JP2010124250A - A/d conversion apparatus, d/a conversion apparatus, and signal processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To correct gain errors of an input gain switching unit on former stages of an A/D conversion unit. <P>SOLUTION: In an A/D conversion apparatus, a control unit 31 allows a base signal of 1 kHz sine-wave generated by a base signal generation unit 25 to be input into an input gain changeover unit 21 at a measuring time, and upon changeover to 6 stages from the minimum gain to the maximum gain, allows the A/D conversion unit 22 to determine and store a compensation coefficient for denying a gain relative error of each stage for the minimum gain from a digital signal level after the A/D conversion. At a normal time, the digital signal level that has been A/D-converted from an input analog audio signal through the input gain changeover unit 21 in the A/D conversion unit 22 is monitored. When a lower level state persists or when it increases to excessively higher, the gain is changed over so that the input level of the A/D conversion unit 22 is optimum, and at the same time, the compensation coefficient corresponding to the changeover gain is set for a level correction unit 23 to allow the gain error to be corrected. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an A / D conversion device, a D / A conversion device, and a signal processing device, and in particular, an A / D conversion device, a D / A conversion device, and a signal processing that can switch a gain for an input analog signal or an output analog signal. Relates to the device.

  For example, in an in-vehicle audio system, the head unit is installed on the dashboard, but the power amplifier and the speaker box are often installed on a trunk or a door away from the head unit. An audio signal of a certain level is output from a source circuit unit such as a tuner unit and a CD playback unit in the head unit, and after the volume is adjusted to a user desired volume by a volume adjustment unit provided in the head unit. And output to the power amplifier via the cable. If you want to perform signal processing such as band limitation with a DSP (digital signal processor) with a power amplifier, the analog audio signal after volume adjustment input from the head unit side is converted into a digital audio signal with an A / D converter. After the processing, the original analog signal is restored by the D / A converter.

By the way, when the user listens to music, there is a big difference in the signal level between the maximum volume and the minimum volume, and the normal volume is lower than the volume that is rarely heard. Sometimes you want to hear a volume that is even lower than the normal volume. The difference between the low volume and the high volume is usually 40 dB, although there are individual differences. When the analog audio signal after volume adjustment is A / D converted, the system needs to be configured not to clip at the maximum signal level of the system. For example, when a 16-bit A / D converter is used to convert a full-bit digital signal when the signal level is maximum, the signal level is 40 dB lower than the maximum volume at a low volume. When the volume is low, the upper 7 bits of the A / D converter are not used, which is equivalent to using a 9-bit A / D converter. There is a problem that the quality is very deteriorated. In this regard, it is conceivable to provide a volume adjustment unit at the subsequent stage of the DSP of the power amplifier and to avoid deterioration in quality by setting the input of the A / D converter to a constant signal level. However, since the volume adjustment operation unit is usually provided in the head unit, it is necessary to wire the volume adjustment signal line separately from the audio signal line from the dashboard head unit to the trunk room power amplifier. Since an input terminal for a volume adjustment signal is not attached, there is a drawback that the versatility of the product is lost, and this is not a realistic solution.
As an additional function of the speaker, the same is true when trying to implement band limiting or the like by signal processing by a DSP. Generally, an analog audio signal after volume adjustment is input to the speaker. There is a drawback that the quality of the audio signal is extremely deteriorated.

  In order to solve this problem, in the case of processing an analog audio signal after volume adjustment whose signal level is largely variable, the A / D converter 2 and the D / A converter before and after the DSP 1 as shown in FIG. 3 is further provided with an input gain variable circuit 4 and an output gain variable circuit 5. When the level of the input analog audio signal Ain increases due to the increase in volume and the output of the A / D converter 2 is likely to clip, the gain of the input gain variable circuit 4 is lowered, and the input level of the A / D converter 2 is decreased. On the other hand, the gain of the output gain variable circuit 5 is increased by that amount so that an output level corresponding to the volume desired by the user can be obtained as a whole. On the contrary, the level of the input analog audio signal Ain is reduced by reducing the volume. When the signal distortion of the output of the A / D converter 2 is likely to become remarkable, the gain of the input gain variable circuit 4 is increased to optimize the input level of the A / D converter 2, while the output is correspondingly increased. It is conceivable to lower the gain of the variable gain circuit 5 so that an output level commensurate with the volume desired by the user can be obtained as a whole. For example, in Japanese Utility Model Laid-Open No. 5-55617, when the level of the input analog audio signal Ain becomes excessive, the gain is decreased by the input gain variable circuit (first VCA), and the output gain variable circuit (first (VCA of 2) has been disclosed (Patent Document 1).

  However, the input gain variable circuit and output gain variable circuit composed of analog circuits have poor gain accuracy due to restrictions on combinations of circuit constants of element parts and variations in circuit constants of element parts. When the gain is varied in a complementary manner by the gain variable circuit, the signal becomes discontinuous and a large noise is generated. For example, FIG. 2 shows an analog gain switching circuit composed of an inverting amplifier OP, resistors R1 to R9, and switch elements SW1 to SW3. 74HC4053 (for example, TC74HC4053AP manufactured by Toshiba) is used as the switch element, and the switch element is relatively accurate with a small number of switch elements. The gain can be switched, and as shown in FIG. 3, the gain is set to x1, x2, x4, x8 by combining "H" and "L" of the 3-bit control signals ContA, ContB, and ContC. , X16, and x32 can be switched to a setting gain that is a multiple of 2 in 6 stages. In the case of the configuration of FIG. 2, the gains for calculation of x1, x2, x4, x8, x16, and x32 are 0.88 (error -12.00%) and 1.76 (error -12.12). 00%), 3.52 (error -12.00%), 7.08 (error -11.50%), 14.06 (error -12.13%), 27.94 (error -12.69%) ), And x2, x4, x8, x16, and x32 are 1.995 times (relative error -0.25%) and 3.991 times (same -0). .22%), 8.014 times (-0.22%), 15.920 times (-0.50%), 31.634 (-1.14%) (however, the resistance value Variation error is not considered). The gain switching circuit is a fairly accurate circuit, but the accuracy is insufficient when the gain is varied following the level change of the input signal. For example, when the gain switching error including the variation in the calculation and the resistance value is 1%, a relatively large pulse noise is generated as an audio signal of about −40 dB with respect to the signal.

  Apart from this, there is an electronic volume circuit that adjusts the volume of an audio signal in the analog domain. This electronic volume circuit has a resistance voltage dividing circuit in which a number of resistance elements are connected in series between the input terminal and the ground, and a number of analog switches interposed between the input terminal and the connection point between each resistor and the output terminal. The analog switch is alternatively closed to divide the voltage value of the input analog audio signal by a variable voltage dividing value and attenuate the signal level. In this electronic volume circuit, for example, when attenuation is performed in steps of 1 dB from 0 dB to -35 dB, 36 resistive elements and 36 analog switches are required, which makes the configuration complicated and large, and increases the cost. . On the other hand, the sound volume can be adjusted by a calculation process using a DSP in the digital tall region. However, when the attenuation is increased, the effective bit length is reduced and the quantization distortion is increased.

Therefore, as shown in FIG. 4, the A / D converter 7 that performs A / D conversion of the input analog audio signal, the DSP 8 that performs attenuation calculation in the digital basin, and the output of the DSP 8 are D / A converted before the electronic volume circuit 6. The D / A converter 9 is provided so that the electronic volume circuit 6 attenuates with a large step width of 6 dB, and the DSP 8 attenuates with a small step width of 1 dB, so that the control circuit 10 has a target sound volume (target attenuation). As described above, a hybrid volume adjustment circuit that controls the electronic volume circuit 6 and the DSP 8 has been proposed (Patent Document 2).
However, the electronic volume circuit 6 composed of an analog circuit has poor linearity as compared with the DSP 8, and for example, the electronic volume circuit 6 may actually have only −5 dB of −6 dB. At this time, in order to set the overall attenuation to −5 dB, the electronic volume circuit 6 is switched to 0 dB and the DSP 8 is switched to −5 dB. When the entire attenuation is decreased by 1 dB to −6 dB, the electronic volume is reduced. There is a problem that the attenuation is almost the same when the attenuation of the circuit 6 is switched to −6 dB and the DSP 8 is switched to 0 dB, and the sound volume hardly changes.

Japanese Utility Model Laid-Open No. 5-55617, paragraph 0025, FIG. Japanese Patent Laid-Open No. 2001-7670, paragraph 0026, FIG.

  An object of the present invention is to provide an A / D conversion device capable of correcting a gain switching error of an input gain switching unit provided in a preceding stage of the A / D conversion unit in view of the above-described problems of the prior art. It is another object of the present invention to provide a D / A converter capable of correcting a gain switching error of an output gain switching unit provided at a subsequent stage of the D / A converting unit. It is another object of the present invention to provide a signal processing apparatus capable of correcting a gain switching error between input gain switching means and output gain switching means provided at the front and rear stages.

In one of the A / D conversion devices of the present invention, an A / D conversion means for A / D converting an input analog signal, and a predetermined number of stages for the input analog signal are provided in the preceding stage of the A / D conversion means. In an A / D converter comprising an input gain switching means capable of switching to a set value gain, a reference signal generating means for generating a reference signal for gain measurement of the input gain switching means, and an input analog signal in normal times Is selected and input to the input gain switching means, and during measurement, the reference signal is selected and input to the input gain switching means, and during measurement, the reference signal is applied to the input gain switching means via the selection means. The level measurement means for measuring the output level of the A / D conversion means when the gain of the input gain switching means is switched to each stage, and the level measurement means. A correction coefficient determining means for determining a correction coefficient for each stage for canceling the gain error of each stage of the input gain switching means based on the level, and an analog input to the input gain switching means through the selection means in normal times. Correction means for performing level correction on the digital signal output from the A / D conversion means according to the gain of the input gain switching means when a signal is applied, based on the correction coefficient determined by the correction coefficient determination means It is characterized by having.
Another one of the A / D converters of the present invention is characterized in that the gain setting values at a plurality of stages of the input gain switching means are values of a multiplier of 2 or a multiplier of 1/2.
In one of the D / A conversion devices of the present invention, a D / A conversion means for D / A converting and outputting a digital signal, and a plurality of stages for the output analog signal are provided after the D / A conversion means. An output gain switching means capable of switching to a gain of a predetermined set value, a reference signal generating means for generating a reference signal for gain measurement of the output gain switching means, and an output in normal times Selects an analog signal and inputs it to the output gain switching means, selects a reference signal for measurement and inputs it to the output gain switching means, and inputs an analog signal output from the output gain switching means for measurement A / D conversion means, and at the time of measurement, when the reference signal is applied to the output gain switching means via the selection means, the gain of the output gain switching means is set to each stage. Level measuring means for measuring the output level of the A / D conversion means at the time of switching to each level, and for each stage for canceling the gain error at each stage of the output gain switching means based on the output level measured by the level measuring means A correction coefficient determining means for determining the correction coefficient of the D / A conversion means, and when the output of the D / A conversion means is applied to the output gain switching means via the selection means in the normal state, D is determined according to the gain of the output gain switching means. Correction means for performing level correction on the digital signal input to the / A conversion means based on the correction coefficient determined by the correction coefficient determination means.
Another one of the D / A converters according to the present invention is characterized in that the gain setting values at a plurality of stages of the output gain switching means are values of multipliers of 2 or multipliers of 1/2.
In one of the signal processing devices of the present invention, A / D conversion means for A / D converting an input analog signal, signal processing means for performing signal processing on an input digital signal output from the A / D conversion means, D / A conversion means for D / A converting and outputting the output digital signal of the signal processing means, and switching to a predetermined setting gain in a plurality of stages with respect to the input analog signal provided in the preceding stage of the A / D conversion means In a signal processing apparatus comprising: a possible input gain switching means; and an output gain switching means provided at a subsequent stage of the D / A conversion means and capable of switching to a predetermined set value gain in a plurality of stages for an output analog signal , A reference signal generating means for generating a reference signal for measuring the gain of the input gain switching means and the output gain switching means; In the first measurement, the reference signal is selected and input to the input gain switching means, and in the second measurement, the output of the output gain switching means is selected and input to the input gain switching means. A first selection means; and a second selection means for selecting an output analog signal during normal operation and inputting it to the output gain switching means, and during a second measurement, selecting a reference signal and inputting it to the output gain switching means. During the first measurement, when the reference signal is applied to the input gain switching means via the first selection means, the A / D conversion means when the gain of the input gain switching means is switched to each stage. When the output level is measured and the reference signal is applied to the input gain switching means via the second selection means, the output gain switching means, and the first selection means during the second measurement, the output gain is measured. The level measurement means for measuring the output level of the A / D conversion means when the gain of the switching means is switched to each stage, and the gain error at each stage of the input gain switching means based on the level measured by the level measurement means. Correction coefficient determining means for determining the correction coefficient for each stage for canceling and the correction coefficient for each stage for canceling the gain error of each stage of the output gain switching means; When the input analog signal is applied to the input gain switching means via the second input means and the output analog signal is applied to the output gain switching means via the second selection means, the gain of the input gain switching means and the output gain switching means Correction coefficient determined by the correction coefficient determination means for the digital signal between the A / D conversion means and the D / A conversion means in accordance with the gain of And a correcting means for performing level correction based on the above.
In another one of the signal processing devices of the present invention, the set values of the gains of the plurality of stages of the input gain switching means are values of a multiplier of 2 or a multiplier of 1/2, and the gains of the plurality of stages of the output gain switching means The set value is a value that is a multiplier of 2 or a multiplier of 1/2.

According to the A / D conversion device of the present invention, the correction coefficient for each step for canceling the gain error of each step of the input gain switching means is determined using the reference signal, and the A / D is supplied to the input gain switching means. Since the level correction is performed on the digital signal output from the A / D conversion means according to the gain of the input gain switching means when the input analog signal to be converted is applied, the analog circuit is configured. Even if there is an error in the gain of each stage of the input gain switching means, the gain for the signal can be switched accurately.
When digital signals after A / D conversion are processed, it is necessary to switch the level of past sample data held in the digital signal processing means in conjunction with the gain switching of the input gain switching means. If the gain setting values in a plurality of stages of the gain switching means are switched at a multiplier factor of 2 or a multiplier factor of 1/2, the level of past sample data can be easily switched by a bit shift operation. .
According to the D / A conversion device of the present invention, the correction coefficient for each step for canceling the gain error of each step of the output gain switching means is determined using the reference signal, and the D / A conversion target digital signal is determined. When an output analog signal obtained by D / A converting the signal is applied to the output gain switching means, level correction is performed on the digital signal to be D / A converted in accordance with the gain of the output gain switching means. Therefore, even if there is an error in the gain at each stage of the output gain switching means constituted by the analog circuit, the gain for the signal can be switched accurately.
Further, when the gain for the digital signal input to the D / A conversion means is varied in an offset manner with the gain of the output gain switching means, the gain setting values at a plurality of stages of the output gain switching means are a multiplier of 2 or 2 If the switching is performed by a multiplier, the switching of the gain for the digital signal input to the D / A conversion means can be easily executed by the bit shift operation.
According to the signal processing device of the present invention, the input gain switching means determines the correction coefficient for each stage for canceling the gain error of each stage of the input gain switching means and the output gain switching means using the reference signal. When an analog signal subject to A / D conversion is applied to and an output analog signal obtained by D / A converting the digital signal subject to D / A conversion is applied to the output gain switching means, the input gain switching means and the output Since level correction is performed on the digital signal between the A / D conversion means and the D / A conversion means in accordance with the gain of the gain switching means, the input gain switching means and the output gain switching configured by analog circuits are performed. Even if there is an error in the gain of each stage of the means, the gain for the signal can be switched accurately.
When digital signals after A / D conversion are processed, it is necessary to switch the level of past sample data held in the digital signal processing means in conjunction with the gain switching of the input gain switching means. If the gain setting values in multiple stages of the gain switching means are switched at a multiplier factor of 2 or a multiplier factor of 1/2, the level of past sample data can be easily switched by a bit shift operation. Similarly, when the gain for the digital signal input to the D / A conversion means is varied in an offset manner with respect to the gain of the output gain switching means, the gain setting values at a plurality of stages of the output gain switching means are multipliers of 2. If the signal is switched by a factor of 2 or a multiplier of 2, the digital signal input to the D / A conversion means is not affected. Gain switching can be easily performed by bit shift operation that.

  Hereinafter, the best mode of the present invention will be described based on examples.

A first embodiment of the present invention will be described with reference to FIG. FIG. 5 is a block diagram showing the configuration of the band-limiting signal processing apparatus according to the present invention.
In FIG. 5, reference numeral 20 denotes a switch unit that selects and outputs an input analog audio signal Ain after volume adjustment in the normal mode, and selects and outputs an analog reference signal ST described later in the measurement mode. Reference numeral 21 denotes an input gain switching unit capable of switching the gain for the analog signal output from the switch unit. The input gain switching unit 21 is configured in the same way as in FIG. 2 here, and the gain can be switched to a multiplier of 2 in the first to sixth stages. The first stage is set to a gain of x1, the second stage is set to x2, the third stage is set to x4, the fourth stage is set to x8, the fifth stage is set to x16, and the sixth stage is set to a gain of x32. However, the actual gain at each stage does not match the set value due to the limitation of the selection of the circuit constants of the element parts and the variation of the circuit constants, and there is an error. This error is corrected in the digital domain as will be described later. Reference numeral 22 denotes an A / D converter, which performs A / D conversion on the analog audio signal after gain adjustment by the input gain switching unit 21. A level correction unit 23 performs level correction on the digital audio signal after A / D conversion so as to cancel the gain error of the input gain switching unit 21. Reference numeral 24 denotes a digital signal processing unit which, as an example, executes signal processing as a low-pass filter that passes only a component having a frequency equal to or lower than a predetermined cutoff frequency in accordance with the signal flow of the third-order FIR filter shown in FIG. To do. Sampling period T of digital audio signal, discrete time nT (where n = 1, 2, 3,...), Input x (nT) and output y (nT) at time nT of digital signal processing unit 24,
y (nT) = a0.x (nT) + a1.x ((n-1) T) + a2.x ((n-2) T) + a3.x ((n-3) T)
a0, a1, a2, a3; calculation coefficients. Here, x ((n-1) T), x ((n-2) T), and x ((n-3) T) are past sample data.

  Reference numeral 25 is a reference signal generation section for generating a digital reference signal for gain measurement of the input gain switching section 21, and 26 is for selecting and outputting a digital audio signal after digital signal processing input from the digital signal processing section 24 in normal times. At the time of measurement, a switch unit for selecting and outputting the digital reference signal input from the reference signal generation unit 25, 27 a D / A conversion unit for D / A converting the digital signal input from the switch unit 26, and 28 This is an output gain switching unit that can switch the gain for the analog signal after D / A conversion, and an output analog signal Aout is output from the output gain switching unit 28. The output gain switching unit 28 enables the gain to be switched to a multiplier factor of 1/2 in the first to sixth stages. The first stage is x 1/32, the second stage is x 1/16, the third stage is x 1/8, the fourth stage is x 1/4, the fifth stage is x 1/2, the sixth stage is x The gain is set to 1/1. The output of the D / A conversion unit 27 is connected to the measurement mode side input terminal of the switch unit 20. A level measurement unit 29 measures the signal level of the digital signal output from the A / D conversion unit 22. Here, the signal level is measured by a peak value, but may be measured by an effective value (RMS). Reference numeral 30 denotes a correction coefficient storage unit that stores correction coefficients at each stage of the input gain switching unit 21.

Reference numeral 31 denotes a control unit, which applies a reference signal composed of a 1 kHz sine wave to the input gain switching unit 21 during measurement and switches the gain of the input gain switching unit 21 in order from the first to sixth stages. Using the signal level measured by the level measuring unit 29, the correction coefficient for each step for canceling the gain error of each step of the input gain switching unit 21 is determined and stored in the correction coefficient storage unit 30. During normal times, the signal level of the input analog audio signal input to the A / D conversion unit 22 via the input gain switching unit 21 is monitored, and a small level state below a predetermined first reference level continues for a certain period of time. Then, in order to optimize the input level of the A / D conversion unit 22 so that the quantization distortion does not become remarkable, the gain of the input gain switching unit 21 is increased by one step, and the gain of the output gain switching unit 28 is increased accordingly. Move down one step. At this time, a correction coefficient corresponding to the gain of the input gain switching unit 21 is set in the level correction unit 23, and the gain error of the input gain switching unit 21 is corrected by multiplying the digital audio signal after A / D conversion by the correction coefficient. At the same time, the digital signal processing unit 24 corrects the past sample data to a double size by arithmetic shift operation (1 bit left shift), and the output analog audio signal Aout output to the outside before and after the gain switching is obtained. Avoid discontinuities.
Further, when the signal level of the input analog audio signal is monitored, if the signal level exceeds a predetermined second reference level (> first reference level), the input level of the A / D conversion unit 22 is exceeded. In order to prevent clipping due to excessive input, the gain of the input gain switching unit 21 is decreased by one step, and the gain of the output gain switching unit 28 is increased by one step accordingly. At this time, a correction coefficient corresponding to the gain of the input gain switching unit 21 is set in the level correction unit 23, and the gain error of the input gain switching unit 21 is corrected by multiplying the digital audio signal after A / D conversion by the correction coefficient. In addition, the digital signal processing unit 24 corrects the past sample data to 1/2 times larger by arithmetic shift operation (1 bit right shift), and outputs an analog audio signal output to the outside before and after the gain switching. Aout should not be discontinuous. The digital domain is from A / D converter 22 to D / A converter 27 in FIG. In the digital domain, sample data is represented by an n-bit length, and MSB represents a sign of ±. However, it is assumed that the digital signal processing unit 24 expands to the lower side by 5 bits, performs an operation with a length of (n + 5) bits, and rounds to n bits at the time of output.

FIG. 7 and FIG. 8 are flowcharts showing control processing at the time of measurement and normal time by the control unit 31, and the operation of the above-described embodiment will be described below with reference to these drawings.
Here, the gain of x1 of the input gain switching unit 21 is used as a reference (not corrected), and correction is performed so as to cancel the relative error of gain at other stages.
(1) During measurement When the apparatus is turned on, the control unit 31 switches the switch units 20 and 26 to the measurement mode side, and the reference signal generation unit 25 has an amplitude that is about 20% smaller than half of the full scale of the digital domain. A digital reference signal composed of a 1 kHz sine wave is generated, and the gain of the input gain switching unit 21 is further switched to x1 in the first stage (steps S10 to S12 in FIG. 7). The digital reference signal is generated by repeatedly outputting sample data having a waveform of one sine wave. The digital reference signal is input to the D / A conversion unit 27 via the switch unit 26, converted into the analog reference signal ST, and then input to the input gain switching unit 21 via the switch unit 20. Here, after the gain is multiplied by 1, it is converted into a digital signal by the A / D converter 22. Then, the peak value is measured by the level measuring unit 29 and input to the control unit 31. The control unit 31 stores the input peak value in the correction coefficient storage unit 30 as the reference level SP, and stores the first-stage gain correction coefficient P1 as 1 (steps S13 to S15).

  Next, the gain of the input gain switching unit 21 is increased by one step and switched to x2, and the reference signal generating unit 25 shifts each sample data of the waveform of one cycle of the digital reference signal by one bit by arithmetic shift calculation. Then, the amplitude is set to ½ times the current amplitude (steps S16 to S18). As a result, the analog reference signal ST is halved in amplitude and input to the input gain switching unit 21, and the gain is set to x2. The control unit 31 calculates the correction coefficient P2 of x2 by dividing the reference level SP by the peak value input from the level measurement unit 29, and stores it in the correction coefficient storage unit 30 (steps S19 and S20). Next, the gain of the input gain switching unit 21 is increased by one step and switched to x4, and the reference signal generation unit 25 further shifts each sample data of the waveform of one cycle of the digital reference signal by one bit to the right by arithmetic shift calculation. By shifting, the amplitude is further halved, the peak value input from the level measuring unit 29 is divided by the reference level SP, and a correction coefficient P3 of x4 is calculated and stored in the correction coefficient storage unit 30 (step S21). NO, S22, S17 to S20). Thereafter, the same processing is repeated up to the sixth stage x32. Thus, every time the power is turned on, the gain correction coefficients P1 to P6 of each stage of the input gain switching unit 21 are determined and stored in the correction coefficient storage unit 30.

(2) Normal time When the correction coefficient is determined, the control unit 31 switches the switch units 20 and 26 to the normal mode side and sets the gain of the input gain switching unit 21 to x1 in the first stage, and the output gain switching unit. The gain of 28 is switched to x1 / 1 in the sixth stage (YES in step S21 in FIG. 7, steps S30 and S31 in FIG. 8). Further, the correction coefficient P1 = 1 corresponding to x1 is set in the level correction unit 23 (step S32). The input analog audio signal Ain after the volume adjustment is input to the input gain switching unit 21 via the switch unit 20, and after being gained by × 1, it is converted into a digital audio signal by the A / D conversion unit 22. Then, after the level correction unit 23 multiplies the correction coefficient 1, the digital signal processing unit 24 extracts a low frequency component by a third-order FIR filter. Then, the signal is input to the D / A conversion unit 27 via the switch unit 26 and converted into an analog audio signal, and then the gain is multiplied by 1/1 and output to the subsequent stage by the output gain switching unit 28. The control unit 31 constantly monitors the level (here, the peak value) of the digital audio signal measured by the level measurement unit 29 (step S33), and a small level state below the first reference level is maintained for a certain time or longer. The determination of whether or not it has continued and the determination of whether or not it has exceeded a predetermined constant level (= second reference level) are repeated (repetition of steps S34 and S35).

  If the low level state continues for a certain time or more and the answer is YES in step S34, the current input gain switching unit 21 is not yet in the sixth stage, so the input gain switching unit 21 is switched to x2 in the second stage. The dynamic range of the A / D conversion unit 22 is optimized, and the output gain switching unit 28 is switched to x1 / 2 of the fifth stage accordingly (steps S36 to S38). Subsequently, the correction coefficient P2 for the second stage is read from the correction coefficient storage section 30 and set in the level correction section 23, and the gain from the input gain switching section 21 to the level correction section 23 is accurately 2 with respect to the first stage. In addition, the past sample data of the FIR filter is shifted left by 1 bit by arithmetic shift operation with respect to the digital signal processing unit 24 so as to be doubled (steps S39 and S40). This prevents the analog audio signal output from the output gain switching unit 28 from becoming discontinuous before and after the gain switching.

  Thereafter, when the low level state continues for a certain time or more again, it is determined as YES in step S34, and the input gain switching unit 21 is switched to x4 in the third stage to optimize the dynamic range of the A / D conversion unit 22. Accordingly, the input gain switching unit 21 is switched to x1 / 4 of the fourth stage (steps S36 to S38). Subsequently, the correction coefficient P3 for the third stage is read from the correction coefficient storage section 30 and set in the level correction section 23, and the gain from the input gain switching section 21 to the level correction section 23 is accurately 4 with respect to the first stage. In addition, the past sample data of the FIR filter is shifted left by 1 bit by arithmetic shift operation with respect to the digital signal processing unit 24 so as to be doubled (steps S39 and S40).

  After that, when the measurement level becomes excessive exceeding a predetermined constant level, it is determined YES in step S35, and the current input gain switching unit 21 is not yet in the first stage. Returning to 2, the dynamic range of the A / D conversion unit 22 is optimized so as not to become an excessive input, and the output gain switching unit 28 is switched to x1 / 2 of the fifth stage accordingly (steps S41 to S43). Subsequently, the correction coefficient P2 for the second stage is read from the correction coefficient storage section 30 and set in the level correction section 23, and the gain from the input gain switching section 21 to the level correction section 23 is accurately 2 with respect to the first stage. In addition, the past sample data of the FIR filter is shifted by one bit to the digital signal processing unit 24 by an arithmetic shift operation so as to be ½ times larger (steps S44 and S45). This prevents the analog audio signal output from the output gain switching unit 28 before and after gain switching from becoming discontinuous.

According to this embodiment, the gain of the input gain switching unit 21 is switched according to the level of the input analog audio signal to be A / D converted, and the level is optimized and input to the A / D conversion unit 22. Quantization distortion caused by / D conversion can be reduced. Further, a correction coefficient for each step for canceling the gain error of each step of the input gain switching unit 21 is determined using the reference signal, and an analog audio signal to be A / D converted is applied to the input gain switching unit 21. Since the level correction is performed on the digital audio signal output from the A / D conversion unit 22 in accordance with the gain of the input gain switching unit 21, the input gain switching unit configured by an analog circuit is used. Even if there is an error in the gain at each stage of 21, the gain for the signal can be switched accurately.
When digital signal processing is performed on the digital audio signal after A / D conversion, it is necessary to switch the level of past sample data held in the digital signal processing unit 24 in conjunction with the gain switching of the input gain switching unit 21. However, since the gain setting values in a plurality of stages of the input gain switching unit 21 are switched by a multiplier of 2, the level of the past sample data can be easily switched by arithmetic shift calculation (right shift or left shift). be able to.

In the above-described embodiment, the correction is performed so as to cancel the relative error of the gains of the second to sixth stages when the gain of the first stage is used as a reference. However, the absolute error of the gain of each stage is canceled. It may be corrected. In this case, the peak value w is measured by the level measuring unit 29 in a state in which the digital reference signal ST having a known amplitude W is D / A converted and then input to the input gain switching unit 21 switched to the third stage gain. When the correction coefficient P3 is
W × 4 = w × P3
P3 = (W / w) × 4
It becomes. Correction coefficients at other stages can be obtained in the same manner.
In the above-described embodiment, the analog signal output from the D / A converter 27 is output to the outside after the level is changed by the output gain switching unit 28 in the normal state. If not, the output of the D / A converter 27 may be output to the outside as it is (see reference Aout ′). Furthermore, if conversion to an analog format is not necessary, the digital signal output from the switch unit 26 may be output to the outside as it is (see reference sign Dout).

A second embodiment of the present invention will be described with reference to FIG. FIG. 9 is a block diagram showing the configuration of the band-limiting signal processing apparatus according to the present invention. The same components as those in FIG.
In the first embodiment shown in FIG. 5, the input is switched before the A / D converter in accordance with the level of the input analog audio signal to be A / D converted to optimize the input level of the A / D converter. A gain switching unit is provided to correct the gain error of the input gain switching unit. On the other hand, in FIG. 9, the D / A conversion unit is preceded by a D / A conversion target digital audio signal level. Provided with a gain switching unit that switches the gain and optimizes the input level of the D / A conversion unit, and an output gain switching unit that switches the gain so as to cancel the gain of the gain switching unit after the D / A conversion unit And a level correction unit is inserted before the D / A conversion unit to correct the gain error of the output gain switching unit.
In FIG. 9, reference numeral 32 denotes a switch unit which selects and outputs an input analog audio signal Ain in the normal mode, and selects an analog reference signal ST ′ output from an output gain switching unit 28 described later in the measurement mode. And output. Reference numeral 24A denotes a digital signal processing unit that executes signal processing as a low-pass filter that passes only components having a frequency equal to or lower than a predetermined cutoff frequency in accordance with the signal flow of the third-order FIR filter shown in FIG. A gain switching unit 33 switches the gain for the band-limited digital audio signal output from the digital signal processing unit 24A. The gain can be switched to a multiplier of 2 in the first to sixth stages by arithmetic shift calculation. And The first stage is set to a gain of x1, the second stage is set to x2, the third stage is set to x4, the fourth stage is set to x8, the fifth stage is set to x16, and the sixth stage is set to a gain of x32. When the gain switching unit 33 sets the gain to x1, the sample data of the digital audio signal output from the digital signal processing unit 24A is output as it is, and the gain is set to x2 ^k (where k = 1, 2, 3, 4 5), the signal level is varied by shifting the sample data to the left by k bits by an arithmetic shift operation.

  An output gain switching unit 28 can switch the gain for the analog audio signal after D / A conversion. The output gain switching unit 28 can switch the gain to a multiplier factor of 1/2 in the first to sixth stages. The first stage is x 1/32, the second stage is x 1/16, the third stage is x 1/8, the fourth stage is x 1/4, the fifth stage is x 1/2, the sixth stage is x The gain is set to 1/1. However, the actual gain at each stage does not match the set value due to the limitation of the selection of the circuit constants of the element parts and the variation of the circuit constants, and there is an error. This error is corrected in the digital domain as will be described later. A level correction unit 34 performs level correction on the digital audio signal to be D / A converted so as to cancel the gain error of the output gain switching unit 28. Reference numeral 25 is a reference signal generation section for generating a digital reference signal for gain measurement of the output gain switching section 28, and 26 is for selecting a digital audio signal after processing the digital signal input via the level correction section 34 in normal times. A switch unit that outputs and selects and outputs the digital reference signal input from the reference signal generation unit 25 at the time of measurement, and 30A is a correction coefficient storage unit that stores correction coefficients at each stage of the output gain switching unit 28. . Reference numeral 35 denotes a switch unit that selects and outputs the output of the level correction unit 34 to the level measurement unit 29 in the normal mode, and selects and outputs the output of the A / D conversion unit 22 in the measurement mode.

Reference numeral 31A denotes a control unit, which applies a reference signal composed of a 1 kHz sine wave to the output gain switching unit 28 during measurement and causes the level measurement unit 29 to pass through the switch unit 32, the A / D conversion unit 22, and the switch unit 35. Input, the gain error at each stage of the output gain switching unit 28 using the signal level measured by the level measurement unit 29 at each stage while sequentially switching the gain of the output gain switching unit 28 from the sixth to the first stage. The correction coefficient for each step for canceling is determined and stored in the correction coefficient storage unit 30A. During normal times, the signal level of the digital audio signal input from the level correction unit 34 to the D / A conversion unit 27 via the switch unit 26 is monitored, and the small level state below the predetermined first reference level is constant. When the time continues, in order to optimize the input level of the D / A conversion unit 27 so that the quantization distortion does not become significant, the gain of the gain switching unit 33 is increased by one step, and the output gain switching unit 28 Decrease gain by one level. At this time, a correction coefficient corresponding to the gain of the output gain switching unit 28 is set in the level correction unit 34, and the gain error of the output gain switching unit 28 is corrected by multiplying the digital audio signal to be D / A converted by the correction coefficient. Thus, the output analog audio signal Aout is prevented from becoming discontinuous before and after the gain switching.
Further, when the signal level of the digital audio signal input to the D / A converter 27 is monitored, when the signal level exceeds a predetermined second reference level (> first reference level), In order to optimize the input level of the D / A conversion unit 27 and prevent clipping due to excessive input, the gain of the gain switching unit 33 is decreased by one step, and the gain of the output gain switching unit 28 is increased by one step accordingly. At this time, a correction coefficient corresponding to the gain of the output gain switching unit 28 is set in the level correction unit 34, and the gain error of the output gain switching unit 28 is corrected by multiplying the digital audio signal to be D / A converted by the correction coefficient. Thus, the output analog audio signal Aout is prevented from becoming discontinuous before and after the gain switching. The A / D converter 22 to the D / A converter 27 in FIG. In the digital domain, sample data is represented by an n-bit length, and MSB represents a sign of ±. However, the digital signal processing unit 24A expands to the lower side by 5 bits and performs calculation with (n + 5) bit length, and outputs to the gain switching unit 33 with (n + 5) bit length. The gain switching unit 33 performs arithmetic shift calculation. Later, when outputting, it shall round to n bits.
The other components are configured in the same manner as in FIG.

FIG. 10 and FIG. 11 are flowcharts showing control processing at the time of measurement and normal time by the control unit 31A, and the operation of the second embodiment will be described below with reference to these drawings.
Here, it is assumed that correction is performed so as to cancel the relative error of gains at other stages, with the gain of 1/11 of the output gain switching unit 28 as a reference (not corrected).
(1) During measurement When the power of the apparatus is turned on, the control unit 31A switches the switch units 32, 26, and 35 to the measurement mode side, and the reference signal generation unit 25 has an amplitude 2 from 1/2 of the full scale of the digital domain. A digital reference signal composed of a 1 kHz sine wave is generated that is relatively small, and the gain of the output gain switching unit 28 is switched to x1 / 1 of the sixth stage (steps S50 to S52 in FIG. 10). The digital reference signal is generated by repeatedly outputting sample data having a waveform of one sine wave. The digital reference signal is input to the D / A conversion unit 27 via the switch unit 26, converted into an analog reference signal, and then input to the output gain switching unit 28. Here, after the gain is multiplied by 1, it is input as an analog reference signal ST ′ to the A / D conversion unit 22 via the switch unit 32 and converted into a digital signal. Then, the peak value is measured by the level measuring unit 29 and input to the control unit 31A. The control unit 31A stores the input peak value in the correction coefficient storage unit 30A as the reference level SQ, and stores the sixth-stage gain correction coefficient Q6 as 1 (steps S53 to S55).

  Next, the gain of the output gain switching unit 28 is lowered by one step and switched to x1 / 2 (steps S56, S57), and the peak value input from the level measuring unit 29 is used to remove a level that is 1/2 of the reference level SQ. Then, a correction coefficient Q5 of x1 / 2 is calculated and stored in the correction coefficient storage unit 30A (steps S58 and S59). Next, the gain of the output gain switching unit 28 is lowered by one step to be switched to × 1/4, and the peak value input from the level measuring unit 29 is divided by the level of 1/4 of the reference level SQ × 1/4. Correction coefficient Q4 is calculated and stored in the correction coefficient storage unit 30A (NO in step S60, S61, S57 to S59). Thereafter, the same processing is repeated up to × 1/32 in the first stage. Thus, each time the power is turned on, the gain correction coefficients Q1 to Q6 of each stage of the output gain switching unit 28 are determined and stored in the correction coefficient storage unit 30A.

(2) Normal time When the correction coefficient is determined, the control unit 31A switches the switch units 32, 26, and 35 to the normal mode side and switches the gain of the gain switching unit 33 to the first stage x1 and the output gain. The gain of the unit 28 is switched to x1 / 1 in the sixth stage (YES in step S61 in FIG. 10, steps S80 and S81 in FIG. 11). Further, the correction coefficient Q6 = 1 corresponding to x1 / 1 is set in the level correction unit 34 (step S82). The input analog audio signal Ain after volume adjustment is input to the A / D conversion unit 22 via the switch unit 32 and converted into a digital audio signal. Then, a low-frequency component is extracted by the third-order FIR filter in the digital signal processing unit 24A. The digital audio signal after the digital signal processing is set to a gain of x1 (no shift) by the gain switching unit 33, further multiplied by the correction coefficient 1 by the level correction unit 34, and then D / A converted via the switch unit 26. After being input to the unit 27 and converted into an analog audio signal, the output gain switching unit 28 gains × 1/1 and is output to the subsequent stage. The control unit 31A constantly monitors the level (here, the peak value) of the digital audio signal measured by the level measurement unit 29 (step S83), and determines whether or not the low level state continues for a certain period of time. The determination of whether the level has exceeded the level is repeated (repetition of steps S84 and S85).

  If the low level state continues for a certain time or more and the answer is YES in step S84, the current gain switching unit 33 is not yet in the sixth stage, so the gain switching unit 21 is switched (input) to the second stage x2. The sample data is shifted to the left by 1 bit by an arithmetic shift operation), and the dynamic range of the D / A converter 27 is optimized, and the output gain switching unit 28 is switched to x1 / 2 of the fifth stage accordingly (step) S86-S88). Subsequently, the correction coefficient Q5 for the fifth stage is read from the correction coefficient storage section 30A and set in the level correction section 34, and the gain from the level correction section 34 to the output gain switching section 28 is exactly 1 with respect to the sixth stage. / 2 times (step S89). This prevents the analog audio signal Aout output from the output gain switching unit 28 before and after gain switching from becoming discontinuous.

  Thereafter, when the low level state continues for a certain time or more again, it is determined YES in step S84, and the gain switching unit 33 is switched to the third stage x4 (2 bits are obtained by arithmetic shift operation on the input sample data). Left-shifted), the dynamic range of the D / A converter 27 is optimized, and the output gain switching unit 28 is switched to x1 / 4 of the fourth stage accordingly (steps S86 to S88). Subsequently, the correction coefficient Q4 for the fourth stage is read from the correction coefficient storage section 30A and set in the level correction section 34, and the gain from the level correction section 34 to the output gain switching section 28 is exactly 1 with respect to the sixth stage. / 4 times (step S89).

  Thereafter, when the measurement level becomes excessive exceeding a certain level, YES is determined in step S85, and the current gain switching unit 33 is not yet in the first stage, so the gain switching unit 33 is switched to x2 in the second stage ( The input sample data is shifted to the left by 1 bit by an arithmetic shift operation), and is optimized so that the dynamic range of the D / A converter 27 does not become an excessive input. Switching to × 1/2 (steps S90 to S92). Subsequently, the correction coefficient Q5 for the fifth stage is read from the correction coefficient storage section 30A and set in the level correction section 34, and the gain from the level correction section 34 to the output gain switching section 28 is exactly 1 with respect to the sixth stage. / 2 times (steps S93 and S94). This prevents the analog audio signal Aout output from the output gain switching unit 28A before and after gain switching from becoming discontinuous.

According to this embodiment, the gain of the gain switching unit 33 is switched according to the level of the digital audio signal to be D / A converted, and the level is optimized and input to the D / A conversion unit 27. The quantization distortion caused by the conversion can be reduced. Further, a correction coefficient for each step for canceling the gain error of each step of the output gain switching unit 28 is determined using the reference signal, and the D / A converted analog audio signal is applied to the output gain switching unit 28. Since the level correction is performed on the digital audio signal to be D / A converted in accordance with the gain of the output gain switching unit 28, each stage of the output gain switching unit 28 composed of an analog circuit is performed. Even if there is an error in the gain, the gain for the signal can be switched accurately.
Further, since the gain switching unit 33 switches the gain with respect to the output of the signal processing unit 24A by a multiplier factor of 2, the calculation for gain switching can be easily executed by the bit shift operation.

In the second embodiment described above, the correction is made so as to cancel the relative error of the gains of the second to fifth stages when the gain of the sixth stage is used as a reference. However, the absolute error of the gain of each stage is canceled. You may correct | amend as follows. In this case, for example, a digital reference signal having a known amplitude Y is D / A converted and input to the output gain switching unit 28 that has been switched to the third stage × 1/8 gain. When y is measured, the correction factor Q3 is
Y × (1/8) = y × Q3
Q3 = (X / y) × (1/8)
It becomes. Correction coefficients at other stages can be obtained in the same manner.
In the second embodiment described above, the analog audio signal Ain is input to the A / D conversion unit 22 through the switch unit 32 at normal times. However, when the preceding device is a digital output, it is shown in FIG. As described above, the digital audio signal Din may be directly input to the digital signal processing unit 24A. In this case, the analog reference signal ST ′ may be directly input to the A / D converter 22.

A third embodiment of the present invention will be described with reference to FIG. FIG. 13 is a block diagram showing the configuration of the volume control device according to the present invention, and the same components as those in FIG. 9 are given the same reference numerals. The volume adjusting device of the third embodiment attenuates in a 1 dB step from 0 to −5 dB in the digital domain, attenuates in a 6 dB step from 0 to −30 dB in the analog domain, and has a total range of 0 to −35 dB. The volume can be adjusted in 1 dB steps.
In FIG. 13, reference numeral 32 denotes a switch unit that selects and outputs an input analog audio signal Ain to be adjusted in the normal mode, and outputs an analog reference signal output from an output gain switching unit 28B described later in the measurement mode. Select ST 'and output. Reference numeral 24B denotes a digital signal processing unit, which performs arithmetic processing (0 dB, -1 dB, -2 dB, -3 dB, -4 dB, -5 dB) for switching the gain with a step width of 1 dB in the range of 0 dB to -5 dB with respect to the digital audio signal. Is executed). Reference numeral 28B denotes an output gain switching unit capable of switching the gain for the analog audio signal after D / A conversion. The output gain switching unit 28B allows the gain (here, attenuation of 0 dB or less) to be switched between the first to sixth steps. The first stage is set to 0 dB, the second stage is set to -6 dB, the third stage is set to -12 dB, the fourth stage is set to -18 dB, the fifth stage is set to -24 dB, and the sixth stage is set to -30 dB. However, the actual gain at each stage does not match the set value due to the limitation of the selection of the circuit constants of the element parts and the variation of the circuit constants, and there is an error. This error is corrected in the digital domain as will be described later. A level correction unit 34 performs level correction on the digital audio signal to be D / A converted so as to cancel the gain error of the output gain switching unit 28B. Reference numeral 25 is a reference signal generation section for generating a digital reference signal for gain measurement of the output gain switching section 28, and 26 is for selecting a digital audio signal after processing the digital signal input via the level correction section 34 in normal times. A switch unit 30B for selecting and outputting the digital reference signal input from the reference signal generation unit 25 during output and 30B is a storage unit, and stores the correction coefficients at each stage of the output gain switching unit 28B. The storage unit 30B stores a control table as shown in FIG. 14 showing combinations of gains set in the digital signal processing unit 24B and the output gain switching unit 28B with respect to the target volume. A level measuring unit 29 measures the signal level of the digital signal output from the A / D conversion unit 22 in the measurement mode.

Reference numeral 31B denotes a control unit, which applies a reference signal composed of a 1 kHz sine wave to the output gain switching unit 28B and inputs it to the level measurement unit 29 via the switch unit 32 and the A / D conversion unit 22 during measurement. While sequentially switching the gain of the gain switching section 28B from the sixth to the first stage, the signal level measured by the level measuring section 29 at each stage is used to cancel the gain error at each stage of the output gain switching section 28B. A correction coefficient for each stage is determined and stored in the storage unit 30B. In normal times, gain switching control is performed on the digital signal processing unit 24B and the output gain switching unit 28B with reference to a control table in the storage unit 30B based on a target volume input from an external volume changing operation unit (not shown) and output. The gain error of the gain switching unit 28B is corrected to improve the linearity of the output signal level with respect to the change in the target volume. The A / D converter 22 to the D / A converter 27 in FIG. 13 are the digital domain, and the D / A converter 27 to the output gain switching unit 28B are the analog domain. In the digital domain, sample data is represented by an n-bit length, and MSB represents a sign of ±. It is assumed that the digital signal processing unit 24B expands to the lower side by 5 bits, performs an operation with a length of (n + 5) bits, and rounds to n bits when outputting.
Other components are configured in the same manner as in FIG.

FIG. 15 and FIG. 16 are flowcharts showing control processing at the time of measurement and normal time by the control unit 31B, and the operation of the third embodiment will be described below with reference to these drawings.
Here, the gain of 0 dB of the output gain switching unit 28B is used as a reference (not corrected), and correction is performed so as to cancel the relative error of gain at other stages.
(1) During measurement When the power supply of the apparatus is turned on, the control unit 31B switches the switch units 32 and 26 to the measurement mode side, and the reference signal generation unit 25 has an amplitude of about 20% from 1/2 of the full scale of the digital domain. A small digital reference signal composed of a 1 kHz sine wave is generated, and the gain of the output gain switching unit 28B is switched to 0 dB in the sixth stage (steps S200 to S202 in FIG. 15). The digital reference signal is generated by repeatedly outputting sample data having a waveform of one sine wave. The digital reference signal is input to the D / A conversion unit 27 via the switch unit 26, converted into an analog reference signal, and then input to the output gain switching unit 28B. Here, after the gain is set to 0 dB, the analog reference signal ST ′ is input to the A / D conversion unit 22 via the switch unit 32 and converted into a digital signal. Then, the peak value is measured by the level measuring unit 29 and input to the control unit 31B. The control unit 31B stores the input peak value in the storage unit 30B as the reference level SQ ′, and stores the sixth-stage gain correction coefficient Q6 ′ as 1 (steps S203 to S205).

  Next, the gain of the output gain switching unit 28B is lowered by one step and switched to −6 dB (steps S206 and S207), and the peak value input from the level measurement unit 29 is divided by the level that is −6 dB lower than the reference level SQ ′. The -6 dB correction coefficient Q5 ′ is calculated and stored in the storage unit 30B (steps S208 and S209). Next, the gain of the output gain switching unit 28B is lowered by one step to switch to −12 dB, and the peak value input from the level measuring unit 29 is divided by a level that is −12 dB lower than the reference level SQ ′ to obtain a correction coefficient of −12 dB Q4 ′ is calculated and stored in the storage unit 30B (NO in step S2100, S211, S207 to S208). Thereafter, the same processing is repeated up to −30 dB in the first stage. Thus, every time the power is turned on, the gain correction coefficients Q1 'to Q6' of the respective steps of the output gain switching unit 28B are determined and stored in the storage unit 30B.

(2) Normal time When the correction coefficient is determined, the control unit 31B switches the switch units 32 and 26 to the normal mode side (YES in step S210 of FIG. 15, step S220 of FIG. 16). Then, the target volume is input, the gain of the digital signal processing unit 24B and the gain of the output gain switching unit 28B are determined by referring to the control table, and the gain of the digital signal processing unit 24B and the gain of the output gain switching unit 28B are switched. Control is performed (step S222). Subsequently, the correction coefficient for the gain of the output gain switching unit 28B is set in the level correction unit 34 (step S223). The input analog audio signal Ain whose volume is to be adjusted is input to the A / D conversion unit 22 via the switch unit 32 and converted into a digital audio signal. Then, the digital signal processing unit 24B performs variable calculation of the signal level by the gain set by the control unit 31B. The digital audio signal after the digital signal processing is multiplied by a correction coefficient in the level correction unit 34, then input to the D / A conversion unit 27 via the switch unit 26, converted into an analog audio signal, and further output gain. The signal level is varied by the gain set by the control unit 31B in the switching unit 28B.

For example, when the target sound volume is -20 dB, the digital signal processing unit 24B attenuates to -2 dB, the output gain switching unit 28B attenuates to -18 dB, and cancels the gain error of -18 dB of the output gain switching unit 28B. Since Q4 ′ is set in the level correction unit 34, the volume of the output audio signal Aout is accurately −20 dB as a whole.
The control unit 31B monitors the change in the target volume. If there is a change, the control unit 31B determines the gain of the digital signal processing unit 24B and the gain of the output gain switching unit 28B by referring to the control table in the same manner as described above. Switching control of the gain of the digital signal processing unit 24B and the gain of the output gain switching unit 28B is performed (YES in steps S225 and S226, S222 and S223). Subsequently, the correction coefficient for the gain of the output gain switching unit 28B is set in the level correction unit 34 (step S224).

  According to this embodiment, the gain switching of the output gain switching unit 28B can be performed in six steps, so that the configuration is simple, and the digital signal processing unit 24B only attenuates in the range of 0 to 5 dB. Quantization distortion does not increase. Further, a correction coefficient for each step for canceling the gain error of each step of the output gain switching unit 28B is determined using the reference signal, and the analog audio signal D / A converted by the output gain switching unit 28 is determined. Is applied, the level of the D / A conversion target digital audio signal is corrected according to the gain of the output gain switching unit 28B. Even if there is an error in the gain at each stage, the gain for the signal can be switched accurately, and the linearity of the output signal level with respect to the target sound volume becomes good.

In the third embodiment described above, correction is made so as to cancel the relative error of the gains in the second to fifth stages when the gain in the sixth stage is used as a reference. However, the absolute error of the gain in each stage is canceled out. You may correct | amend as follows.
In the third embodiment described above, the analog audio signal Ain is normally input to the A / D converter 22 via the switch unit 32 in the normal case. It may be directly input to the digital signal processing unit 24B. In this case, the analog reference signal ST ′ may be directly input to the A / D converter 22.

A fourth embodiment of the present invention will be described with reference to FIG. FIG. 17 is a block diagram showing the configuration of the band-limiting signal processing apparatus according to the present invention. The same components as those in FIGS. 5 and 9 are given the same reference numerals.
FIG. 17 has a configuration in which FIG. 5 and FIG. 9 are combined. The input gain switching unit is upstream of the A / D conversion unit, the gain switching unit is upstream of the D / A conversion unit, and the output is downstream of the D / A conversion unit. A gain switching unit is provided to correct the gain error of the input gain switching unit and the gain error of the output gain switching unit.
In FIG. 17, 20 is a switch unit for selecting the input analog audio signal Ain in the normal mode and for selecting the reference signal ST in the first measurement mode, and 21 is an input gain switching unit, and the input analog audio after volume adjustment. The gain for the signal Ain is switched to a gain of x1 in the first stage, x2 in the second stage, x4 in the third stage, x8 in the fourth stage, x16 in the fifth stage, and x32 in the sixth stage. Is possible. 32 selects the output of the input gain switching unit 21 in the normal mode and the first measurement mode, and selects the reference signal ST ′ in the second measurement mode. Reference numeral 23C denotes a first level correction unit that performs level correction for canceling the gain error of the input level switching unit 21, and reference numeral 33 denotes a gain switching unit. The first stage is obtained by performing an arithmetic shift operation on the gain of the digital audio signal after the signal processing. The gain can be switched to a gain of x1, the second stage is x2, the third stage is x4, the fourth stage is x8, the fifth stage is x16, and the sixth stage is x32. Reference numeral 28 denotes an output gain switching unit, and the gain for the analog audio signal after D / A conversion is x 1/32 in the first stage, x 1/16 in the second stage, x 1/8 in the third stage, The stage can be switched to a gain of x1 / 4, the fifth stage can be switched to a gain of x1 / 2, and the sixth stage can be switched to a gain of x1 / 1. Reference numeral 34C denotes a second level correction unit that performs level correction on the digital audio signal to be D / A converted so as to cancel the gain error of the output gain switching unit 28.

  Reference numeral 25 denotes a reference signal generator for generating a digital reference signal for gain measurement of the input gain switching unit 21 and the output gain switching unit 28, and 29C measures the level of the digital signal output from the A / D converter 22 from the peak value. The first level measuring unit 36, a second level measuring unit 36 that measures the level of the digital signal input to the D / A converting unit 27 from the peak value, and 30C is a correction coefficient storage unit. The correction coefficient at each stage of the output gain switching unit 28 is stored.

  Reference numeral 31C denotes a control unit. During the first measurement, a reference signal composed of a 1 kHz sine wave is applied to the input gain switching unit 21 through the switch unit 20, and the first signal is input through the switch unit 32 and the A / D conversion unit 22. The input gain switching unit uses the signal level measured by the first level measuring unit 29C at each stage while sequentially switching the gain of the input gain switching unit 21 to the first to sixth stages. The correction coefficient for each stage for canceling the gain error of each stage of 21 is determined and stored in the correction coefficient storage unit 30C. At the time of the second measurement, a reference signal composed of a 1 kHz sine wave is applied to the output gain switching unit 28 and input to the first level measuring unit 29C via the switch unit 32 and the A / D conversion unit 22, and the output gain switching unit. Each of the steps for canceling the gain error at each stage of the output gain switching section 28 using the signal level measured by the first level measuring section 29C at each stage while switching the gain of 28 from the sixth to the first stage in order. The correction coefficient for each stage is determined and stored in the correction coefficient storage unit 30C.

  In normal times, the control unit 31C monitors the signal level of the input analog audio signal input to the A / D conversion unit 22 via the input gain switching unit 21, and a small level state below a predetermined first reference level is detected. In order to optimize the input level of the A / D conversion unit 22 and prevent the quantization distortion from becoming noticeable when it continues for a certain time, the gain of the input gain switching unit 21 is increased by one step, and the output gain switching unit 28 is correspondingly increased. Decrease the gain by one step. Further, when the signal level of the input analog audio signal is monitored, if the signal level exceeds a predetermined second reference level (> first reference level), the input level of the A / D conversion unit 22 is exceeded. In order to prevent clipping due to excessive input, the gain of the input gain switching unit 21 is decreased by one step, and the gain of the output gain switching unit 28 is increased by one step accordingly. Then, the correction coefficient corresponding to the gain of the input gain switching unit 21 is set in the first level correction unit 23C, the digital audio signal output from the A / D conversion unit 22 is multiplied by the correction coefficient, and the input gain switching unit 21 The gain coefficient of the output gain switching section 28 is set in the second level correction section 34C, and the digital audio signal to be D / A converted is multiplied by the correction coefficient to switch the output gain. The gain error of the unit 28 is corrected. The digital signal processor 24A corrects the past sample data to double or 1/2 times by arithmetic shift operation (1-bit left shift or 1-bit right shift), and outputs analog data before and after gain switching. The audio signal Aout is prevented from becoming discontinuous.

Further, at the normal time, the control unit 31C monitors the signal level of the digital audio signal input from the second level correction unit 34C to the D / A conversion unit 27 via the switch unit 26, and falls below a predetermined first reference level. When the small level state continues for a certain time, in order to optimize the input level of the D / A converter 27 and prevent the quantization distortion from becoming noticeable, the gain of the gain switching unit 33 is increased by one step. The gain of the output gain switching unit 28 is lowered by one step. Further, when the signal level of the digital audio signal input to the D / A converter 27 is monitored, when the signal level exceeds a predetermined second reference level (> first reference level), In order to optimize the input level of the D / A conversion unit 27 and prevent clipping due to excessive input, the gain of the gain switching unit 33 is decreased by one step, and the gain of the output gain switching unit 28 is increased by one step accordingly. Then, the correction coefficient corresponding to the gain of the output gain switching unit 28 is set in the second level correction unit 34C, the digital audio signal to be D / A converted is multiplied by the correction coefficient, and the gain error of the output gain switching unit 28 is set. The correction is made so that the output analog audio signal Aout does not become discontinuous before and after the gain switching. The A / D converter 22 to the D / A converter 27 in FIG. In the digital domain, sample data is represented by an n-bit length, and MSB represents a sign of ±. However, the digital signal processing unit 24A expands to the lower side by 5 bits and performs calculation with (n + 5) bit length, and outputs to the gain switching unit 33 with (n + 5) bit length. The gain switching unit 33 performs arithmetic shift calculation. Later, when outputting, it shall round to n bits.
Other components are configured in the same manner as in FIGS.

FIG. 18 to FIG. 21 are flowcharts showing control processing at the time of measurement and normal time by the control unit 31C, and the operation of the above-described fourth embodiment will be described below with reference to these drawings.
Here, the gain of x1 / 1 of the input gain switching unit 21 is used as a reference (not corrected), and correction is performed so as to cancel the relative error of gains at other stages. Similarly, the output gain switching unit It is assumed that 28 × 1/1 gain is used as a reference (no correction is made), and correction is performed so as to cancel the relative error of gain at other stages.
(1) At the time of first measurement When the apparatus is turned on, the control unit 31C switches the switch units 20, 32, and 26 to the first measurement mode side, and the reference signal generation unit 25 has an amplitude W of full scale in the digital domain. A digital reference signal composed of a 1 kHz sine wave smaller than 1/2 is generated, and the gain of the input gain switching unit 21 is switched to the first stage x1 (steps S100 to S102 in FIG. 18). The digital reference signal is input to the D / A conversion unit 27 via the switch unit 26, converted into the analog reference signal ST, and then input to the input gain switching unit 21 via the switch unit 20. Here, after the gain is multiplied by 1, it is converted into a digital signal by the A / D converter 22 via the switch 32. Then, the peak value is measured by the first level measurement unit 29C and input to the control unit 31C. The control unit 31C stores the input peak value as the reference level SP in the correction coefficient storage unit 30C, and stores the first stage gain correction coefficient P1 of the input gain switching unit 21 as 1 (steps S103 to S105). ).

  Next, the gain of the input gain switching unit 21 is increased by one level and switched to x2, and the output amplitude of the reference signal generation unit 25 is halved from that (steps S106 to S108). As a result, the analog reference signal ST is halved in amplitude and input to the input gain switching unit 21, and the gain is set to x2. The control unit 31C calculates the correction coefficient P2 of x2 by dividing the reference level SP by the peak value input from the first level measurement unit 29C, and stores it in the correction coefficient storage unit 30C (steps S109 and S110). Thereafter, the same processing is repeated until the sixth stage x32 (repetition of steps S111, S112, and S107 to S110). Thus, every time the power is turned on, the gain correction coefficients P1 to P6 of each stage of the input gain switching unit 21 are determined and stored in the correction coefficient storage unit 30C.

(2) Second measurement When the correction coefficient of the input gain switching unit 21 is determined, the control unit 31C switches the switch units 20, 32, and 26 to the second measurement mode side, and causes the reference signal generation unit 25 to detect the amplitude Y. Is about 20% smaller than 1/2 of the full scale of the digital domain, and a digital reference signal composed of a 1 kHz sine wave is generated, and the gain of the output gain switching unit 28 is switched to × 1/1 of the sixth stage (FIG. YES in step S111 of 18, and steps S120 to S122 of FIG. 19). The digital reference signal is input to the D / A conversion unit 27 via the switch unit 26 and converted into an analog reference signal. After the gain is set to 1 by the output gain switching unit 28, the digital reference signal is switched as the analog reference signal ST '. The signal is input to the A / D converter 22 via the unit 32 and converted into a digital signal. Then, the peak value is measured by the first level measurement unit 29C and input to the control unit 31C. The control unit 31C stores the input peak value in the correction coefficient storage unit 30C as the reference level SQ, and stores the gain correction coefficient Q6 of the sixth stage of the output gain switching unit 28 as 1 (steps S123 to S125). ).

  Next, the gain of the output gain switching unit 28 is lowered by one step to x1 / 2 (steps S126 and S127). The analog reference signal output from the D / A converter 27 is input to the output gain switching unit 28, and the gain is set to x1 / 2. The control unit 31C calculates the fifth-stage gain correction coefficient Q5 of the output gain switching unit 28 by dividing 1/2 of the reference level SQ by the peak value input from the first level measurement unit 29C, and stores the correction coefficient storage Store in the unit 30C (steps S128, S129). Thereafter, the same processing is repeated up to × 1/32 in the first stage (repetition of steps S130, S131, S127 to S129). Thus, each time the power is turned on, the gain correction coefficients Q1 to Q6 of each stage of the output gain switching section 28 are determined and stored in the correction coefficient storage section 30C.

(3) Normal time When the correction coefficient is determined, the control unit 31C switches the switch units 20, 32, and 26 to the normal mode side, and changes the gain of the input gain switching unit 21 to the first stage × 1, gain switching. The gain of the unit is switched to × 1, and the gain of the output gain switching unit 28 is switched to the sixth stage of × 1/1 (YES in step S130 of FIG. 19, steps S140 and S141 of FIG. 20). Further, the correction coefficient P1 = 1 corresponding to x1 of the input gain switching unit 21 is set in the first level correction unit 23C, and the correction coefficient Q6 = corresponding to x1 of the output gain switching unit 28 is set in the second level correction unit 34C. 1 is set (step S142). The input analog audio signal Ain after the volume adjustment is input to the input gain switching unit 21 via the switch unit 20, and after the gain is set to × 1, it is converted into a digital audio signal by the A / D conversion unit 22 via the switch unit 32. Is done. Then, after the first level correction unit 23C multiplies the correction coefficient 1, the digital signal processing unit 24A extracts a low frequency component by a third-order FIR filter. Then, the gain switching unit 33 multiplies the gain by 1, and the second level correction unit 34C multiplies the correction coefficient 1. Then, the gain is input to the D / A conversion unit 27 via the switch unit 26 and converted into an analog audio signal. The Then, the output gain switching unit 28 decrements the gain by 1/1 and outputs it to the subsequent stage. The control unit 31C constantly monitors the level (here, peak value) of the digital audio signal measured by the first level measurement unit 29C and the second level measurement unit 36 (step S143, step S160 in FIG. 21). The determination as to whether or not the low level state continues for a certain time or more and the determination as to whether or not the predetermined level exceeds the predetermined level are repeated (steps S144 and S145 in FIG. 20 and steps S161 and S162 in FIG. 21).

  As a result of the measurement by the first level measurement unit 29C, when the low level state continues for a certain time or more and the answer is YES in step S144, the current input gain switching unit 21 is not yet in the sixth stage, but the output gain switching unit. Since 28 is not yet in the first stage, the input gain switching unit 21 is switched to × 2 in the second stage to optimize the dynamic range of the A / D conversion unit 22, and the output gain switching unit 28 is correspondingly changed in the fifth stage. Switching to × 1/2 (steps S146 to S148). Subsequently, the correction coefficient P2 for the second stage of the input gain switching unit 21 is read from the correction coefficient storage unit 30C and set in the first level correction unit 23C, and from the input gain switching unit 21 to the first level correction unit 23C. The gain is accurately doubled with respect to the first stage of the input gain switching unit 21, and the correction coefficient Q5 for the fifth stage of the output gain switching unit 28 is read from the correction coefficient storage unit 30C to obtain the second level correction unit. 34C so that the gain from the second level correction unit 34C to the output gain switching unit 28 is exactly ½ times the sixth stage of the output gain switching unit 28. Further, the past sample data of the FIR filter is shifted by 1 bit to the digital signal processing unit 24A by an arithmetic shift operation to double the size (steps S149 and S150). This prevents the analog audio signal Aout output from the output gain switching unit 28 from becoming discontinuous before and after the gain switching.

  After that, when the signal level measured by the first level measurement unit 29C continues to be in a low level state for a certain time or more, it is determined YES in step S144, and the input gain switching unit 21 is switched to x4 in the third stage. The dynamic range of the A / D conversion unit 22 is optimized, and the output gain switching unit 28 is switched to x¼ of the fourth stage accordingly (steps S146 to S148). Subsequently, the third-stage correction coefficient P3 of the input gain switching unit 21 is read from the correction coefficient storage unit 30C and set in the first level correction unit 23C, and from the input gain switching unit 21 to the first level correction unit 23C. The gain is accurately quadrupled with respect to the first stage, the fourth stage correction coefficient Q4 of the output gain switching section 28 is read from the correction coefficient storage section 30C, set in the second level correction section 34C, The gain from the two-level correction unit 34C to the output gain switching unit 28 is accurately set to 1/4 times the sixth level of the output gain switching unit 28. Then, the past sample data of the FIR filter is shifted by 1 bit to the digital signal processing unit 24A by an arithmetic shift operation to double the size (steps S149 and S150).

  Thereafter, when the measurement level at the first level measurement unit 29C becomes excessive exceeding a predetermined constant level, YES is determined in step S145, and the current input gain switching unit 21 is not yet in the first stage, but the output gain is switched. Since the unit 28 is not yet in the sixth stage, the input gain switching unit 21 is returned to × 2 in the second stage and optimized so as not to become an excessive input with respect to the dynamic range of the A / D conversion unit 22, and the output gain is switched accordingly. The unit 28 is switched to x1 / 2 of the fifth stage (steps S151 to S153). Subsequently, the correction coefficient P2 for the second stage of the input gain switching unit 21 is read from the correction coefficient storage unit 30C and set in the first level correction unit 23C, and the fifth of the output gain switching unit 28 from the correction coefficient storage unit 30C. The correction coefficient Q5 for the stage is read and set in the second level correction unit 34C, and the past sample data of the FIR filter is shifted to the right by one bit by arithmetic shift operation with respect to the digital signal processing unit 24A, and is ½ times larger. (Steps S154 and S155). This prevents the analog audio signal Aout output from the output gain switching unit 28 before and after gain switching from becoming discontinuous.

  After that, as a result of the measurement of the second level measurement unit 36, the low level state continues for a certain time or more, and when YES is obtained in step S161 in FIG. 21, the control unit 31C has not yet reached the sixth stage in the current gain switching unit 33. Since the output gain switching unit 28 is not yet in the first stage, the gain switching unit 33 is switched to × 2 in the second stage to optimize the dynamic range of the D / A conversion unit 27, and the output gain switching unit 28 is correspondingly changed. Decrease by one level and switch to × 1/4 of the fourth level (steps S163 to S165). Subsequently, the correction coefficient Q4 for the fourth stage of the output gain switching unit 28 is read from the correction coefficient storage unit 30C and set in the second level correction unit 34C, and from the second level correction unit 34C to the output gain switching unit 28. The gain is accurately set to 1/4 times the sixth level of the output gain switching unit 28 (step S166). This prevents the analog audio signal Aout output from the output gain switching unit 28 from becoming discontinuous before and after the gain switching.

  Thereafter, when the signal level measured by the second level measurement unit 36 continues to be in a small level state for a certain time or more, it is determined YES in step S161, and the gain switching unit 33 is switched to x4 in the third stage. The dynamic range of the / A conversion unit 27 is optimized, and the output gain switching unit 28 is lowered by one step and switched to x1 / 8 of the third step (steps S163 to S165). Subsequently, the third-stage correction coefficient Q3 of the output gain switching section 28 is read from the correction coefficient storage section 30C and set in the second level correction section 34C, and from the second level correction section 34C to the output gain switching section 28. The gain is set to exactly ８ times the sixth stage of the output gain switching unit 28. (Step S166).

  After that, when the measurement level in the second level measurement unit 36 becomes excessive exceeding a predetermined constant level, it is determined YES in step S162, and the current gain switching unit 33 is not yet in the first stage, but the output gain switching unit. 28 is not yet in the sixth stage, the gain switching unit 33 is returned to x2 in the second stage to optimize the dynamic range of the D / A conversion unit 27 so that it does not become an excessive input, and the output gain switching unit 28 correspondingly. Is switched to × 1/4 of the fourth stage (steps S167 to S169). Subsequently, the third-stage correction coefficient Q3 of the output gain switching section 28 is read from the correction coefficient storage section 30C and set in the second level correction section 34C (step S170). This prevents the analog audio signal Aout output from the output gain switching unit 28 before and after gain switching from becoming discontinuous.

According to this embodiment, the gain of the input gain switching unit 21 is switched according to the level of the input analog audio signal to be A / D converted, and the level is optimized and input to the A / D conversion unit 22. Quantization distortion caused by the / D conversion can be reduced, the gain of the gain switching unit 33 is switched according to the level of the digital audio signal to be D / A converted, and the level is optimized to the D / A conversion unit 27. By inputting, quantization distortion caused by D / A conversion can be reduced.
Further, a correction coefficient for each step for canceling the gain error of each step of the input gain switching unit 21 is determined using the reference signal, and an analog audio signal to be A / D converted is applied to the input gain switching unit 21. Since the level correction is performed on the digital audio signal output from the A / D conversion unit 22 in accordance with the gain of the input gain switching unit 21, the input gain switching unit configured by an analog circuit is used. Even if there is an error in the gain at each stage of 21, the gain for the signal can be switched accurately. Similarly, a correction coefficient for each step for canceling the gain error of each step of the output gain switching unit 28 is determined using the reference signal, and the analog audio signal D / A converted by the output gain switching unit 28 Since the level correction is performed on the digital audio signal to be D / A converted in accordance with the gain of the output gain switching unit 28 when it is applied, each of the output gain switching unit 28 configured by an analog circuit is performed. Even if there is an error in the gain of the stage, the gain for the signal can be switched accurately.
Further, when digital signal processing is performed on the digital audio signal after A / D conversion, it is necessary to switch the level of past sample data held in the digital signal processing unit 24A in conjunction with the gain switching of the input gain switching unit 21. However, since the gain setting values in a plurality of stages of the input gain switching unit 21 are switched by a multiplier of 2, the level of the past sample data can be easily switched by arithmetic shift calculation (right shift or left shift). be able to.
Further, since the gain switching unit 33 switches the gain with respect to the output of the signal processing unit 24A by a multiplier factor of 2, the calculation for gain switching can be easily executed by the bit shift operation.

  In the fourth embodiment described above, the input gain switching unit is corrected so as to cancel the relative error of the second to sixth stage gains when the first stage gain is used as a reference. The absolute error may be corrected so as to cancel out. Similarly, the output gain switching unit has been corrected so as to cancel the relative error of the gains of the second stage to the fifth stage when the gain of the sixth stage is used as a reference, but so as to cancel the absolute error of the gain of each stage. It may be corrected.

  The present invention is similarly applicable to A / D conversion or D / A conversion of other types of signals such as video signals in addition to audio signals.

It is a block diagram which shows an example of the conventional signal processing system. It is a circuit diagram which shows an example of an analog gain switching circuit. It is explanatory drawing of the circuit characteristic of FIG. It is a block diagram which shows an example of the conventional volume control circuit. 1 is a block diagram of a signal processing device for band limitation according to a first embodiment of the present invention (first embodiment). FIG. It is explanatory drawing which shows the signal flow of the digital signal processing part in FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a block diagram of the signal processing apparatus for band limitation which concerns on 2nd Example of this invention (Example 2). It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a block diagram which shows the structure of the modification of FIG. It is a block diagram of the volume control apparatus which concerns on 3rd Example of this invention (Example 3). It is explanatory drawing of the control table memorize | stored in the memory | storage part of FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG. FIG. 10 is a block diagram of a signal processing device for band limitation according to a fourth embodiment of the present invention (fourth embodiment). It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG. It is a flowchart which shows the control processing of the control part of FIG.

Explanation of symbols

20, 26, 32, 35 Switch unit 21 Input gain switching unit 22 A / D conversion unit 23, 34 Level correction unit 23C First level correction unit 24, 24A, 24B Digital signal processing unit 25 Reference signal generation unit 27 D / A Conversion unit 28, 28B Output gain switching unit 29 Level measurement unit 29C First level measurement unit 30, 30A, 30C Correction coefficient storage unit 30B Storage unit 31, 31A, 31B, 31C Control unit 33 Gain switching unit 34C Second level correction unit 36 Second Level Measurement Unit

Claims (6)

A / D conversion means for A / D converting an input analog signal, and input gain switching means provided in a preceding stage of the A / D conversion means and capable of switching to a predetermined set value gain in a plurality of stages for the input analog signal In an A / D conversion device comprising:
Reference signal generating means for generating a reference signal for gain measurement of the input gain switching means;
A selection means for selecting an input analog signal for normal input and inputting it to the input gain switching means, and for selecting a reference signal for input to the input gain switching means for measurement,
Level measurement that measures the output level of the A / D conversion means when the gain of the input gain switching means is switched to each stage when the reference signal is applied to the input gain switching means via the selection means during measurement. Means,
Correction coefficient determining means for determining a correction coefficient for each stage for canceling the gain error of each stage of the input gain switching means based on the level measured by the level measuring means;
Normally, when an input analog signal is applied to the input gain switching means via the selection means, the correction coefficient is determined for the digital signal output from the A / D conversion means according to the gain of the input gain switching means. Correction means for performing level correction based on the correction coefficient determined by the means;
An A / D conversion device comprising: The gain setting value of the multiple steps of the input gain switching means is a value that is a multiplier of 2 or a multiplier of 1/2.
The A / D converter according to claim 1. D / A conversion means for D / A converting and outputting a digital signal, and output gain switching provided at a subsequent stage of the D / A conversion means and capable of switching to a predetermined set value gain for the output analog signal And a D / A conversion device comprising:
Reference signal generating means for generating a reference signal for gain measurement of the output gain switching means;
A selection means for selecting an output analog signal for normal operation and inputting it to the output gain switching means, and for selecting a reference signal for input to the output gain switching means for measurement,
A / D conversion means to which an analog signal output from the output gain switching means is input during measurement;
Level measurement that measures the output level of the A / D conversion means when the gain of the output gain switching means is switched to each stage when the reference signal is applied to the output gain switching means via the selection means during measurement. Means,
Correction coefficient determining means for determining a correction coefficient for each stage for canceling the gain error of each stage of the output gain switching means based on the output level measured by the level measuring means;
Normally, when the output of the D / A conversion means is applied to the output gain switching means via the selection means, the digital signal input to the D / A conversion means according to the gain of the output gain switching means Correction means for performing level correction based on the correction coefficient determined by the correction coefficient determination means;
A D / A conversion device comprising: The gain setting value of the plurality of stages of the output gain switching means is a value that is a multiplier of 2 or a multiplier of 1/2.
The D / A converter according to claim 3. A / D conversion means for A / D converting the input analog signal, signal processing means for performing signal processing on the input digital signal output from the A / D conversion means, and D / A for the output digital signal of the signal processing means D / A conversion means for converting and outputting; input gain switching means provided in the preceding stage of the A / D conversion means and capable of switching to a predetermined set value gain in a plurality of stages for the input analog signal; and D / A An output gain switching means provided at a subsequent stage of the conversion means and capable of switching to a predetermined set value gain of a plurality of stages for the output analog signal;
Reference signal generating means for generating a reference signal for gain measurement of the input gain switching means and the output gain switching means;
During normal measurement, the input analog signal is selected and input to the input gain switching means. During the first measurement, the reference signal is selected and input to the input gain switching means. During the second measurement, the output of the output gain switching means is output. First selection means for selecting and inputting to the input gain switching means;
An output analog signal is selected and input to the output gain switching means at normal time, and a reference signal is selected and input to the output gain switching means at the time of the second measurement;
During the first measurement, when the reference signal is applied to the input gain switching means via the first selection means, the output of the A / D conversion means when the gain of the input gain switching means is switched to each stage. When the level is measured and the reference signal is applied to the input gain switching means via the second selection means, the output gain switching means, and the first selection means during the second measurement, the output gain switching means Level measuring means for measuring the output level of the A / D conversion means when the gain is switched to each stage;
Based on the level measured by the level measuring means, the correction coefficient for each stage for canceling the gain error of each stage of the input gain switching means and for each stage for canceling the gain error of each stage of the output gain switching means Correction coefficient determining means for determining the correction coefficient of
Normally, when the input analog signal is applied to the input gain switching means via the first selection means and the output analog signal is applied to the output gain switching means via the second selection means, the input gain Correction for performing level correction on the digital signal between the A / D conversion means and the D / A conversion means according to the gain of the switching means and the gain of the output gain switching means based on the correction coefficient determined by the correction coefficient determination means Means,
A signal processing apparatus comprising: The multi-stage gain setting value of the input gain switching means is a multiple of 2 or a multiplier value of 1/2, and the multi-stage gain setting value of the output gain switching means is a multiplier of 2 or 1/2. A multiplier value,
The signal processing device according to claim 5.

Images