CN107104656B - Filter time constant changing circuit and D/A converting circuit - Google Patents

Abstract

The invention discloses a filter time constant changing circuit and a D/A conversion circuit, which reduces the ripple of the output of a filter in a structure that the time constant of the filter is changed according to the change of an input signal. The filter time constant changing circuit (4) includes: a change amount calculation unit (40, 41) that calculates a change amount h (T, T) ═ f (T) — f (T-T) per a predetermined time T, which is obtained from a value f (T) of an input signal at a current time and a value f (T-T) of an input signal T (T-T) T time before the current time; and a comparison unit (42) that outputs a control signal for controlling the time constant of the filter (2) on the basis of the result of comparison between the variation h (T, T) and a predetermined threshold value.

Description

Filter time constant changing circuit and D/A conversion circuit

technical field

The present invention relates to a filter time constant changing circuit for changing the time constant of a filter according to a change of an input signal, and a D/A conversion circuit using the filter time constant changing circuit.

Background technique

In the output of the D/A conversion circuit, a method in which the waveform of the pulse is smoothed by a low-pass filter and output is generally adopted. Depending on the size of the time constant of the filter at this time, the response speed and the ripple cannot be achieved at the same time. In fact, when determining the time constant of the filter, for example, the response speed as fast as possible is determined in the value of the allowable ripple. On the other hand, when a large change in the output of the D/A conversion circuit is carried out stepwise, there is a case where a delay in follow-up due to the limitation of the response speed becomes a problem.

As a solution to such a problem, a technique has been proposed in which, when there is a large and abrupt change in the output of the D/A conversion circuit, by reducing the time constant of the filter, the ripple is limited and the response is accelerated (refer to the patent Reference 1). FIG. 14 is a block diagram showing the configuration of the D/A conversion circuit disclosed in Patent Document 1. As shown in FIG. The D/A conversion circuit of FIG. 14 uses an ASIC (Application Specific Integrated Circuit) 300 to improve the responsiveness. When the digital data for creating the PWM signal is changed, it is switched to an analog filter with a small time constant formed by the resistor 302 and the capacitor 303 to improve the responsiveness. When the digital data does not change, it switches to an analog filter with a large time constant composed of the resistor 301 and the capacitor 303 to suppress the output ripple.

When the time constant of the filter is changed in accordance with the magnitude of the change in the input signal input from the D/A conversion circuit to the filter, as common sense, a configuration such as that shown in FIG. 15 is conceivable. In the example of FIG. 15 , the change amount calculation unit 100 calculates the change amount Δf(t)/Δt of the input signal f(t), and based on the change amount Δf(t)/Δt, the time constant changing unit 101 changes the value of the filter 102 time constant. In the example of FIG. 16, when the input signal f(t) changes, the time constant of the filter 102 is changed. On the other hand, in the example of FIG. 17 , when the change amount Δf(t)/Δt exceeds the threshold value TH, the time constant of the filter 102 is changed.

Generally, the processing performed by the change amount calculation unit 100 corresponds to differentiation, and the generation of the control signal for changing the time constant of the filter 102 (the filter time constant change flag in FIGS. 16 and 17 ) is limited to the input signal f ( t) A period with a slope. Therefore, when the time width in which the input signal f(t) has a slope is very short, the time width in which the control signal for changing the time constant is generated is also shortened. The appropriate time width for changing the time constant of the filter 102 depends on the time constant of the filter 102 and the purpose of the filter 102. Therefore, if the time width for generating the control signal for changing the time constant is shortened, there will be There is a possibility that the function for the purpose of changing the time constant of the filter 102 cannot be sufficiently exerted.

For example, consider a case where the time constant of the filter 102 is temporarily reduced in response to a change in the input signal f(t) in order to respond to a sudden change in the input signal f(t) above a certain level. In this case, the time width for changing the time constant needs to be a length that can sufficiently follow the input signal f(t) according to the time constant of the filter 102 after the change. However, even if the time constant of the filter 102 is reduced only during the change period of the input signal f(t) (only during the period when the differential value exceeds the threshold value TH), the output of the filter 102 becomes as shown in FIG. 18, and the filtering cannot be sufficiently exerted. function of the device 102. In the example of Fig. 18, g(t) represents the actual output of the filter 102, and g'(t) represents the ideal output. It can be seen from FIG. 18 that the output g(t) of the filter 102 cannot follow the change of the input signal f(t).

With regard to the problem of the configuration shown in FIG. 15 , it is easy to think that the output of the variation calculation unit 100 (differential value of the input signal f(t)) is triggered by using a timer, and the output of the variation calculation unit 100 exceeds the threshold value. In the case of the filter 102, the function of the filter 102 can be fully exhibited by changing the time constant of the filter 102 and holding it for a certain period of time. In the example of FIG. 19 , the timer 103 is provided between the change amount calculation unit 100 and the time constant changing unit 101, and the timer 103 generates a signal of a time width necessary to follow the input signal f(t), As shown in FIG. 20 , this provides a time width for changing the time constant necessary to follow the input signal f(t).

However, in the configuration shown in FIG. 19 , when the output of the change amount calculation unit 100 (the differential value of the input signal f(t)) is used as a trigger for changing the time constant of the filter 102 , there is an increase that is unintentionally large. The risk of misoperation in the scene. For example, as shown in FIG. 21 , the time constant of the constant-time filter 102 is changed in response to the instantaneous change of the input signal f(t), and there is an unintended region in the output g(t) of the filter 102 Ripple problem.

prior art literature

Patent Literature

Patent Document 1: Japanese Patent Laid-Open No. 2003-101413

SUMMARY OF THE INVENTION

Invention to solve problem

As described above, in the configuration in which the time constant of the filter is changed in accordance with the change of the input signal, there is a problem that if the output of the filter is made to sufficiently follow the change of the input signal, the output of the filter has the following problem. larger ripple.

In addition, the above problem is not limited to the filter that takes the output of the D/A conversion circuit as input, but also occurs similarly to a filter that filters an analog signal and inputs it to the A/D conversion circuit, for example.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a filter time constant change capable of reducing the ripple in the output of the filter in a configuration in which the time constant of the filter is changed in accordance with the change of the input signal. circuit.

technical solutions to problems

The present invention is a filter time constant changing circuit for determining a time constant of a filter for filtering an input signal, wherein the filter time constant changing circuit is characterized by a change amount calculation unit that The change amount h(t, T)=f(t)- f(t-T); and a comparison unit that outputs a control signal for controlling the time constant of the filter based on the result of comparing the change amount h(t, T) with a predetermined threshold value.

Further, in one configuration example of the filter time constant changing circuit according to the present invention, when the amount of change h(t, T) is equal to or smaller than the threshold value, the comparison unit outputs the output that causes the The time constant of the filter is a control signal of the first filter time constant τ1; when the amount of change h(t, T) exceeds the threshold, the comparison unit outputs a change in the time constant of the filter It is a control signal of the second filter time constant τ2 smaller than the first filter time constant τ1.

In addition, in one configuration example of the filter time constant changing circuit according to the present invention, the predetermined time T is from a time point when the input signal starts to change until the second filter time constant In τ2, the time until the value of the output signal of the filter reaches a predetermined ratio with respect to the final value of the change.

In addition, the D/A conversion circuit of the present invention is characterized by comprising: a D/A conversion part that converts a digital input signal into an analog signal; a filter that smoothes the output signal of the D/A conversion part; and the filter time constant changing circuit according to any one of claims 1 to 3, which takes the digital input signal as an input and outputs a control signal for controlling the time constant of the filter.

effect of invention

According to the present invention, by providing the change amount calculation unit and the comparison unit, the output of the filter can be made to follow the change of the input signal, and the ripple of the output of the filter can be reduced. In addition, in the present invention, since the ripple reduction of the output of the filter can be realized without using a conventional timer, the circuit installation can be easily performed.

Description of drawings

FIG. 1 is a waveform diagram illustrating the principle of the variation calculation according to the present invention.

FIG. 2 is a waveform diagram illustrating the principle of the variation calculation according to the present invention.

FIG. 3 is a waveform diagram illustrating the operation of changing the time constant of the filter in the present invention.

FIG. 4 is a waveform diagram showing another example of the calculation of the amount of change according to the present invention.

FIG. 5 is a waveform diagram showing another example of the calculation of the amount of change according to the present invention.

FIG. 6 is a waveform diagram illustrating the operation of changing the time constant of the filter in the present invention.

FIG. 7 is a waveform diagram showing another example of the calculation of the amount of change according to the present invention.

FIG. 8 is a waveform diagram showing another example of the calculation of the amount of change according to the present invention.

FIG. 9 is a waveform diagram illustrating the operation of changing the time constant of the filter in the present invention.

10 is a block diagram showing a configuration of a D/A conversion circuit according to an embodiment of the present invention.

11 is a diagram illustrating a method of setting a predetermined time in the embodiment of the present invention.

12 is a waveform diagram showing an example of the response of the D/A conversion circuit according to the embodiment of the present invention.

13 is a waveform diagram showing an actual measurement result of the response of the D/A conversion circuit according to the embodiment of the present invention.

FIG. 14 is a block diagram showing the configuration of a conventional D/A conversion circuit.

FIG. 15 is a block diagram showing a configuration for changing the time constant of the filter.

FIG. 16 is a waveform diagram illustrating the operation of changing the time constant of the filter in the configuration of FIG. 15 .

FIG. 17 is a waveform diagram illustrating another operation of changing the time constant of the filter in the configuration of FIG. 15 .

FIG. 18 is a waveform diagram illustrating a problem of the configuration of FIG. 15 .

FIG. 19 is a block diagram showing another configuration for changing the time constant of the filter.

FIG. 20 is a waveform diagram illustrating the operation of changing the time constant of the filter in the configuration of FIG. 19 .

FIG. 21 is a waveform diagram illustrating a problem of the configuration of FIG. 19 .

Detailed ways

[principle of invention]

FIGS. 1(A) to 1(C) and FIGS. 2(A) to 2(C) are waveform diagrams for explaining the principle of the variation calculation according to the present invention. In the present invention, the existing configuration of calculating the amount of change (differential value) of the input signal is replaced by the configuration of calculating the amount of change for each time width T corresponding to the change in the time constant that is the purpose of the filter, thereby , so that the effect of changing the time constant of the filter is appropriately reflected. Specifically, the value f(t) of the input signal at the current time point at which the change amount is to be calculated and the value f(t-T) of the input signal before T time from the current time point are calculated for each predetermined time. The amount of change h(t, T)=f(t)-f(t-T), and the time constant of the filter is determined according to the calculation result. In the examples of FIGS. 1(A) to 1(C) and 2(A) to 2(C), it is shown that the change amount h(t) is successively calculated according to the input signal that changes with time , T) condition.

When the amount of change h(t, T) calculated as described above exceeds the predetermined threshold value TH, if the time constant of the filter becomes small (the filter time constant change flag is set to "1"), the filter The output g(t) of is as shown in FIG. 3 . g'(t) in Figure 3 represents the ideal output.

Comparing with the case of FIG. 20, it can be seen that the output g(t) of the filter of the present invention follows the input signal f(t) earlier.

In addition, in the present invention, since the amount of change h(t, T) is in a format that includes the total amount of the magnitude of the change in the input signal f(t) as information, the ripple as described in FIG. 21 does not occur. question. FIG. 4(A) to FIG. 4(D) and FIG. 5 show the results of the calculation of the change amount of the present invention for the instantaneous change of the input signal f(t) as shown in FIG. 21 .

When the time constant of the filter is reduced when the calculation result of FIG. 4 (A) to FIG. 4 (D) and FIG. 5 exceeds the threshold value TH, the output g(t) of the filter is as shown in FIG. 6 . as shown. In the present invention, the threshold value TH is defined according to the maximum value that can be obtained by the difference between two points of the input signal (f(t)-f(t-T), so that malfunction due to noise lower than the threshold value TH does not occur. .

In addition, in the present invention, even when noise of a very sharp and large value is mixed into the input signal f(t), the filter is applied at a time width within the time width of the peak of the noise × 2. Since the time constant is changed, the influence on the filter output can be expected to be smaller than that in the case of using the timer in FIG. 19 . 7(A) to 7(C), FIG. 8(A), and FIG. 8(B) show the results of the variation calculation of the present invention for such noise input.

When the calculation results of FIGS. 7(A) to 7(C), 8(A), and 8(B) exceed the threshold value TH, the time constant of the filter is reduced. The output g(t) of the filter is as shown in FIG. 9 .

As described above, the present invention can solve the problem that ripples are generated in the output g(t) of the unintended area filter, which is caused in the prior art.

[Embodiment]

Hereinafter, embodiments of the present invention will be described with reference to the drawings. 10 is a block diagram showing a configuration of a D/A conversion circuit according to an embodiment of the present invention. The D/A conversion circuit of this embodiment is composed of the following parts: a D/A conversion unit 1, which converts the 16-bit digital input signal f(t) into an analog signal; and a low-pass filter 2, which is used for the D/A conversion unit The output signal of 1 is smoothed; the buffer circuit 3 is connected to the output terminal of the low-pass filter 2; time constant of device 2.

As the D/A converter 1, various configurations such as a ΔΣ modulator, a PWM (Pulse Width Modulation) modulator, and a T-resistor type D/A converter can be applied, but the present invention is not limited to D/A conversion method.

The low-pass filter 2 is composed of the following parts: resistors R1 and R2, which are provided in series between the output terminal of the D/A converter 1 and the input terminal of the buffer circuit 3, and a switch SW1, which is changed according to the time constant from the filter. The control signal of the circuit 4 short-circuits the resistor R1; and the capacitor C1, which is provided between the input terminal of the buffer circuit 3 and the ground.

The filter time constant changing circuit 4 includes a delay unit 40, a subtraction unit 41, and a comparison unit 42 (comparator). The delay unit 40 and the subtraction unit 41 constitute a change amount calculation unit.

The delay unit 40 delays the digital input signal f(t) by a predetermined time T. As the delay unit 40, a multi-stage cascaded flip-flop can be used.

The subtraction unit 41 subtracts the digital output signal of the delay unit 40 from the digital input signal f(t). In this way, the amount of change of the input signal at each time T can be calculated, that is, h(t, T)=f(t)-f(t-T) can be calculated.

The comparison unit 42 compares the digital output signal of the subtraction unit 41 with a predetermined threshold value TH. When the value of the digital output signal of the subtraction unit 41 exceeds the threshold value TH, the comparison unit 42 sets the control signal (filter time constant change flag) for controlling the switch SW1 of the low-pass filter 2 to "1" (High, high), when the value of the digital output signal of the subtraction unit 41 is equal to or less than the threshold value TH, the comparison unit 42 sets the control signal to “0” (Low). The threshold value TH may be set in advance in consideration of the maximum value of the difference (f(t)−f(t−T) between two points of the input signal) that can be obtained and the magnitude of noise to be removed.

The low-pass filter 2 is an active filter capable of changing the time constant. When the control signal output from the comparison unit 42 is "1", the switch SW1 is turned on. Thereby, since the resistor R1 is short-circuited, the time constant of the low-pass filter 2 becomes small. When the control signal output from the comparison unit 42 is "0", the switch SW1 is turned off. Accordingly, since the resistors R1 and R2 are connected in series, the time constant of the low-pass filter 2 increases.

As shown in FIG. 11 , the predetermined time T serving as the delay time of the delay unit 40 may be set to T in advance as a time from the time when the input signal f(t) starts to change until D/A The time until the value of the output signal g(t) of the conversion circuit reaches a predetermined ratio A (for example, 63.2% or 90%) with respect to the final value (100%) of the change. The ratio A may be set according to the desired performance of the D/A conversion circuit.

In this embodiment, the speed at which the output signal g(t) of the D/A conversion circuit follows the input signal f(t) and the level of the ripple of the output signal g(t) are determined by the time constant of the low-pass filter 2, The speed and the level of ripple cannot be compromised at the same time. In the present embodiment, the low-pass filter 2 is based on, for example, a time constant of τ1=10 msec, but on the other hand, a high-speed response of 1 msec may be required. However, if the time constant of the low-pass filter 2 is τ2=1 msec, the ripple of the output signal g(t) becomes large. Therefore, T=1 msec, A=63.2%, and a high-speed response mode capable of following abrupt changes in the input signal f(t) is prepared.

According to the present embodiment, as shown in FIG. 12 , when the input signal f(t) changes, the operation is as follows: at time t1 when the change amount h(t, T) exceeds the threshold value TH, the low-pass filter The time constant of 2 is changed from, for example, τ1=10msec to τ2=1msec (switch SW1 is turned on); at the time point t2 when the output signal g(t) reaches the ratio A=63.2% of the input signal f(t), the low-pass filter The time constant of the device 2 is changed from τ2 = 1 msec to τ1 = 10 msec (switch SW1 is turned off).

FIG. 13 is a waveform diagram showing an actual measurement result of the response of the D/A conversion circuit of the present embodiment. 130 denotes the output signal g(t) of the D/A conversion circuit when the time constant of the low-pass filter 2 is fixed at 1 msec, and 131 denotes the output signal g(t) when the time constant of the low-pass filter 2 is fixed at 10 msec The output signal g(t) of the D/A conversion circuit, 132 represents the output signal g(t) of the D/A conversion circuit of the present embodiment.

It can be seen from FIG. 13 that if the time constant of the low-pass filter 2 is reduced, a large ripple appears in the output signal g(t). Therefore, it is desired to reduce the time constant of the low-pass filter 2 only during the period of following the change of the input signal f(t), but it is not enough to reduce the time constant only at the moment of the change of the input signal f(t). Therefore, in the present embodiment, by calculating the change amount h(t, T) of the input signal f(t), the time of the low-pass filter 2 is reduced only when the change amount h(t, T) exceeds the threshold value TH constant to achieve a sufficient time for changing the time constant.

In the present embodiment, since the output ripple can be reduced without using the timer shown in FIG. 19 , it can be easily mounted on the D/A conversion circuit of the integrated circuit.

In addition, in the present embodiment, the first-order filter has been described as an example of the low-pass filter 2, but the present invention is not limited to this. The present invention can also be applied to other filters such as filters.

In the present embodiment, an example of application to a D/A conversion circuit has been described, but the filter time constant changing circuit 4 of the present invention may be applied as an input portion of an A/D conversion circuit. In this case, the analog input signal f(t) may be input to the low-pass filter 2, and the output signal of the low-pass filter 2 may be input to the A/D conversion circuit. The filter time constant changing circuit 4 is constituted by an analog circuit. As an example of the delay unit 40 , there is, for example, a CCD (Charge-Coupled Device, charge-coupled device).

【Industrial possibilities】

The present invention can be applied to a technique of changing the time constant of a filter based on an input signal.

Symbol Description

1...D/A conversion unit, 2...low-pass filter, 3...buffer circuit, 4...filter time constant changing circuit, 40...delay unit, 41...subtraction unit, 42...comparison unit, R1, R2...resistors, C1...capacitor, SW1...switch.

Claims (2)

1. A filter time constant changing circuit that determines a time constant of a filter for filtering an input signal, the filter time constant changing circuit being characterized by comprising: A change amount calculation unit that calculates a change per predetermined time T obtained from the value f(t) of the input signal at the current time point and the value f(t-T) of the input signal before T time from the current time point The quantity h(t, T)=f(t)-f(t-T); and a comparison unit that outputs a control signal for controlling the time constant of the filter based on the result of comparing the change amount h(t, T) with a predetermined threshold value, When the amount of change h(t, T) is equal to or smaller than the threshold value, the comparison unit outputs a control signal for setting the time constant of the filter to be the first filter time constant τ1; When the amount of change h(t, T) exceeds the threshold, the comparison unit outputs a second filter time for changing the time constant of the filter to be smaller than the first filter time constant τ1 control signal of constant τ2, The predetermined time T is the value of the output signal of the filter relative to the value of the output signal of the filter after the time constant of the filter is changed to the second filter time constant τ2 from the time point when the change of the input signal starts. The time until the final value of the change of the input signal reaches a predetermined ratio. 2. a D/A conversion circuit, is characterized in that, comprises: D/A conversion part, which converts the digital input signal into an analog signal; a filter that smoothes the output signal of the D/A conversion section; and 1. The filter time constant changing circuit according to claim 1, which takes the digital input signal as an input and outputs a control signal for controlling the time constant of the filter.

Images