scispace - formally typeset
Search or ask a question

Showing papers on "Butterworth filter published in 1977"


Journal ArticleDOI
J. Kaiser1, R. Hamming
TL;DR: A simple, powerful method for suitably combining the results of several passes through the same filter is described in detail, and its computational efficiency is compared to the best possible filter designs meeting the same specifications.
Abstract: When processing data by filters, we often find it necessary to improve the performance of the filter, either by increasing the out-of-band rejection (loss) or by decreasing the error in the passband, or both. A first approach is to process the data by repeated passes through the same filter. Each pass, while increasing the out-of-band loss, also increases the passband error, often to an undesirable level. It also increases the length (order) of the equivalent filter. How can we do a better job of filtering by suitably combining the results of several passes through the same filter? By "better" we mean both less passband error and greater out-of-band, or stopband, loss. This process is called filter sharpening. A simple, powerful method for doing this is described in detail, and its computational efficiency is compared to the best possible filter designs meeting the same specifications. The design method, based on the idea of the amplitude change function, is restricted to symmetric nonrecursive (finite impulse response) filters with piecewise constant pass- and stopbands. Several illustrative examples are given.

242 citations


Proceedings ArticleDOI
J. Kaiser1, R. Hamming
09 May 1977
TL;DR: In this paper, the amplitude change function is used to improve the performance of symmetric non-recursive (finite impulse response) filters with piecewise constant pass and stopbands.
Abstract: When processing data by filters we often find it necessary to improve the performance of the filter, either by increasing the out-of-band rejection (loss) or by decreasing the error in the passband, or both. A first approach is to process the data by repeated passes through the same filter. Each pass, while increasing the out-of-band loss, also increases the passband error, often to an undesirable level. It also increases the length (order) of the equivalent filter. How can we do a better job of filtering by suitably combining the results of several passes through the same filter? By "better" we mean both less passband error and greater out-of-band, or stopband, loss. This process is called filter sharpening. A simple, powerful method for doing this is described in detail, and its computational efficiency is compared to the best possible filter designs meeting the same specifications. The design method, based on the idea of the amplitude change function, is restricted to symmetric nonrecursive (finite impulse response) filters with piecewise constant pass and stopbands. Several illustrative examples are given.

137 citations


Journal ArticleDOI
TL;DR: The paper shows that the most practical RLC load can be optimally matched to a resistive generator over a finite frequency band to achieve the Butterworth or Chebyshev transducer power-gain characteristic of arbitrary order.
Abstract: The paper shows that the most practical RLC load can be optimally matched to a resistive generator over a finite frequency band to achieve the Butterworth or Chebyshev transducer power-gain characteristic of arbitrary order, justifying and extending the procedure given in the literature. Explicit formulas for the optimum design of these matching networks are presented, thus avoiding the necessity of using the design curves and solving the nonlinear equations for selecting the optimum design parameters. The significance of these formulas is that they reduce the design problem to simple arithmetic. Illustrative examples are provided to demonstrate the usage of the formulas.

57 citations


Patent
14 Jun 1977
TL;DR: In this article, the use of two-pole crystal bandpass filters improves the frequency stability of the system and significantly reduces post detection noise levels, as compared with conventional limiter/discriminator techniques.
Abstract: Demodulation of FM signals employing an FM feedback loop of the present invention includes a first mixer circuit for mixing an IF input signal with a voltage controlled oscillator signal to down-convert the IF input signal to a predescribed center frequency IF signal. This signal is coupled through a tuned amplifier having an automatic gain control loop and is then filtered in a predetection filter and amplified. The predetection filter is composed of a two-pole crystal bandpass filter having a very narrow bandwidth. The filtered signal is then coupled to a frequency discriminator circuit comprised of a frequency detection second mixer and filter circuit, the frequency detection filter circuit being composed of a second two-pole crystal bandpass filter, the bandwidth of which is an order of magnitude larger than that of the predetection filter. The output of the frequency discriminator is coupled through a loop compensation filter to the voltage controlled oscillator so that changes in the output frequency of the voltage controlled oscillator reduce the deviation of the FM signal in the intermediate frequency and frequency discriminator stages. The use of two-pole crystal bandpass filters improves the frequency stability of the system and significantly reduces post detection noise levels, as compared with conventional limiter/discriminator techniques. The two-pole crystal bandpass filters have a Butterworth configuration. In addition to the FM feedback loop, a squelch circuit is also coupled to the output of the frequency discriminator.

24 citations


Journal ArticleDOI
TL;DR: In this article, a new general second-order digital filter section is proposed and two quantizers used in the filter apply controlled rounding which has the effect that limit cycles are completely eliminated under constant-input conditions.
Abstract: In this paper a new general second-order digital filter section is proposed. One of the two quantizers used in the filter applies controlled rounding which has the effect that limit cycles are completely eliminated under constant-input conditions. Furthermore, the configuration of the filter is such that multipliers can be saved in case the filter is used as an all-pass filter or as a filter with zeros of transmission on the unit circle in the z domain.

23 citations


Journal ArticleDOI
TL;DR: In this article, a filter tuning method based on the match of measured and computed input impedances for a short-circuited filter is described, and two singly terminated filters, an 8-pole Chebyshev filter and a 6-pole pseudoelliptic function filter, are tuned using this method.
Abstract: This paper describes a filter tuning method based upon the match of measured and computed input impedances for a short-circuited filter. Two singly terminated filters, an 8-pole Chebyshev filter, and a 6-pole pseudoelliptic function filter tuned by using this method have demonstrated excellent performance.

23 citations


Patent
05 Dec 1977
TL;DR: In this paper, a bandpass filter is described for filtering a television signal, which has a passband which may be selectively centered for filtering signals in one of two channels of the television receiver.
Abstract: A bandpass filter is described for filtering a television signal. The filter has a passband which may be selectively centered for filtering signals in one of two channels of the television receiver. The television channel at which the filter passband is centered is selectable by applying a voltage to a control terminal. The filter frequency response for each channel selected may be independently controlled and a linear phase response is realized for the selected channel minimizing distortion.

19 citations


Journal ArticleDOI
TL;DR: The applications of the least-mean-square-approximation technique to filter synthesis are discussed, and explicit expressions for the characteristic function of all-pole lowpass filters are derived using a weighted least mean-square error norm as discussed by the authors.
Abstract: The applications of the least-mean-square-approximation technique to filter synthesis are discussed, and explicit expressions for the characteristic function of all-pole lowpass filters are derived using a weighted least mean-square error norm. The weight function depends on one variable parameter which controls the shape of the magnitude response both in the passband and in the stopband. It is shown that most of the filter functions in common use are special cases of this approximation procedure, including Legendre monotonic passband filters as a limiting, degenerate case. Also, a useful generalization of Legendre filter functions is proposed.

11 citations


Journal ArticleDOI
TL;DR: In this paper, a class of low-pass filters with no finite transmission zeros is introduced satisfying a number of phase flatness conditions at the origin and providing more stopband rejection than the corresponding Butterworth filters.
Abstract: Transfer functions of low-pass filters with no finite transmission zeros are introduced satisfying a number of phase flatness conditions at the origin and providing more stopband rejection than the corresponding Butterworth filters. If only one of the available degrees of freedom is used to shape the passband magnitude response, while all remaining are assigned to phase linearity, the largest fractional banwidth of phase linearity is obtained but the maximum passband loss cannot be specified beforehand and increases with increasing the order of the network. It is shown, however, that assigning two parameters to magnitude requirements in the resulting transfer function almost any prescribed maximum passband loss can be met while still retaining excellent phase linearity over a large part of the passband. In addition, a class of transitional filter functions is obtained the frequency and time domain characteristics of which can be adjusted over a wide range of filter specifications by varying one parameter only. It is shown that the time domain characteristics of these filters compare favorably with those for the well-known transitional Butterworth-Thomson filters.

8 citations


Patent
25 Mar 1977
TL;DR: In this article, a filtering system for use in an FM modulating system is described, which consists of a first filter and a second filter, separated by an overshoot compensator, which modifies the output of the first filter in such a way that, when the modified signal is filtered by the second filtering, an output signal will be provided which does not overshoot a preselected amplitude.
Abstract: A filtering system for use in an FM modulating system. The filtering system provides frequency limiting of an input signal and does not overshoot amplitude limitations thereof. This filtering system includes a first filter and a second filter, separated by an overshoot compensator. The compensator modifies the output of the first filter in such a manner that, when the modified signal is filtered by the second filter, an output signal will be provided which does not overshoot a preselected amplitude. The overshoot compensator includes a threshold clipper which separates out any overshoot component generated by the first filter, an amplifier which amplifies the overshoot component, and a subtractor circuit which subtracts the amplified overshoot component from the output of the first filter to provide the modified signal. The second filter is of a type which provides a constant delay for all frequencies within the passband on the filtering system.

8 citations


Journal ArticleDOI
TL;DR: In this paper, a current-tunable 2nd-order active filter is described, which is based on the variation of the transconductance of a bipolar transistor with bias current, which can give a lowpass or band-pass 2ndorder response or can be cascaded to give a high-order all-pole tunable filter.
Abstract: A new current-tunable 2nd-order active filter is described which is based on the variation of the transconductance of a bipolar transistor with bias current. The circuit can give a lowpass or bandpass 2nd-order response or can be cascaded to give a high-order all-pole tunable filter. Measurements have indicated that a tuning range of at least four decades is possible with a dynamic range of 75 dB.

Patent
28 Sep 1977
TL;DR: In this paper, an active filter can be controlled over a 50:1 or greater frequency range by varying the gain of a variable gain amplifier connected in series with the capacitive or resistive element of the multiplier networks, while maintaining a fixed filter characteristic and a constant, near zero DC offset voltage.
Abstract: An active filter, capable of being electronically controlled over a 50:1 or greater frequency range, while maintaining a fixed filter characteristic and a constant, nearzero DC offset voltage. The filter utilizes passive element-multipliers, either capacitor-multiplier or resistor-multiplier circuits, as the frequency determining networks, depending upon whether the filter is a low-pass or a high-pass filter. Tuning is accomplished by varying the gain of a variable gain amplifier connected in series with the capacitive or resistive element of the multiplier networks. Also disclosed is an implementation of the variable gain amplifier which permits the filter's cut-off frequency to be controlled digitally. The filter may be designed to have any desired type of filter characteristic. Several filter sections may be connected in cascade to provide any desired number of filter poles.


Journal ArticleDOI
TL;DR: A novel method for constructing charge-coupled device (CCD) transversal filters is presented, and a programmable filter function has been confirmed by operating the devices as Gaussian weight filter and band pass filter.
Abstract: A novel method for constructing charge-coupled device (CCD) transversal filters is presented. In the filter, transversal filtering action is implemented by introducing weighted input signals into each CCD register stage. The aim of such an input weighted CCD transversal filter was to eliminate some of the disadvantages in conventional CCD filters, i.e., the split-electrode filters or tapped delay line filters. The split electrode filter is non-programmable and requires a complex sensing circuit. The tapped delay line filter has a low transfer efficiency and requires an external summing circuit. A 64-stage programmable filter device has been developed. The device contains dual gate FET's for input-weighting and a 4-phase surface channel CCD register for time delay and summing. The filter output is obtained from usual floating diffusion amplifier connected to the final register stage. Fundamental delay line characteristics were measured at 10.7 MHz clock frequency. A 3 dB bandwidth of more than 3 MHz and a differential gain of less than 5 percent were achieved. Using the devices, a programmable filter function has been confirmed by operating the devices as Gaussian weight filter and band pass filter.

Journal ArticleDOI
TL;DR: A design procedure for cascades of grating resonators to construct Tchebysheff, Butterworth, Gaussian, Legendre and other passband filters is given in this paper.
Abstract: A design procedure is given for cascades of grating resonators to construct Tchebysheff, Butterworth, Gaussian, Legendre and other passband filters.

Patent
20 Dec 1977
TL;DR: In this article, a transversal filter with a single passband was proposed for frequency-converting filtering with an abrupt transition from stopband to passband, no involved oscillator circuit, no mixing circuit, and is capable of being integrated on a single semiconductor chip.
Abstract: OF THE DISCLOSURE Disclosed is a frequency converting filter comprised of a charge transfer device transversal filter and a non-sampling filter. The transversal filter has an output lead coupled to an input of the non-sampling filter, and has frequency passbands of width .DELTA.f centered at the fixed frequencies nfs ? fo where n is an integer multiple of the sampling frequency, fs is the sampling frequency, and fo is the center frequency removed from the sampling frequency of the desired frequency passband. The non-sampling filter has a single passband. The single passband has a width of less than fs - fo - .DELTA.f/2 and it is centered to include only a selected one of the passbands of the transversal filter. The invention provides an improved device for filtering having an abrupt transition from stopband to passband, no involved oscillator circuit, no mixing circuit, and is capable of being integrated on a single semiconductor chip.

Journal ArticleDOI
01 Jan 1977
TL;DR: In this paper, a low-pass power filter using a D.c. machine as a reactive element, replacing the conventional electrostatic capacitors, is proposed, and the critical parameters in the compensation circuit necessary to overcome the effects of these factors are derived together with the required VA ratings of the machine and inductance in a typical L-type filter.
Abstract: A novel low-pass power filter is proposed, in which a d.c. machine is used as a reactive element, replacing the conventional electrostatic capacitor. The equivalent capacitance of a suitably operated d.c. machine is, typically, several times larger than commercially obtainable electrostatic capacitors, and, therefore, a filter using a machine as one of its elements should have exceptional characteristics. The primitive machine filter, i.e. one in which the machine directly replaces the capacitance, is shown to be unsatisfactory because of the rotor's leakage inductance and resistance. The critical parameters in the compensation circuit necessary to overcome the effects of these factors are derived together with the required VA ratings of the machine and inductance in a typical L-type filter. Based on the derived principles, a machine filter, using a standard solid-iron-frame d.c. machine, has been assembled and tested, and the results show an output-voltage waveform which contains less than 0.333% total distortion. A theoretical study of the filter is presented using the proposed equivalent circuit. Finally, the machine filter is compared with classical passive-element systems under steady-state and transient conditions, and possible improvements are suggested. It is concluded that the machine filter is somewhat better than the classical filter in terms of cost and size, and that the noise and vibration problem may be offset in a number of applications by the machine filter's superior transient performance and extended temperature range.

01 Jun 1977
TL;DR: Expressions for a generalized Modified Transitional Butterworth-Chebyshev (MTBC) filter are derived and it is shown that by adjusting location and order of the inserted zeros, the cut-off slope rate can be traded for maximum attenuation in the stop band.
Abstract: : Expressions for a generalized Modified Transitional Butterworth-Chebyshev (MTBC) filter are derived. The characteristics of this filter as applied to digital filter design are investigated. It is shown that by adjusting location and order of the inserted zeros, the cut-off slope rate of the filter can be traded for maximum attenuation in the stop band. The performance of this MTBC filter is compared to that of Butterworth, Chebyshev, transitional Butterworth-Chebyshev filters together with those suggested by other investigators. It is shown that the stop-band attenuation can be significantly increased without great sacrifice of cut-off slope rate. Step response of this MTBC filter is also obtained and compared with other filters. Various tabulations as well as graphs of this filter are given for design purposes. A computer program is developed for the design of this filter. (Author)

Journal ArticleDOI
01 Jan 1977-Frequenz
TL;DR: In this article, a block diagram of the universal active-active-Η filter is shown and analyzed; the corresponding circuit is given; and a design procedure and some curves are given for practical purposes.
Abstract: Different novel biquadratic active-Η filters are proposed; all of them use the first pole model of op. amp. instead of external capacitors. At first, the general block diagram of the universal filter is shown and analyzed; the corresponding circuit is given. As special cases L. P., B. P., H. P., notch, and allpass filters are derived from the proposed configuration; the simplest one of them is a H. P. — B. P. one which consists of two op. amps, and two resistors only. A design procedure and some curves are given for practical purposes. This universal filter has the advantage that it is of low sensitivities usable for high frequencies, and tunable by varying the resistors ratios. Verschiedene neue biquadratische aktivc-R Filter werden vorgeschlagen, die anstelle externer Kapazit ten den durch die inneren Kapazit ten der Operationsverst rker verursachten Frequenzgang der Verst rkung ausnutzen. Zuerst wird das Blockdiagramm des Universalfilters analysiert sowie eine Realisierung angegeben. Als Sonderf lle der universellen Schaltung werden Filter mit Tief-, Bandund Hochpa sowie Bandsperrencharakter angegeben. Die einfachste Anordnung besteht aus nur zwei Widerst nden und zwei Operationsverst rkern und verwirklicht eine Hochund Bandpa bertragungsfunktion gleichzeitig. Ein Entwurfsverfahren mit zugeh rigen Kurven wird vorgeschlagen. Das neue Universalfilter zeigt niedrige Empfindlichkeiten und ist auch f r hohe Frequenzen verwendbar. Das Netzwerk kann durch Variation von Widerstandsverh ltnissen abgestimmt werden. F r die Dokumentation: Aktive-R Filter / Operationsverst rker / Filtertheorie

Journal ArticleDOI
TL;DR: In this paper, a simple high-Q insensitive active band-pass filter with low closed-loop gains was proposed, which is capable of operating over a wide frequency range with relatively stable Q and resonant frequency.
Abstract: The paper describes a simple high-Q insensitive active band-pass filter having relatively few passive components and active elements with low closed-loop gains, using positive feedback Q enhancement scheme. The proposed filter has minimum dependence on the bandwidth of the operational amplifiers and is capable of operating over a wide frequency range with relatively stable Q and resonant frequency.

Journal ArticleDOI
TL;DR: In this paper, a multiple-notch filter is proposed to track and remove a periodic interfering signal and its harmonics, and it is shown that, to first order, the filter notches are not dependent on incomplete charge transfer.
Abstract: An application of c.c.d. is described to realise a multiple-notch filter which can be made to track and remove a periodic interfering signal and its harmonics. It is shown that, to first order, the filter notches are not dependent on incomplete charge transfer.


Journal ArticleDOI
TL;DR: In this article, epsilon quadrature is extended to a numerical transform associated with tuned integration, which is shown to improve the performance of a digital Butterworth filter and the accuracy of a new numerical transform.

Journal ArticleDOI
TL;DR: In the above paper, Brodie presents a biquad low-pass filter which requires only two operational amplifiers, and it is shown that if one allows one additional resistor for biasing, a true bandpass output is available.
Abstract: In the above paper, Brodie (see ibid., vol.SC-11, p.552-5, 1976) presents a biquad low-pass filter which requires only two operational amplifiers. The use of current differencing amplifiers for reduced operational amplifier count implementation of the biquad filter has been around for some time. However, Brodie does not mention that the low-pass form is not the only filter that can be achieved. If one allows one additional resistor for biasing, a true bandpass output is available.

ReportDOI
01 Aug 1977
TL;DR: In this article, P Log P maximum entropy analysis has been applied to a bank of Butterworth filters and resolution enhancement was obtained with the prewhitening autocorrelation method.
Abstract: P Log P maximum entropy analysis has been applied to a bank of Butterworth filters. Resolution enhancement was obtained. Comparison is done with the prewhitening autocorrelation method.