JP2008256452A - Pulse signal generator - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a pulse signal generator which uses two digital synthesizers, and moreover can apply amplitude control to an envelope in a free shape including an envelope having smooth rises and falls, without requiring a multiplier. <P>SOLUTION: A sum phase and a difference phase to be acquired by adding an envelope phase to be expressed by the inverse cosine function of a normalized envelope function normalized so that the maximum value of a predetermined envelope function becomes 1, to the signal phase of a predetermined frequency signal not being amplitude-controlled, and by subtracting the envelope phase from the signal phase, are formed. Referring to a waveform table, the sum phase is converted into a sum phase waveform signal and the difference phase is converted into a difference phase waveform signal, and those are composited to obtain a pulse signal. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a pulse signal generator that generates a pulse signal with controlled amplitude such as a transmission pulse used in a pulse radar device.

  In a pulse radar device, a transmitter generates a transmission pulse signal in which a high-frequency transmission frequency signal is modulated in a pulse shape with a short time width, and transmits the transmission pulse signal to the outside via an antenna. The reflected signal from the antenna is received by the antenna and the distance, direction, moving speed, etc. of the target are measured.

  A digital synthesizer is used as a method of creating a pulse signal such as this transmission pulse signal. Since the digital synthesizer uses a method of generating a high frequency signal waveform from a digitally generated phase signal, the phase of the signal waveform can be easily controlled, and is used in a signal processing system that requires phase information.

  The digital synthesizer has a mechanism for generating a signal by converting an instantaneous phase into a waveform function as shown in FIG. 13 showing a basic configuration (Non-patent Document 1). In FIG. 13, the phase increment is input to the adder 11 in digital form at a predetermined period. In synchronization with the input, the value of the phase register is updated by adding the phase increment to the value of the phase register 12, and the instantaneous phase at that time is obtained. Using the instantaneous phase value, the instantaneous value of the waveform function is obtained from the waveform table 14 stored in the memory, and the waveform function is output from the waveform function converter 15.

  The waveform table 14 is a memory area for obtaining an instantaneous value of a waveform function from a given phase, and normally stores a value of a sine function. In the waveform generation by the waveform table 14 and the waveform function converter 15, the value of the phase register 12, that is, the instantaneous phase is given as an address of the waveform table (ROM or the like) 14, and a sine wave is obtained by reading the value from the address. Of course, as in the case of normal waveform generation, when the phase of the output sine wave makes a round, the address specified by the value of the phase register 12 is reset.

  The digital waveform function value obtained by referring to the waveform table 14 is converted into an analog signal by the DA converter 17 based on the reference voltage. Since the converted analog signal is a stepped signal as shown in FIG. 13, it is smoothed through a low-pass filter 18 to obtain an output waveform function (output signal) of a desired analog waveform. By generating this output signal continuously for a predetermined period T, a pulse signal having a pulse width T is generated.

  The input phase increment value is an amount that defines the frequency of the generated pulse signal. For example, in the signal example of FIG. 13, since the phase increment is set to a constant value, the output signal is a sine wave having a constant frequency. On the other hand, if the phase increment is input so as to increase linearly or decrease linearly with respect to time, a linear frequency modulation wave used in an FM-CW radar or the like can be obtained.

  In addition, a phase modulation wave can be generated using a digital synthesizer. In this case, an adder is provided on the output side of the phase register 12 in FIG. 13, and the digital modulation signal is added to the instantaneous phase of the phase register 12 to make the output of the adder a new instantaneous phase. . As this digital modulation signal, for example, a binary phase signal, a quaternary signal, or the like is used to switch the instantaneous phase discontinuously, thereby realizing digital phase modulation.

As described above, since the basic digital synthesizer generates a waveform from the instantaneous phase, the signal frequency can be freely controlled, but the signal amplitude of the output signal is always constant. However, the output signal of the digital synthesizer can be amplitude-modulated by controlling the reference voltage of the DA converter 17 as shown in FIG. FIG. 14 shows the addition of the reference voltage Vref of the DA converter 17 of FIG. 13 which is composed of an inverting circuit composed of an operational amplifier OP1 and a resistor, an operational amplifier OP2 and a resistor, and adds the modulation signal Vm to the inverted reference voltage. A modulation reference voltage (Vref−Vm) is obtained by superimposing a modulation signal Vm input in the form of an analog signal on a reference voltage Vref by a circuit. Since the modulation reference voltage is a voltage that controls the conversion scale of the DA converter, even if the same waveform function value is input, the voltage value of the output signal changes in proportion to the modulation reference voltage. . That is, the output signal of the digital synthesizer is subjected to amplitude modulation by the input analog signal Vm.
EDN Japan “Using DDS to realize a high-precision, high-purity sine wave oscillator”, [online], Reed Electronics Group, “February 27, 2007 search”, Internet <URL: http: // www. ednjapan. com / content / issue / 2005/09 / content03. html>

  Since the phase of the output signal generated by the digital synthesizer can be controlled, application examples that handle the phase signal can be implemented relatively easily. Even in the pulse radar device, the phase information in the pulse signal is used, so that the signal-to-noise ratio is improved by pulse compression and integration processing using the phase signal, and a farther target can be detected. is there.

  On the other hand, a pulse signal that rises quickly tends to cause spurious because the side lobe attenuation is small when viewed in the frequency domain. In addition, since a fast rise is distorted by a power amplifier or the like, a range side lobe appears because pulse compression radar tends to cause a correlation shift at the rise and fall of the pulse. Since the transmission pulse width of radar equipment is shorter than a few microseconds, in order to solve the problems related to spurious and range side lobes, the signal generated by a digital synthesizer can be stably generated within a short time, for example within 1 microsecond. Amplitude modulation is required.

  With the circuit configuration of FIG. 14, it is possible to obtain an amplitude-modulated output signal from a digital synthesizer. However, when a narrow pulse signal of several microseconds or less handled by a pulse radar device or the like is obtained, in the amplitude control for the output signal using the circuit configuration of FIG. 14, the reference voltage is set for a very short time (for example, 1 microsecond). )) To make a large change (from high level to low level, or from low level to high level), a relatively large delay or distortion occurs in the DA converter, etc. It was difficult to get.

  In order to cope with this problem, the present inventor uses a plurality of digital synthesizers, generates signals having frequencies separated from each synthesizer by a predetermined frequency interval, and performs pulse amplitude control by synthesizing these signals. A pulse signal generator has been proposed (Japanese Patent Application No. 2007-062063; hereinafter referred to as the prior invention). According to the prior invention, the amplitude control of a short pulse is possible. For example, assuming that the angular frequency of the signal is ω and the transmission time (pulse width) of the pulse is T, signals having angular frequencies ω−1 / 2T and ω + 1 / 2T are generated and synthesized with the same amplitude, as shown in FIG. A signal with an envelope of sin (πt / T) is obtained, and discontinuities can be removed from the signals at both ends of the pulse. However, the envelope shapes that can be generated by only two synthesizers are limited in this way, so that the signals at both ends of the pulse are smooth (a shape that does not include a discontinuity in the first derivative) like a Hanning window. Requires three or more synthesizers.

  Therefore, the present invention uses two digital synthesizers in a pulse signal generator that synthesizes the outputs of a plurality of digital synthesizers, and includes an envelope that rises and falls smoothly without requiring a multiplier. An object of the present invention is to provide a pulse signal generator capable of controlling the amplitude of a free-form envelope.

The pulse signal generator according to claim 1, wherein the pulse signal generator generates a pulse signal having a predetermined envelope shape with a predetermined frequency fo and a predetermined time width T.
The inverse cosine function (cos ⁻¹ function or arc cosine) of the normalized envelope function normalized so that the maximum value of the predetermined envelope function becomes 1 with respect to the signal phase of the predetermined frequency signal that is not amplitude-controlled. A sum / difference phase generation means for generating a sum phase obtained by adding the envelope phases expressed by (function) and a difference phase obtained by subtracting the envelope phase from the signal phase;
Sum phase waveform function conversion means for converting the sum phase into a sum phase waveform function value (sum phase waveform signal) by referring to a waveform table prepared in advance in a memory;
A differential phase waveform function conversion means for converting the differential phase into an output frequency waveform function value (differential phase waveform signal) by referring to the waveform table or another waveform table equivalent to the waveform table;
And a synthesis means for synthesizing the sum phase waveform function value and the difference phase waveform function value to output the pulse signal.

The pulse signal generator according to claim 2 is the pulse signal generator according to claim 1,
When the envelope phase increment of the envelope phase is represented by an Nth order polynomial (where N is 0 and a positive integer),
The N-th order envelope phase register means to the first order envelope phase register means each having an initial phase increment value and accumulatively adding and outputting the input value to the initial phase increment value in synchronization with the input timing. Connected in series,
An envelope initial phase register means having an initial phase value of the envelope and accumulating and adding an input value from the primary envelope phase register means in synchronization with an input timing is output. Connected to the output side of the means,
A predetermined constant envelope N + 1 order phase increment is inputted to the N-th order envelope phase register means in synchronization with an input timing, and an output of the envelope initial phase register means is obtained as the envelope phase. Features.

  According to the pulse signal generator of the present invention, a relatively free-form envelope can be obtained simply by using two sets of digital synthesizers (sum / difference phase generation means, sum phase waveform function conversion means, difference phase waveform function conversion means). A pulse signal whose amplitude is controlled to a line can be generated. Since the phase of the pulse signal is not converted to an unintended phase by amplitude control, and the signal phase can be controlled in the same manner as an unmodulated pulse signal, the pulse signal according to the present invention is a signal processing system using phase information. Can be used. The amplitude control of the pulse signal of the present invention is useful in a pulse compression radar, and the range side lobe of the compressed pulse can be suppressed by transmitting the amplitude-controlled pulse signal.

  In addition, the pulse signal generator of the present invention can suppress waveform distortion that occurs during signal amplification and filtering by controlling the amplitude to an envelope shape that smooths both ends of a pulse such as a Hanning window, which is unnecessary. Signals with small frequency components (spurious) can be realized.

  Furthermore, even when the envelope phase is a high-order polynomial with respect to time, the envelope phase can be given only by an adder using a register arranged in series in multiple stages, so that the present invention can be used without using any multiplier. Since the amplitude can be controlled by implementing, the circuit scale does not increase.

  Embodiments of a pulse signal generator according to the present invention will be described below with reference to the drawings.

  The pulse signal generator of the present invention uses only two sets of digital synthesizers (sum / difference phase generation means, sum phase waveform function conversion means, difference phase waveform function conversion means). Amplitude modulation is realized so that both ends of a pulse signal requiring a synthesizer are smooth. Furthermore, the present invention can control the amplitude without using any multiplier. Therefore, in the invention of the prior application, the angular frequency (envelope phase increment) Ω (t) of the envelope is constant, but in the present invention, the angular frequency (envelope phase increment) Ω (t) of the envelope is changed with time. By changing, the amplitude control of the pulse is realized by only two sets of digital synthesizers.

  First, a case where a pulse signal having an envelope as shown in FIG. 1 and an envelope phase increment for that purpose (defined by a first order polynomial in FIG. 1) is generated will be described as an example.

  Assume that the original signal to be transmitted is cos θ (t), and the amplitude of the original signal cos θ (t) is controlled by an envelope of cos Θ (t). Here, θ (t) is the instantaneous phase of the original signal cos θ (t) to be transmitted at time t, and Θ (t) is the instantaneous phase of the envelope cos Θ (t). At this time, the amplitude-controlled pulse signal is expressed by the following Equation 1.

  As can be seen from Equation 1, the cosine function cos [θ (t) + Θ (t)] of the sum of the instantaneous phase θ (t) and the instantaneous phase Θ (t) and the instantaneous phase θ (t) and the instantaneous phase Θ ( By synthesizing the cosine function cos [θ (t) −Θ (t)] of the difference from t), an amplitude-controlled pulse signal can be obtained.

Here, if the instantaneous angular frequencies of the original signal cosθ (t) and the envelope cosΘ (t) are ω (t) and Ω (t), respectively, the instantaneous phases θ (t) and Θ (t) are It can be expressed by Equation 2. However, θ ₀ and Θ ₀ are the initial phases of the original signal and the envelope, respectively.

  In the present invention, the amplitude control of the pulse is realized by only two sets of digital synthesizers by changing the angular frequency (envelope phase increment) Ω (t) of the envelope with time. Taking FIG. 1 as an example, the pulse transmission period is between time 0 and T, and the angular frequency (envelope phase increment) Ω (t) of the envelope is increased linearly until time T / 2. Consider a case where the angular frequency Ω (t) of the envelope is linearly decreased. Here, the angular frequency of the envelope at time 0 and T is set to zero, and Θ (0) = − π / 2, Θ (T / 2) = 0, and Θ (T) = π / 2. If the above conditions are imposed, the angular frequency and instantaneous phase of the envelope can be expressed by Equation 3.

  FIG. 2 schematically shows the generation of such an envelope shape. FIG. 2A shows the shape of an envelope when the phase is taken on the horizontal axis, that is, a sinusoidal shape. Here, by changing the angular frequency Ω (t) of the envelope as time t passes, the rotational speed of the phase at both ends is reduced. Therefore, the envelope shape when the time t is taken on the horizontal axis is similar to the Hanning window function as shown in FIG. Amplitude control that smoothes the rise and fall of the pulse signal requires three or more synthesizers in the prior invention, but in the present invention, by making the angular frequency of the envelope variable. Amplitude control similar to the Hanning window is possible using only two synthesizers.

  Similarly, when the angular frequency Ω (t) of the envelope is a quadratic function of time, the amplitude can be controlled so that both ends of the pulse are smooth. For example, when the angular frequency Ω (t) of the envelope has a parabolic shape as shown in FIG. 3, the phase Θ (0) = − π / 2, Θ (T / 2) = 0, Θ (T ) = Π / 2, and the coefficient is obtained under the condition that satisfies the condition, the angular frequency and phase of the envelope are determined as shown in Equation 4.

  In the case of FIG. 3, although the envelope tail is narrower than before, the frequency of the envelope at the tail is zero, so that amplitude control can be realized so that the rise and fall of the pulse are smooth. .

  To make the rise and fall of the pulse smooth, select the envelope phase so that the envelope phase at both ends of the pulse is -π / 2 or π / 2 and the frequency at both ends of the pulse is zero. That's fine. In order to effectively use the dynamic range of the synthesizer, it is desirable to select an envelope phase such that the envelope phase becomes zero or a value in the vicinity thereof within the transmission period.

When this envelope phase is generally expressed, the inverse cosine function (cos ⁻¹ function or arc cosine function) of the normalized envelope function normalized so that the maximum value of the predetermined envelope function becomes 1; It can be. The general expression of the envelope phase is the same for the other embodiments of the present invention.

  The original signal of a predetermined frequency is a signal whose amplitude is not controlled, and the predetermined frequency of the original signal may of course be a constant frequency that is not modulated, and is subjected to frequency modulation such as linear frequency modulation. The same applies to those that are present.

  To implement the present invention, two sets of synthesizers (sum / difference phase generation means, sum phase waveform function conversion means, difference phase waveform function conversion means) and a synthesis circuit for synthesizing the output signals of the synthesizers are provided. Thus, it suffices to have phase generation means for giving them a signal phase and an envelope phase.

  FIG. 4 is a diagram illustrating an example of digitally synthesizing the sum phase waveform signal and the difference phase waveform signal.

  At a predetermined period, the signal phase increment is input to the adder 11-1 in digital form. In synchronization with the input of the signal phase increment, the value of the signal phase register 12-1 is updated by adding the signal phase increment to the initial value of the signal stored in the signal phase register 12-1, and at that time Signal phase θ (t). This signal phase θ (t) may be a signal phase of a predetermined frequency signal whose amplitude is not controlled. The adder 11-1 and the signal phase register 12-1 constitute a signal phase register means.

Also, in the same cycle, the envelope phase increment is input to the adder 11-2 in digital form. In synchronization with the input of the envelope phase increment, the value of the envelope phase register 12-2 is updated by adding the envelope phase increment to the initial value stored in the envelope phase register 12-2, and The envelope phase Θ (t) at the time is obtained. This envelope phase Θ (t) is expressed by the inverse cosine function (cos ⁻¹ function or arc cosine function) of the normalized envelope function normalized so that the maximum value of the predetermined envelope function becomes 1. The envelope phase to be used may be used. The adder 11-2 and the envelope phase register 12-2 constitute a comprehensive line phase register means (the same applies hereinafter).

  The envelope phase Θ (t) is added to the signal phase θ (t) by the adder 13-1 to generate a sum phase, and the envelope phase Θ (t) is calculated from the signal phase θ (t). Then, the difference phase is generated by subtraction in the adder 13-2. The generation means by the butterfly calculator using the adders 13-1 and 13-2 that generates the sum phase and the difference phase is the sum / difference phase generation means.

  By referring to the waveform table 14 prepared in advance in the memory, the sum phase is converted into a sum phase waveform function value (sum phase waveform signal) by the sum phase waveform function converting means 15-1. Similarly, by referring to the waveform table 14 (or another waveform table equivalent to the waveform table 14), the difference phase is converted into the waveform function value for the output frequency (difference phase waveform signal) by the difference phase waveform function conversion means 15-2. ). As described above, the present invention has an advantage that the sum phase waveform signal and the difference phase waveform signal to be synthesized are not weighted unlike the prior invention, and therefore a multiplier is unnecessary.

  The sum phase waveform function value and the difference phase waveform function value are synthesized by an adder 16, converted to an analog signal by a DA converter 17, and smoothed by a low-pass filter (LPF) 18. Output a signal.

  FIG. 5 is a diagram illustrating an example of synthesizing the synthesis of the sum phase waveform signal and the difference phase waveform signal. The process is the same as in FIG. 4 until the sum phase waveform signal and the difference phase waveform signal are obtained. The sum phase waveform signal is converted to an analog signal by the DA converter 17-1 and smoothed by the LPF 18-1, the difference phase waveform signal is converted to an analog signal by the DA converter 17-2, and smoothed by the LPF 18-2. The smoothed sum / difference phase waveform signal is synthesized by an analog adder circuit including an operational amplifier 19 and resistors R, R, and R0, and a pulse signal is output.

  In FIG. 5, the resistor through which the sum phase analog waveform signal and the difference phase analog signal pass immediately after the low-pass filters 18-1 and 18-2 is based on the fact that no multiplier is required for signal synthesis of the present invention. The same resistor R is used, but a variable resistor may be used in order to adjust the amplitude of the analog voltage.

  When the present invention is implemented using commercially available digital synthesizers 1 and 2, the configuration shown in FIG. In the configuration of FIG. 6, the signal phase register 12-1 and the envelope phase register 12-2 are behind the butterfly calculators 13-1 and 13-2, and therefore the order is reversed compared to FIG. 5. . Since the update processing of the phase registers 12-1 and 12-2 and the butterfly computations 13-1 and 13-2 are both linear processing, the butterfly computing unit shown in FIG. The same processing result as in FIG. 5 can be obtained. In addition, the present invention can be implemented by inputting the result of the butterfly operation to the phase increment input of a commercially available digital synthesizer by changing the order.

  As described above, in the case of analog synthesis, the quality of amplitude control is not as good as that of digital synthesis. However, since the present invention can be implemented by combining a commercially available digital synthesizer, it can be said to be an embodiment for an inexpensive system.

  Next, the relationship between the desired envelope shape and the envelope phase increment input therefor will be described.

  Taking FIG. 1 as an example, Ω (t) in the figure is the angular frequency of the envelope, and corresponds to the envelope phase increment input to the adder 11-2 in FIG. Θ (0) is the initial phase of the envelope, and corresponds to the initial value of the envelope phase register 12-2. FIG. 1 shows an example in which the phase increment (angular frequency) of the envelope is realized by a first-order polynomial in order to obtain an envelope in which both ends of the pulse signal are smooth.

In FIG. 1, the pulse transmission period is between time 0 and T, and the angular frequency (envelope phase increment) Ω (t) of the envelope is increased to time T / 2, Ω (t) = 4π (t / T ² ) and linearly increased from time T / 2 to T and decreased linearly at Ω (t) = 4π ((T−t) / T ² ). Ω (0) = Ω (T) = 0, Ω (T / 2) = 2π / T. The phase of the envelope may be on the condition that Θ (0) = − π / 2 and Θ (T / 2) = 0. Thereby, a pulse signal having a smooth envelope as shown in FIG. 1 is obtained.

  FIG. 7 to FIG. 10 are diagrams showing the relationship between other desired envelope shapes and the envelope phase increments input therefor.

FIG. 7 shows an example in which the phase increment (angular frequency) of the envelope is realized by a second-order polynomial. In FIG. 7, the angular frequency (envelope phase increment) Ω (t) of the envelope between times 0 and T, which is the pulse transmission period, is expressed as Ω (t) = 6π (t (T−t) / T ³ ) is continuously changed. Ω (0) = Ω (T) = 0. The phase of the envelope may be on the condition that Θ (0) = − π / 2, Θ (T / 2) = 0, and Θ (T) = π / 2.

FIG. 8 shows an example in which the phase increment (angular frequency) of the envelope is realized by a third-order polynomial. In FIG. 8, the angular frequency (envelope phase increment) Ω (t) of the envelope between times 0 and T, which is the pulse transmission period, is expressed as Ω (t) = 32π (t (t−T / 2 ) (T−T) / T ⁴ ). Ω (0) = Ω (T) = 0. The phase of the envelope may be on the condition that Θ (0) = − π / 2, Θ (T / 2) = 0, and Θ (T) = π / 2.

  FIG. 1, FIG. 7 and FIG. 8 are examples in which the phase increment (angular frequency) of the envelope is realized by a first-order to third-order polynomial in order to obtain an envelope in which both ends of the pulse signal are smooth.

FIG. 9 is a diagram showing a relationship between a triangular waveform envelope shape and an envelope phase increment input therefor. In FIG. 9, the pulse transmission period is between time 0 and T, and the angular frequency (phase increment) Ω (t) of the envelope is up to time T / 2, Ω (t) = − (2 / T) [ 1 + 2 (t / T) ² ] is reduced by a quadratic curve, and from time T / 2 to T, quadratic at Ω (t) = (2 / T) [1 + 2 ((T−t) / T) ² ] The curve is decreasing. The phase of the envelope may be on the condition that Θ (0) = π / 2, Θ (T / 2) = 0, and Θ (T) = π / 2. As a result, a pulse signal having a triangular wave envelope as shown in FIG. 9 is obtained.

This shape is not a strict shape but an approximate shape. In order to obtain a precise shape, Equation 5 must be used. As an example that can be easily realized, a linear approximation of 1 / (1-x) ^1/2 ≈1 + x / 2 was used. If higher accuracy is required, a quadratic approximation or the like may be used.

  In FIG. 10, the amplitude of the entire period T of the pulse signal is not controlled, but the pulse signal rises (increases) at the rise time Tr, and the pulse signal falls (decays) at the fall time Td. Is realized by a constant phase increment. In FIG. 10, the pulse transmission period is between time 0 and T, the angular frequency (phase increment) Ω (t) of the envelope is set to a constant value of Ω (t) = π / 2Tr until time Tr, From T-Td to T, Ω (t) = π / 2Tr is a constant value, and T-Td is 0 from time Tr during that time. The phase of the envelope may be on the condition that Θ (0) = π / 2, Θ (T / 2) = 0, and Θ (T) = π / 2. As a result, the amplitude changes sinusoidally during the rising and falling periods as shown in FIG. 9, and the amplitude is substantially constant with respect to time near the maximum value of the amplitude. A pulse signal whose envelope is close to a trapezoid can be realized.

  In FIGS. 1 and 7 to 10, if the phase increment of the envelope is correctly input, it is possible to obtain the respective envelopes of the scheduled pulse signal. However, when the angular frequency (phase increment) of the envelope is a first-order or higher-order polynomial with respect to time, a multiplier is required to calculate this according to the formula, and an amplitude modulation method without using a multiplier is required. As a result, the advantages of the present invention will be diminished.

  Therefore, when the envelope phase increment of the envelope phase is expressed by an Nth order polynomial (where N is 0 and a positive integer), each has an initial phase increment value and inputs an input value to this initial phase increment value. Nth-order envelope phase register means to primary envelope phase register means, which are cumulatively added in synchronization with the timing and output, are connected in series in the order of the order. An envelope initial phase register means having an initial phase of the envelope and accumulating and outputting the input value from the primary envelope phase register means in synchronization with the input timing is output from the primary envelope register means. Connect to the output side. Then, a predetermined constant envelope N + 1-order phase increment is input to the N-th order envelope phase register means in synchronization with the input timing, and the output of the envelope initial phase register means is obtained as the envelope phase. .

  By arranging the phase register means in series, it is possible to calculate the required phase increment only by the adder even if the phase increment is a high order polynomial. The multiplier in the digital circuit is almost the same as the scale in which adders are arranged by the number of bits necessary for the operation. Therefore, the effect of eliminating the multiplier by arranging the phase register means in series is great.

FIG. 11 is a diagram illustrating an example of an input configuration of an envelope phase increment that becomes the second order polynomial (ie, N = 2) in FIG. A secondary envelope phase register means comprising an adder 11-4 and a register 12-4; a primary envelope phase register means comprising an adder 11-3 and a register 12-3; and an adder 11-2 Envelope initial phase register means comprising a register 12-2 are connected in series. Then, a predetermined constant number of envelope tertiary phase increments are input to the secondary envelope phase register means in synchronization with the input timing, and the output of the envelope initial phase register means is obtained as the envelope phase. In this example, the envelope third-order phase increment is −12π (Δt / T) ³ , and the initial value of the second-order envelope phase register means is 6π (Δt / T) ² (1-2Δt / T). Yes, the initial value of the primary envelope phase register means is π (Δt / T) ² (3−2Δt / T), and the initial value of the envelope initial phase register means is −π / 2. Here, the transmission time of the pulse signal is T, and the clock interval Δt for phase increment input is assumed.

  Here, the initial value to be set in the register is obtained as follows. Since the phase of the envelope at time t is a value obtained by integrating Ω (t), it can be obtained from Equation 6.

Assuming that the clock generation time for the nth data input after the start of transmission is t _n ≡n · Δt and the envelope phase at that time is Θ _n , it is expressed by Equation 7.

This is a value to be stored in the phase register of each order at time tn. Since the value of the primary envelope phase register means is defined as ΔΘ _n ≡Θ _{n + 1} −Θ _n , Equation 8 is obtained.

Further, the value of the envelope phase secondary envelope phase register means of the envelope phase is defined as Δ ² Θ _n ≡ΔΘ _{n + 1} −ΔΘ _n , and thus is given by Equation 9.

Further, since the envelope third-order phase increment is defined as Δ ³ Θ _n ≡Δ ² Θ _{n + 1} −Δ ² Θ _n , it settles to a constant expressed by Equation 10.

  Therefore, if the third-order phase increment is input as a constant, the correct envelope phase increment and the envelope phase can be given only by the adder. The initial value given to each register can be obtained by substituting n = 0 into each equation described above.

  The concept of FIG. 11 can be applied to, for example, FIG. 8 when the envelope phase increment of the envelope phase is expressed by an Nth order polynomial.

However, since the envelope phase increments given in FIGS. 1, 9, and 10 have inflection points and discontinuities, they must be switched during the transmission time. As an implementation example in such a case, FIG. 12 shows a circuit for providing the envelope phase increment of FIG. The initial value of the register is shown in FIG. 12, using the procedure described in FIG. 10, at the time on the discrete time axis of the envelope phase theta _n differentiating _{(ΔΘ n ≡Θ n + 1 -Θ} n) Determined by seeking. Since the envelope phase increment shown in FIG. 1 tends to increase immediately after the start of transmission, the switch 12-5 in the figure is on the A side (constant input; 4π (Δt / T) ² ). However, since the envelope phase increment tends to decrease when the inflection point is reached at time T / 2, the switch 12-5 is switched to the B side (constant input; −4π (Δt / T) ² ). The switch 12-5 can be realized by a selector circuit using a logic circuit. At the moment when the switch 12-5 is switched to the B side, 2π · Δt / T is stored in the envelope phase increment register 12-3, but in order to avoid an error at the end of transmission, the switch 12-5 The register 12-3 is preferably changed to 2π [(Δt / T) − (Δt / T) ² ] at the moment of switching −5.

  Generally expressing this, when inputting a predetermined constant number of envelope N + 1-order phase increments to the N-th order envelope phase register means in synchronism with the input timing, the inflection point or inflection point of the envelope phase increment. At a continuous point, it is necessary to switch the envelope constant N + 1-order phase increment described above to the envelope constant N + 1-order phase increment of another predetermined constant. Further, the other envelopes shown in FIGS. 9 and 10 can be designed in a substantially similar procedure.

  The original signal of a predetermined frequency is a signal whose amplitude is not controlled, and the frequency of the original signal is subjected to frequency modulation such as linear frequency modulation (linear chirp modulation) in addition to a constant frequency that is not modulated. It may be what you have.

  When a linear frequency-modulated original signal is used, the signal phase with respect to time is a quadratic function. For example, when FIG. 1 is taken as an example, it is constituted by an adder 11-1 and a signal phase register 12-1. It is preferable that the signal phase register means composed of an adder and a register is connected in series before the signal phase register means, and a predetermined phase increment is input to the signal phase register means as a constant.

Example of amplitude control by inputting first order phase increment Example of instantaneous frequency of envelope for the envelope shape of FIG. Example when the instantaneous frequency of the envelope is a quadratic function of time Example in which the present invention is realized by digital synthesis Example of realizing the present invention by analog synthesis Example of the present invention for analog synthesis using a commercially available digital synthesizer Example of amplitude control by inputting second order phase increment Example of amplitude control by inputting third-order phase increment Example of phase increment input to obtain an approximately triangular wave envelope. Example of phase increment input to obtain approximately trapezoidal envelope Input example of envelope phase increment to be the second order polynomial in FIG. Example of phase increment input realizing the envelope of FIG. Example of basic configuration of digital synthesizer Example of amplitude modulation of the output of a conventional digital synthesizer Example of generating an envelope of a cosine function shape in the invention of the prior application

Explanation of symbols

1,... Synthesizer 3.. Synthesizer (adder circuit), 11-1 to 11-4, adder, 12-1 to 12-4, phase register, 13-1, 13-2,. Adder,
14, 14-1, 14-2 ... Waveform table, 15-1, 15-2 ... Waveform function converter, 16. Adder, 17, 17-1, 17-2 ... DA converter,
18, 18-1, 18-2 ··· LPF, 19 · · operational amplifier, 12-5 ··· switch

Claims (2)

In a pulse signal generator for generating a pulse signal having a predetermined envelope shape with a predetermined frequency and a predetermined time width,
The envelope phase expressed by the inverse cosine function of the normalized envelope function normalized so that the maximum value of the predetermined envelope function becomes 1 is added to the signal phase of the predetermined frequency signal not subjected to amplitude control. Sum / difference phase generating means for generating a sum phase and a difference phase obtained by subtracting the envelope phase from the signal phase;
Sum phase waveform function conversion means for converting the sum phase into a sum phase waveform function value by referring to a waveform table prepared in advance in a memory;
A differential phase waveform function conversion means for converting the differential phase into an output frequency waveform function value by referring to the waveform table or another waveform table equivalent to the waveform table;
A pulse signal generating apparatus comprising: a synthesizing unit for synthesizing the sum phase waveform function value and the difference phase waveform function value to output the pulse signal. When the envelope phase increment of the envelope phase is represented by an Nth order polynomial (where N is 0 and a positive integer),
The N-th order envelope phase register means to the first order envelope phase register means each having an initial phase increment value and accumulatively adding and outputting the input value to the initial phase increment value in synchronization with the input timing. Connected in series,
An envelope initial phase register means having an initial phase value of the envelope and accumulating and adding an input value from the primary envelope phase register means in synchronization with an input timing is output. Connected to the output side of the means,
A predetermined constant envelope N + 1 order phase increment is inputted to the N-th order envelope phase register means in synchronization with an input timing, and an output of the envelope initial phase register means is obtained as the envelope phase. The pulse signal generator according to claim 1, characterized in that it is characterized in that:

Images