Bandwidths of 25 and 80 MHz are achieved for vector summation and RMS-DC conversion, respectively.

The use of bipolar semiconductor junctions in linear circuit design

The concept of translinear circuits has been around for many years, dating back to the use of semiconductor diodes as a dc controlled wide-band attenuator in 1961. With the advent of bipolar integrated circuit technology such circuits have proved to be very effective in providing many functions in linear ICs. The translinear concept basically relies on utilising, to advantage, the fundamental logarithmic characteristics of a semiconductor junction and the ability to match device parameters very closely on integrated circuits. This paper will present the fundamental relationship of a bipolar semiconductor junction and show how it affects the operation of known circuit configurations and also will introduce new circuit configurations taking advantage of the fundamental junction properties. The semi-custom Monochip is ideally suited to these applications due to the availability of large numbers of identical geometry, matched transistors.

Simple, wide-range approximations to trigonometric and inverse trigonometric functions useful in real-time signal processing

By means of straightforward methods, simple approximations are derived for the cosine, sine and tangent functions as well as for their inverse functions. The approximations are simple in the sense that they contain few terms and have numerical coefficients with, at most, three significant figures. The trigonometric approximations presented are valid in all four quadrants, including the singularity points in the case of the tangent. The inverse trigonometric approximations are valid over the full natural argument intervals. The absolute approximation errors are shown to be near-minimax with peak values all less than 0.25% of maximum. The results are expected to be useful in real-time signal processing circuits and systems where speed and simplicity rather than extreme accuracy are the dominating factors. A translinear circuit realisation example is given in the Appendix.

Sinusoidal frequency doubler

A translinear circuit is described that can perform exact frequency doubling, free of any d.c. component and with linear amplitude transfer. The basic circuit consists of matched-transistor monolithic arrays. Theory and circuit operation with results are given.