Showing papers on "Butterworth filter published in 1989"

Adaptive line enhancement by using an IIR lattice notch filter

Abstract: Two simple methods for retrieving a single sinusoid corrupted with noise are proposed. They are based on the lattice form realization of an adaptive infinite-impulse-response (IIR) notch filter. The IIR filter is a cascade of second-order all-pole and all-zero filters, and the coefficients of the finite-impulse-response (FIR) section are adapted. The proposed algorithms keep the poles of the filter inside the unit circle. The computer simulation results show that the algorithms have considerable potential in adaptive notch filter applications, especially when the input signal-to-noise ratio is low. >

High-frequency fully differential filter using operational amplifiers without common-mode feedback

Abstract: A technique for realizing a fully differential filter using an operational amplifier (op amp) without common-mode feedback (CMFB) is presented. This technique results in a smaller area, a reduced power dissipation, and an improved speed and dynamic range of the filter. Experimental results of a biquad switched-capacitor bandpass filter with a center frequency of 3.5 MHz and a Q of 24 are presented. The op amp has an open-loop unity-gain bandwidth of 230 MHz, a phase margin greater than 50 degrees , and a DC gain of 76 dB. The prototype filter occupies only 0.26 mm/sup 2/ and dissipates 6 mW with a 5-V supply. >

Normalized frequency domain LMS adaptive filter

Abstract: A normalized frequency domain Least Means Square filter. The feedback coefficient for each frequency bin is adjusted separately in the filter by use of an input power estimate. The power estimate is incorporated directly into the filter algorithm as a data-dependent, time-varying stochastic feedback coefficient. The filter is particularly useful in applications in which the input signal has large spectral variations, which could lead to filter instabilities in some frequency bins if a single feedback coefficient were employed in all frequency bins.

Digital filtering of EMG-signals.

Abstract: Even under field conditions the registration of EMG-signals is often interfered with artifacts. For quantitative signal processing it is necessary to eliminate artifacts and noise from the EMG raw data. We show a simple and efficient way with the help of a recursive digital first order Butterworth filter. This filter-type has a great slope steepness and requires minimal calculations. The algorithm is derived and easily programmable.

Optical fibre digital pulse-position modulation

Abstract: A performance analysis has been presented for digital PPM transmitted over an optical fibre channel and detected using both optimum and sub-optimum pre-detection filters. Receiver sensitivity calculations, carried out at a bit-rate of 140Mbit/s and a wavelength of 1.3 μm, show that the optimum digital PPM system considered offers an 8.6 dB improvement over a typical PCM system. The sub-optimum pre-detection filters considered were a matched filter, an optimised 3-pole filter and a third-order Butterworth filter. These led to sensitivity degradations of 0.4 dB, 0.9 dB and 1.1 dB respectively. This clearly illustrates that receiver complexity can be simplified without large reductions in sensitivity. In particular, the well known and simple Butterworth filter can be employed with only 1.1 dB degradation in sensitivity. The timing requirements for digital optical fibre PPM have been analysed. An original spectral characterisation of the PPM format using its cyclostationary properties has been presented. The characterisation was used to evaluate the inherent systematic jitter associated with the extracted slot clock. An optimisation of the extracted slot clock timing variance and system wrong slot errors (due to imperfect slot synchronisation) was shown to be feasible in terms of the PLL bandwidth and the PPM order. Frame synchronisation was analysed for an original class of frame synchronisers that utilises natural sequences. The extracted frame clock timing variance was evaluated and the probability of wrong slot errors due to the non-ideal frame clock was assessed. The frame clock timing variance and wrong slot errors were shown to be minimisable provided that the proper number of natural sequences is tracked and the appropriate PLL bandwidth is utilised. The analysis has provided a performance evaluation of the optical fibre PPM system in the presence of inherent systematic slot and frame jitter.

Multiple bandwidth filter system for phase locked loop

Abstract: A method and apparatus for combining a plurality of different filtering paths of varying bandwidth to provide selectable bandwidth signal filtering without transients caused by initialization of filter components. The signal to be filtered is applied to a plurality of parallel connected passband filters ranging from a widest passband filter to a narrowest passband filter. The bandwidth of the parallel filter combination is progressively narrowed by sequentially holding the output level of each of the filters in declining order of their passband frequency range, beginning with the widest passband filter, until all of the output levels of the filters are held constant except for the narrowest passband filter. This filter system is especially adapted for use as a loop filter in a phase locked loop control system incorporated in a data storage disc drive.

Radio receiver circuit with tuning feedback control for a tunable bandpass filter

Abstract: A zero-IF radio receiver circuit comprises an input filter (3), quadrature mixers (9,10), d.c.-blocking capacitors (24,26) and a demodulator (22). The circuit is partly integrated on a semiconductor chip and an inductive component of the input filter comprises one or more of the chip bond-wires. In order to compensate for the inevitable variation of the inductance of these bond-wires from circuit to circuit part of the signal from the local oscillator (12) is added to the input signal before its application to the input filter and the d.c. component of the resulting output from one mixer (10), which component is representative of the phase shift of the local oscillator signal produced by the input filter and hence of any tuning error of this filter, is applied to a tuning control input (30) of the filter to reduce the error. Alternatively the tuning control signal may be adjusted to maximize the sum of the squares of the d.c. components of the signals in the two IF channels and hence minimize the attenuation produced by the input filter.

An efficient FIR adaptive filter using DPCM and the sign algorithm

Abstract: An adaptive filter structure that requires zero multiplications for its implementation is introduced. The primary input signal is quantized using differential pulse code modulation (DPCM), and the DPCM outputs are processed by a conventional adaptive filter. The filter coefficients are updated using the sign algorithm. Under the assumption that the primary and reference input signals to the adaptive filter are zero-mean, wide-sense stationary, and Gaussian, theoretical results for the mean and mean-squared behaviour of the filter are also derived. A simulation example demonstrating the usefulness of the filter, as well as verifying the theoretical results and comparing the multiplication-free adaptive digital filter to the sign algorithm, is also presented. >

Frequency interpolation filter bank

Abstract: A technique is presented for designing a bank of filters with the interpolating property: given N arbitrary frequencies (f/sub m/), the mth filter has unity response at the frequency f/sub m/ and zero response at the other frequencies. The filter bank realization may be viewed either as a statistic digital filter or as an adaptive filter wherein N oscillators are summed with time-varying complex weights; the modulated weights represent the filter bank outputs. These filter banks are useful as front-end processors for telephone tone receivers, acoustic signal processing, and general spectrum analysis. >

Analog and Digital Filtering of the Brain Stem Auditory Evoked Response

Abstract: This study compared the filtering effects on the auditory evoked potential of zero and standard phase shift digital filters (the former was a mathematical approximation of a standard Butterworth filter). Conventional filters were found to decrease the height of the evoked response in the majority of waveforms compared to zero phase shift filters. A 36-dB/octave zero phase shift high pass filter with a cutoff frequency of 100 Hz produced a 16% reduction in wave amplitude compared to the unfiltered control. A 36-dB/octave, 100-Hz standard phase shift high pass filter produced a 41 % reduction, and a 12-dB/octave, 150-Hz standard phase shift high pass filter produced a 38 % reduction in wave amplitude compared to the unfiltered control. A decrease in the mean along with an increase in the variability of wave IV/V latency was also noted with conventional compared to zero phase shift filters. The increase in the variability of the latency measurement was due to the difficulty in waveform identification caused by the phase shift distortion of the conventional filter along with the variable decrease in wave latency caused by phase shifting responses with different spectral content. Our results indicated that a zero phase shift high pass filter of 100 Hz was the most desirable filter studied for the mitigation of spontaneous brain activity and random muscle artifact.

Hybrid form digital filter

Abstract: A hybrid form digital filter. A digital filter is provided wherein the zeros of its transfer function are provided by a direct form digital filter portion and the poles of the transfer function are provided by a normal form digital filter portion. The hybrid filter provides truncation noise and stability characteristics like those of a normal form digital filter, but its filter coefficients may be calculated more rapidly than those for a normal form digital filter having the same transfer function, thereby enabling the coefficients to be calculated in real time for tuning the filter. While the preferred embodiment is a second order hybrid digital filter, the same principles are applicable to higher order digital filters as well.

Switched filter for the compensation of power processors

Abstract: An analog-amplitude discrete-time recursive filter is proposed for the frequency compensation of power processors. It has been applied to a converter of the type described by S. Cuk (1986), and it was allowed to reach an almost ideal profile of G/sub loop/. The filter copes with converters that chop up to 0.5 MHz and seems to be better than time-invariant compensation networks and DSP-based filters. Exact small-signal analysis of the switching cell and of the compensation filter is carried out, and experimental results of gain and phase frequency dependencies are reported. >

Real-time determination of IIR coefficients for cascaded Butterworth filters

Abstract: Beginning with an s-domain description of the basic second-order RCL filter network, the digital filter equivalent is derived using the bilinear transformation. The resulting formulas describing the z-domain digital response are algebraically reduced to a form that gives some intuitive insight into the fundamental symmetry relationships between analog and digital filter structures. A simplified description of the digital filter coefficients whose form lends itself to real-time computational techniques is presented. This analysis is extended to the cascaded second-order network characterized by Butterworth damping factors, so that higher order filters can be designed and implemented in real-time using high-speed digital signal processors such as the DSP 56001 and DSP 96002. >

Signal tracking electronically tunable filter

Abstract: A tunable bandpass filter for radio frequency energy with a phase-locked loop for tracking an input signal and to control the filter to keep the center frequency of the passband coincident with the frequency of the input signal is shown Using a Yttrium Iron Garnet (YIG) filter as a frequency determining element and as a passive dispersive reference element for a frequency discriminator, the bandpass filter uses the output signal of the discriminator to form a fine tuning signal to control the center frequency of the passband of the YIG filter

Variable rate rectangular matched filter

Abstract: A novel variable rate near perfect ractangular matched filter (Fig. 2) is provided with a low pass filter (22) coupled to the input symbol data stream (11). The output of the filter is coupled to a sampler (24) and the output (25) of the sampler (24) is coupled to an analog to digital converter (26) to provided digital samples (at 27) indicative of the data stream at a time occuring between input data symbols. The output (27) of the analog to digital converter (26) is applied to a digital adder (28) which has a filter correction input (at 32) factor to adjust the output (at 33) of the digital adder (28) so that the digital samples approximate very closely the output of a perfect rectangular matched filter (Fig. 1).

Sound field correction device

Abstract: PURPOSE: To obtain a natural sound in terms of listening sense by providing a peak correction filter to smooth only a steep peak in a transmission characteristic of a sound field correction filter. CONSTITUTION: An impulse response at a listening point measured in advance is inputted to an input terminal 1 of a filter coefficient calculation means 6 and a desired impulse response is inputted to an input terminal 2, then they are discrete Fourier transformed by an FFT(Fast Fourier Transform) unit 61. A filter characteristic calculator 62 receiving the output calculates the filter characteristic and the peak is corrected by a peak correction filter 63 and inverse discrete Fourier transformation is applied by an inverse FFT 64 to obtain a correction coefficient, which is set in the sound field correction filter 2 and the filter characteristic is obtained. Then the characteristic of a transfer function of the sound field correction filter 2 is preserved and only a steep peak is smoothed. Thus, a natural signal in terms of listening sense is reproduced from a speaker 5. COPYRIGHT: (C)1990,JPO&Japio

An integrated continuous-time bipolar transconductor-capacitor filter

Abstract: The design and some results obtained from a low-frequency low-pass transconductor-capacitor filter integrated in a standard bipolar process are discussed. A comparison is made between the results of the transconductor-capacitor filter design approach and of a gyrator filter design approach. The comparison indicates the advantages of the transconductor-capacitor design over a gyrator filter design. >

Radio receiver circuit arrangement

Abstract: A zero-IF radio receiver circuit comprises an input filter (3), quadrature mixers (9,10), d.c.-blocking capacitors (24,26) and a demodulator (22). The circuit is partly integrated on a semiconductor chip and an inductive component of the input filter comprises one or more of the chip bond-wires. In order to compensate for the inevitable variation of the inductance of these bond-wires from circuit to circuit part of the signal from the local oscillator (12) is added to the input signal before its application to the input filter and the d.c. component of the resulting output from one mixer (10), which component is representative of the phase shift of the local oscillator signal produced by the input filter and hence of any tuning error of this filter, is applied to a tuning control input (30) of the filter to reduce the error. Alternatively the tuning control signal may be adjusted to maximise the sum of the squares of the d.c. components of the signals in the two IF channels and hence minimise the attenuation produced by the input filter.

Variable bandwidth crystal filter with varactor diodes

Abstract: A crystal filter with continuously variable passband, and a constant input and output impedance using quartz crystals, fixed inductors and varactor diodes. The width of the filter passband is set by a DC voltage on a single control line through a resistor biasing network.

A low-complexity multiplierless half-band recursive digital filter design

Abstract: A low-complexity multiplierless design of a half-band recursive digital filter is presented. The filter structure is realized as a bireciprocal lattice wave digital filter. The filter coefficients are represented in a canonic signed-digit code with only two nonzero digits, and thus each filter coefficient can be implemented with only a single adder or subtracter. A fifth-order elliptic filter section is presented which requires only nine adders and five delays. Thus, a cascade of three or four of these fifth-order building blocks could be implemented on a single VLSI chip to obtain a high-performance half-band filter. >

1 Gbit/s integrate-and-dump filter for digital communication systems

Abstract: An integrate-and-dump filter operating at 1 Gbit/s has been demonstrated. For NRZ signal pulses, noise filtering with the I&D circuit yielded the same error rate performance as an 800 MHz lowpass filter. Unlike the lowpass filter, the I&D filter can reduce the degradation caused by certain kinds of timing jitter, and introduces no intersymbol interference.

Filter design for polyphase filter banks with arbitrary number of subband channel

Abstract: Consideration is given to the filter design problem for a recently proposed polyphase filter bank with an arbitrary number of subband channels. On the basis of the perfect reconstruction condition for the 1-D case, an analytical formula for the filter design is developed. Compared with direct numerical design algorithms, this analytical formula allows one to design the required low-pass FIR prototype filter with much less computational complexity and to obtain better filter bank performance. A design example is given for illustration. >

Linear phase IIR filter design: a model reduction approach

Abstract: A method is presented for the design of a linear-phase IIR filter. It involves designing a FIR filter satisfying the given frequency response specifications and subsequently obtaining a significantly lower-order IIR filter using model reduction. The method is illustrated by a design example and is compared with a direct method for the design of linear-phase IIR filters using equalizers. The filter designed using the proposed technique gives a better constant group delay for an almost equal order. >

Computer-aided design of evanescent mode waveguide bandpass filter with nontouching E-plane fins

Abstract: The authors present an efficient computer-aided-design procedure that systematically utilizes a lookup table containing scattering parameters from variously dimensioned E-plane fins. The main idea is to achieve an optimal combination of filter parameters by proper selection of the E-plane fins from the table and appropriate determination of the other filter elements in order to satisfy the given filter specifications. The technique of selecting the proper fin from the table is explained. The relationship between the desired center frequency of the filter and the approximate resonant frequency of the single fin in the table is shown. The relationship between each design parameter and the filter characteristic is presented. The algorithm is applied to the design of a bandpass filter with two E-plane fins operating in the Ka-band. The algorithm is verified by comparing the characteristics of the designed filter with the experimental results. >

FDT-a design tool for switched capacitor filters

Abstract: The authors present FDT (filter design tool), a complete design tool for switched-capacitor filter synthesis from specifications to circuit schematic generation. The tool supports both ladder and biquad structure synthesis. It provides analysis and optimization capabilities at both pole-zero level and circuit level. It also extends these capabilities for noise calculation and noise minimization of the synthesized filter. These facilities together with a completely interactive and flexible user interface provide an efficient design tool for filter synthesis and design evaluation in a short time. FDT was successfully used in the design of a high-pass filter and two delay equalizers in an analog interface chip. >

Design of FIR filters with reduced computations

Abstract: The performance of a symmetric nonrecursive filter can be improved by multiple use of the same filter. The method is based on an Amplitude Change Function (ACF). An approach to the design of nonrecursive filters using an ACF is discussed in this paper. The prototype filter chosen is a Recursive Running Sum (RRS) filter which does not require any multipliers for its implementation. The required filter specifications are met by multiple use of the RRS filters. The overall filter requires a much smaller number of multiplications and adders than the one designed using the conventional method. It is shown that this method provides reduced noise due to coefficient quantization and product quantization compared with the conventional design technique.

3-D f-k filtering

Abstract: Stewart and Schieck (1993) show that a conventional 2-pass 2-D f-k filter fails to provide an axially-symmetric response. They suggest using a true 3-D one-pass filter for specific pass ranges and symmetric results. They also approximate a true 3-D f-k symmetric response by axially rotating a 2-D f-k filter. In this paper a true 3-D f-k filter will be compared with this 2-D rotated filter. During the investigation, it is found that an interpolation scheme is required when rotating a discrete filter to avoid an aliasing problem. A special 2-D filter that has a 2-pass symmetric 3-D f-k response can be defined. This Gaussian shaped filter is applied in two passes and produces a symmetric response, however it has a very gradual cut-off region.

A new approach to the design of bilinear-LDI switched-capacitor filters having low passband sensitivity

Abstract: An approach to the design of low-sensitivity switched-capacitor (SC) filter is described. In this approach, the continuous-time reference transfer function is decomposed into a sum of two individual functions, and each function is realized as the transfer function of a voltage-divider network consisting of a resistance as its series arm and a reactive impedance as its shunt arm. The bilinear-LDI (lossless discrete integrator) design technique is applied to the SC realization of the two voltage-divider networks. These individual realizations are combined to form the SC realization of the overall filter. The resulting filter requires n+1 operational amplifiers (OAs) for its realization, where n is the order of the reference transfer function. For illustration purposes, the proposed approach is applied to the bilinear-LDI SC design of a practical sixth-order elliptic bandpass filter. It is shown that the filter exhibits low sensitivity to dominant-pole OA effects. >

Application of symmetry results in 3-D digital filter design

Abstract: The authors present a systematic way of designing a three-dimensional (3-D) filter, using the symmetries in the given specifications A step-by-step procedure for optimization-based design is given To illustrate the advantage of this procedure, two filter design examples are presented The first example is the design of a 3-D spherically symmetric low-pass filter The second is the design of a conic filter, similar to a fan filter in two dimensions The examples show a significant reduction in CPU time >