JP2008010968A - Noise canceller circuit and noise rejection method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a noise canceller circuit which can remove the impact of external noise occurring over a long time with high density in time without being affected by variation in electric field strength of an incoming radio wave. <P>SOLUTION: An external noise component N resulting from external noise is extracted from a received signal Scv converted into an intermediate frequency and then charged/discharged through a low-pass filter 3 to generate a charge/discharge signal R. When the electric field strength of an incoming radio wave is strong, the time constant τa of the low-pass filter 3 is set at a small value during charging and it is switched to a larger value as the electric field strength of the incoming radio wave becomes weaker. A threshold signal Rth is generated by regulating the level of the charge/discharge signal R by an amplifier 4, a comparator 5 compares the external noise component N with the level of the threshold signal Rth and detects the noise generation period T of a high level external noise component N, and an interpolation circuit 7 interpolates a noise component in a demodulated signal Smd outputted from the detector in a broadcast receiver 200 in that noise generation period T. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a noise canceller circuit provided in a vehicle-mounted broadcast receiver or the like, and relates to a noise canceller circuit and a noise removal method for removing the influence of external noise generated in an automobile or the like and mixed in a received signal.

  In vehicle-mounted broadcast receivers that receive radio broadcasts, if pulsed external noise such as ignition noise generated from an automobile is mixed into a received signal via a receiving antenna, etc., noise is detected from a detector that detects the received signal. A demodulated signal including the component is output, and an unpleasant noise is generated in the reproduced sound reproduced by a speaker or the like. Therefore, a noise canceller circuit is provided to remove the influence of external noise.

Conventionally, a noise canceller circuit shown in FIGS. 1A and 2A is known.
The noise canceller circuit of FIG. 1A includes a noise detection circuit and an interpolation circuit, and is provided in, for example, a superheterodyne broadcast receiver that receives FM broadcast or AM broadcast.

  When a reception signal (hereinafter referred to as “intermediate frequency reception signal”) Scv that has been converted to an intermediate frequency is output from the frequency converter in the broadcast receiver, the noise detection circuit outputs the reception signal Scv to the intermediate frequency reception signal Scv. Performs absolute value processing by full-wave rectification, and performs AM detection (envelope detection) while performing automatic gain control (AGC: Auto Gain Controll) taking into account the dynamic range within the power supply voltage range. The manifested envelope signal Senv is generated, and the envelope signal Senv is compared with a predetermined threshold value (fixed value) Vth to detect the noise generation period T of the external noise component N having a level higher than the threshold value Vth. .

  The interpolation circuit performs a mute process (attenuation or cutoff process) on a noise component (noise component caused by external noise) generated in the demodulated signal Smd output from the detector in the broadcast receiver during the noise generation period T. By performing this interpolation processing, a demodulated signal Sx from which the influence of external noise has been removed is output.

  By the way, in the noise canceller circuit of FIG. 1A, individual external noise components N generated in the envelope signal Senv are detected one by one by comparison with a predetermined threshold value Vth, and from the demodulated signal Smd in each detected noise generation period T. Since the noise components are removed one by one, it is possible to effectively remove the adverse effects when the external noise is mixed once.

  However, when external noise generated over a long period of time with high density is mixed, the noise generation period T of each external noise component N is set as shown schematically in FIGS. If the noise components D in the demodulated signal Smd are detected and removed one by one, the demodulated signal Sx has a waveform that is cut into strips. Then, the signal component left in the strip shape of the demodulated signal Sx becomes a discontinuous sound and is reproduced by a speaker or the like, so that there is a problem that it is difficult for the listener to listen and uncomfortable.

  In order to improve such a problem, a noise canceller circuit shown in FIG. 2 has been developed (see Patent Document 1).

  2A includes an AM detector, a high-pass filter, a low-pass filter, an amplifier, a comparator, a gate signal generation circuit, and an interpolation circuit.

  Here, in the AM detector, the noise detection circuit performs absolute value processing by full-wave rectification on the intermediate frequency reception signal Scv output from the frequency converter, and the power supply voltage While performing automatic gain control (AGC) considering the dynamic range within the range, AM detection (envelope detection) makes the external noise component N generated in the intermediate frequency received signal Scv due to external noise manifest. The envelope signal Senv is generated.

  The high-pass filter has a low-frequency cutoff frequency set to a frequency between the center frequency (carrier frequency) of the IF filter provided in the broadcast receiver and the high-frequency cutoff frequency, and the low-frequency cutoff from the envelope signal Senv. A component in a high frequency range is extracted from the frequency and supplied to a low-pass filter.

  That is, if the center frequency of the IF filter (the bandpass filter for extracting the IF signal that is the modulated signal) is fc and the high-frequency cutoff frequency is fh, the low-frequency cutoff frequency fL of the high-pass filter is The frequency is set to a frequency between fc and fh, and a component in a high frequency region is extracted from the envelope signal Senv from the low-frequency cutoff frequency fL and supplied to the low-pass filter. Thus, since the low-pass cutoff frequency fL of the high-pass filter is set, the external noise component N in the audible frequency range included in the envelope signal Senv can be extracted.

  The low-pass filter is formed of a variable cutoff frequency (variable time constant) low-pass filter that charges the external noise component N at high speed and discharges it at low speed. That is, when the level of the external noise component N starts to rise, the low-pass filter starts high-speed charging based on a small time constant (fixed time constant) τa, and the external noise component N drops from the peak level to charge voltage. If it falls below, it will discharge slowly based on a large time constant (fixed time constant) τr. The charge-discharge signal R generated by performing high-speed charge and low-speed discharge on the external noise component N is set to the amplifier with the period during high-speed charging as the “attack period” and the period during low-speed discharge as the “recovery period”. Supply.

  Here, as described above, since the low-frequency cutoff frequency fL of the high-pass filter is set to a frequency between the frequencies fc and fh, the low-pass filter includes not only the external noise component N but also the audible frequency range. A modulation signal component is also input. Therefore, in order to be able to charge the external noise component N, the high-frequency cutoff frequency fa of the low-pass filter in the attack period is higher than the maximum frequency of the modulated signal component and lower than the maximum frequency of the external noise component N. And the value of the small time constant τa is set in advance based on the relationship of the following equation (1).

  The amplifier is provided to finely adjust the level of the charge / discharge signal R, and the level is adjusted to the high level side (large level side) by amplifying the charge / discharge signal R with a predetermined (fixed) amplification factor g. Then, the level-adjusted signal is output as the threshold signal Rth.

  The comparator compares the external noise component N with the threshold signal Rth, detects the external noise component N having a level greater than the threshold signal Rth, and outputs a signal Sa indicating the noise generation period T of the external noise component N. .

  The gate signal generation circuit generates a gate signal Sg indicating the noise generation period T by performing waveform shaping or the like on the signal Sa described above, and supplies the gate signal Sg to the interpolation circuit.

  Then, the interpolation circuit performs interpolation such as mute processing (attenuation or cutoff processing) on the noise component (noise component caused by external noise) generated in the demodulated signal Smd in the noise generation period T specified by the gate signal Sg. By performing the processing, a demodulated signal Sx from which the influence of external noise is removed is output.

  As described above, in the noise canceller circuit of FIG. 2A, when external noise generated over a long period of time is mixed as shown in FIG. A charge / discharge signal R is generated by performing charge / discharge processing on the generated external noise component N, and a threshold signal Rth is generated by amplifying the charge / discharge signal with an amplification factor g.

  For this reason, when the external noise component N is generated over a long period of time with a high density, the threshold signal Rth becomes a sawtooth waveform, and between the high level external noise component N and the small level external noise component N. Is held at an intermediate level. Then, since the threshold signal Rth held at the intermediate level and the external noise component N are compared, the low level external noise component N is not detected, and only the high level external noise component N is detected and the high level of the external noise component N is detected. Only the noise generation period T of the level external noise component N is detected.

  As shown in FIGS. 2C and 2D, only the noise component corresponding to the detected noise generation period T among the noise component D generated in the demodulated signal Smd due to the external noise is the demodulated signal Smd. Will be removed. For this reason, a small-level noise component D remains in the demodulated signal Sx output from the interpolation circuit, but it is possible to reduce the generation of strip-shaped signal components, which is useful for the listener. It prevents the choppy choppy sound from being played.

  That is, when external noise generated over a long period of time with high density is mixed, the demodulated signal Sx has a lower level than the demodulated signal Sx that generates many strip-shaped cut signal components. Since the noise component D of the noise does not give the listener a sense of incongruity, the noise canceller circuit is adapted to the actual situation.

  Further, since the high level external noise component N can be detected by comparison with the threshold signal Rth, the high level noise component D generated in the demodulated signal Smd is removed, giving the listener a sense of incongruity. It is possible to avoid it.

JP 2005-197813 A

  The conventional noise canceller circuit shown in FIG. 2 (a) exhibits an excellent effect as described above, but its operating characteristics change depending on the electric field strength of the incoming radio wave, and the level of the threshold signal Rth varies. In some cases, the influence of external noise cannot be sufficiently removed.

  A specific example will be described with reference to the characteristic diagram of FIG. 3. First, in the AM detector shown in FIG. 2A, as described above, the automatic gain control (AGC) is performed on the intermediate frequency received signal Scv. AM detection is performed while performing the above, and the envelope signal Senv is generated. For this reason, an envelope signal Senv subjected to automatic gain control is generated. In general, since a broadcast receiver performs signal processing while performing automatic gain control, an envelope signal Senv with automatic gain control is generated even when the automatic gain control is not performed by the AM detector.

  For this reason, when the electric field strength of the incoming radio wave is weak, as shown in FIG. 3, the level of the modulated signal component of the envelope signal Senv decreases, while the level of broadband noise such as white noise distributed over a wide frequency range. Rises. When the low-pass filter performs the above-described charging / discharging process on the envelope signal Senv whose broadband noise level is rising, the base level BLV of the charging / discharging signal R shown in FIG. It will be.

  Furthermore, when the external noise generated over a long period of time is high when the electric field strength of the incoming radio wave is weak, the low-pass filter is used in the attack period. When charging the component N, the external noise component N is started to be charged with the rising base level BLV as a reference level. For this reason, the charge / discharge signal R becomes higher by the rising base level BLV, and the level is further adjusted to a higher level by the amplifier set to the amplification factor g. The level of Rth also becomes higher, and the level of the threshold signal Rth may be higher than that of the high level external noise component N.

  When the comparator compares the threshold signal Rth adjusted to a high level with the external noise component N, the high level external noise component N cannot be detected, and the high level noise component generated in the demodulated signal Smd. May not be able to be removed accurately.

  On the other hand, when the electric field strength of the incoming radio wave is strong, the level of the modulated signal component of the envelope signal Senv increases due to the above-described automatic gain control, and the level of broadband noise such as white noise decreases. When the above-described charge / discharge process is performed, the base level BLV of the charge / discharge signal R shown in FIG. 2B is maintained at the original low level.

  For this reason, when the external noise generated over a long period of time is high when the electric field strength of the incoming radio wave is strong, the low-pass filter will cause the external noise component during the attack period. When N is charged, the external noise component N starts to be charged with reference to the original low base level BLV described above. Therefore, the charge / discharge signal R is maintained at the original level, and the threshold signal Rth is also maintained at the original level. For this reason, when the comparator compares the threshold signal Rth of the main body with the external noise component N, the high level external noise component N can be detected, and the high level noise component generated in the demodulated signal Smd can be accurately detected. Can be removed.

  As described above, the noise canceller circuit of FIG. 2A has a problem that the operation characteristics change (that is, the noise detection sensitivity of the external noise changes) when the electric field strength of the radio wave coming from the broadcasting station changes. Specifically, when the electric field strength of the incoming radio wave is strong (when the reception condition of the broadcast receiver is good), the influence of external noise generated over a long period of time is accurately removed. However, when the electric field strength of the incoming radio wave becomes weak (when the reception condition of the broadcast receiver deteriorates), the influence of high-level external noise among the external noise generated over a long period of time is high. There was a problem that it was difficult to remove.

  The present invention has been made in view of these conventional problems, and eliminates the influence of external noise generated over a long period of time with high density without being affected by changes in the electric field strength of incoming radio waves. An object of the present invention is to provide a noise canceller circuit and a noise removal method that can be used.

  According to the first aspect of the present invention, the demodulation output from the detector provided in the receiver due to the pulsed external noise generated over a long period of time and mixed in the receiver. A noise canceller circuit for removing noise components generated in a signal, an interpolation means for performing interpolation processing on the demodulated signal output from the detector, and an intermediate frequency received signal generated by a frequency converter in the receiver AM detection means and a high-pass filter for generating a signal including an external noise component generated in the intermediate frequency reception signal due to the external noise by envelope detection, and the external noise component with a predetermined attack time constant A low-pass filter that generates a charge / discharge signal by charging at a low speed with a predetermined recovery time constant, and a threshold value by adjusting the level of the charge / discharge signal An amplifying means for generating a signal, a comparing means for comparing the threshold signal and the external noise component, detecting a noise generation period of the external noise component having a level higher than the threshold signal, and a gate signal indicating the noise generation period. By supplying to the interpolation means, a gate signal generation means for interpolating the noise generated in the demodulated signal due to the external noise, and detecting the electric field strength of the radio wave coming from the broadcasting station to the receiver When the electric field strength detected by the detecting means and the detecting means is strong, the attack time constant of the low-pass filter is set to a small value, and the attack time constant of the low-pass filter becomes larger as the electric field strength becomes weaker. Time constant switching means for setting the value to charge the external noise component.

  A preferred embodiment of the present invention will be described with reference to FIGS. 4A, 4B, and 4C are block diagrams showing the configuration of the noise canceller circuit according to the present embodiment, and FIGS. 5A, 5B, and 5C are the functions of the noise canceller circuit according to the present embodiment. It is explanatory drawing for demonstrating operation | movement.

  4A, the noise canceller circuit 100 is provided in a vehicle-mounted broadcast receiver (for example, a superheterodyne broadcast receiver) 200 in order to remove the influence of pulsed external noise such as ignition noise. ing.

  Here, since the superheterodyne broadcast receiver has been described with reference to FIG. 2 (a), the details are omitted, but the channel selection circuit converts the reception output of the reception antenna into an RF reception signal, and the frequency The converter frequency-converts the RF reception signal into an intermediate frequency reception signal Scv, and an IF filter having a predetermined pass bandwidth extracts an IF signal as a desired wave from the intermediate frequency reception signal Scv. The IF signal is detected (AM detection when AM broadcasting is being received, and FM detection when FM broadcasting is being received) to generate a demodulated signal Smd, which is output to an interpolation circuit 7 described later.

  For convenience of explanation, although not shown in FIG. 4A, the noise canceller circuit 100 receives the intermediate frequency from the processing time required for the broadcast receiver 200 to generate the demodulated signal Smd from the intermediate frequency received signal Scv. Since the processing time required to generate a gate signal Sg, which will be described later, from the signal Scv is longer, the detector, the interpolation circuit, and the demodulator Smd and the gate signal Sg are input to the interpolation circuit 7 at the same timing. A delay circuit for delaying the demodulated signal Smd by a predetermined time and supplying the demodulated signal Smd to the interpolation circuit is provided.

  The noise canceller circuit 100 includes an AM detector 1, a high pass filter 2, a low pass filter 3, an amplifier 4, a comparator 5, a gate signal generation circuit 6, an interpolation circuit 7, an electric field strength detection unit 8, and a time constant switching unit 9. Configured.

  Here, the AM detector 1, the high-pass filter 2, the amplifier 4, the comparator 5, the gate signal generation circuit 6 and the interpolation circuit 7 are the same as the AM detector, high-pass filter, amplifier, comparator shown in FIG. It has the same configuration as the gate signal generation circuit and the interpolation circuit, and further exhibits the same function.

  That is, the AM detector 1 performs absolute value processing by full-wave rectification on the intermediate frequency received signal Scv output from the frequency converter in the broadcast receiver 200, and considers the dynamic range within the power supply voltage range. By performing AM detection (envelope detection) while performing automatic gain control (AGC), an envelope signal Senv in which an external noise component generated in the intermediate frequency reception signal Scv due to the external noise is revealed is generated.

  The high-pass filter 2 has a low-frequency cutoff frequency fL set to a frequency between the center frequency (carrier frequency) fc of the IF filter provided in the broadcast receiver 200 and the high-frequency cutoff frequency fh, and is an envelope signal. A component in the high frequency range is extracted from the low cutoff frequency fL from Senv and supplied to the low pass filter. Therefore, the external noise component N in the audible frequency range included in the envelope signal Senv can be extracted. Then, the extracted external noise component N is supplied to the comparator 5 and the time constant switching unit 9.

  The amplifier 4 is provided to finely adjust the level of a charge / discharge signal R, which will be described later, and amplifies the charge / discharge signal R with a predetermined (fixed) amplification factor g that has been finely adjusted to increase the level of the charge / discharge signal R The level is adjusted, and the signal after the level adjustment is output as the threshold signal Rth.

  The comparator 5 compares the external noise component N output from the high pass filter 2 with the threshold signal Rth, detects the external noise component N having a level higher than the threshold signal Rth, and the noise generation period of the external noise component N A signal Sa indicating T is output.

  The gate signal generation circuit 6 generates a gate signal Sg indicating the noise generation period T by performing waveform shaping or the like on the signal Sa described above, and supplies the gate signal Sg to the interpolation circuit 7.

  The interpolation circuit 7 performs an interpolation process such as a mute process (attenuation or cutoff process) on a noise component (noise component caused by external noise) generated in the demodulated signal Smd in the noise generation period T specified by the gate signal Sg. As a result, the demodulated signal Sx from which the influence of external noise has been removed is output.

  Next, the configuration of the electric field strength detection unit 8, the time constant switching unit 9, and the low pass filter 3 will be described.

  The electric field strength detection unit 8 receives the intermediate frequency reception signal Scv, and extracts a signal component substantially the same as the IF signal by a band pass filter (not shown) having a pass bandwidth substantially the same as the pass bandwidth of the IF filter. Then, the electric field strength of the incoming radio wave is detected by measuring the effective power value of the extracted signal component. Then, the measured effective power value is supplied to the time constant switching unit 9 as the detection signal Se indicating the level of the electric field strength. The electric field strength detector 8 of the present embodiment outputs a voltage proportional to the measured effective power value as a detection signal Se indicating the level of the electric field strength.

  The time constant switching unit 9 receives the output of the high-pass filter 2, the detection signal Se from the electric field intensity detection unit 8, and the charge / discharge signal R output from the low-pass filter 3 that performs charge / discharge processing. Then, a period during which the level of the external noise component N included in the output of the high-pass filter 2 is higher than the charge / discharge signal R (ie, “attack period”) is detected, and a signal indicating the detected attack period (hereinafter referred to as “attack period”). Sar (referred to as “attack period signal”) is supplied to the low-pass filter 3. Further, the voltage level of the detection signal Se is detected, and a time constant switching signal SEL for automatically variably adjusting the high-frequency cutoff frequency of the low-pass filter 3 according to the voltage level is generated and supplied to the low-pass filter 3. To do.

  Here, a specific configuration example of the time constant switching unit 9 will be described with reference to FIG.

  In FIG. 6B, the time constant switching unit 9 is provided with a detection circuit 9a for detecting the above-described attack period and a switching signal generation circuit 9b for generating the above-described switching signal SEL.

  The detection circuit 9a includes a comparator OPA. When an external noise component N is generated at the output of the high pass filter 2, the level of the external noise component N and the external noise component N output from the low pass filter 3 are detected. Is compared with the level of the charge / discharge signal R indicating the charge voltage (charge voltage generated delayed from the external noise component N). Then, a period in which the level of the external noise component N is higher than the level of the charge / discharge signal R is detected as an attack period, and an attack period signal Sar having a logical value “H” is output. On the other hand, a period in which the level of the external noise component N is lower than the level of the charge / discharge signal R is detected as a recovery period, and an attack period signal Sar having a logical value “L” is output.

  Next, as shown in the figure, the switching signal generation circuit 9b includes a plurality of voltage dividing resistors (not shown) connected between a predetermined reference voltage (stabilized fixed voltage) Vref and the ground, and individual divisions. A plurality of comparators (not shown) that compare the divided voltage divided by the piezoresistor and the detection signal Se, and the comparison result output from these comparators are decoded to represent the level of the electric field strength. It is formed of a voltage comparison type analog-digital converter having a decoder (not shown) for decoding into digital data. The digital data is supplied as a time constant switching signal SEL to a switching circuit SW (see FIG. 4C) provided in the low-pass filter 3. In the decoder, the digital data indicating the electric field strength of any one of j-stage electric field strengths E1, E2 to Ej determined with a predetermined resolution (that is, the electric field strength corresponding to the comparison result output from the comparator). To decode.

  Next, as shown in FIG. 4C, the low-pass filter 3 receives an amplifier (not shown) that inputs the output of the high-pass filter 2 by power amplification or the like, and inputs the output of the high-pass filter 2 through the amplifier. J switched capacitor filters SCF1 to SCFj and a switching circuit SW for switching and outputting the outputs of the switched capacitor filters SCF1 to SCFj according to the time constant switching signal SEL. The number j of the switched capacitor filters SCF1 to SCFj is determined in accordance with the number j of switching stages of an attack time constant τa described later.

  The switched capacitor filters SCF1 to SCFj basically have the same configuration as shown in the figure.

  The switched capacitor filter SCF1 will be described as a representative. The capacitor C01 on the input side, the analog switch SA1, the operational amplifier OP1, the capacitor Ca1 connected between the output terminal and the inverting input terminal of the operational amplifier OP1, and the capacitor Ca1 The capacitor Cr1 and the analog switch SR1 connected to both ends are provided.

  Here, the analog switch SA1 repeats the switching operation in synchronization with the clock signal CK having a frequency fck sufficiently higher than the frequency of the external noise component N, and the output of the high-pass filter 2 supplied via the amplifier is a capacitor. It accumulates in C01 and is applied to the non-inverting input terminal of the operational amplifier OP1. The clock signal CK is generated by an oscillator (not shown) incorporated in the noise canceller circuit 100.

  The analog switch SR1 is turned on / off in accordance with the attack period signal Sar generated by the detection circuit 9a, and is turned off (cut off) in the attack period in which the attack period signal Sar is at the logical value “H”. On the other hand, when the connection of Cr1 is cut off and the attack period signal Sar becomes the logical value “L”, the capacitor Cr1 is connected to the capacitor Ca1 in parallel.

  Note that the capacitance values of the capacitors C01, Ca1, and Cr1 are predetermined values, which will be described later in detail.

  The switched capacitor filter SCF1 having such a configuration realizes a cutoff frequency variable type (time constant variable type) low-pass filter that switches a high cutoff frequency (in other words, a time constant) between an attack period and a recovery period. Yes.

  That is, in the attack period in which the attack period signal Sar is the logical value “H”, the analog switch SR1 is turned off, the capacitor Cr1 is opened, and only the capacitor Ca1 is feedback-connected to the operational amplifier OP1. As expressed by the equations (2a) and (2b), the high cutoff frequency fa1 and the attack time constant τa1 of the switched capacitor filter SCF1 are determined according to the relationship between the capacitance values of the capacitors C01 and Ca1 and the frequency fck of the clock signal CK. The low-pass filtering is applied to the external noise component N included in the output of the high-pass filter 2.

  On the other hand, in the recovery period in which the attack period signal Sar is at the logical value “L”, when the analog switch SR1 is turned on and the capacitors Ca1 and Cr1 are feedback-connected to the operational amplifier OP1, the feedback capacitance to the operational amplifier OP1 is increased. According to the relationship between the capacitance values of the capacitors C01, Ca1, and Cr1 and the frequency fck of the clock signal CK, as shown by the following equations (3a) and (3b), the value of the switched capacitor filter SCF1 increases. The high-frequency cutoff frequency fr1 and the recovery time constant τr1 are determined, and low-pass filtering is applied to the external noise component N included in the output of the high-pass filter 2.

  Next, the capacitors C01, Ca1, and Cr1 are determined to have a small capacitance value because it is necessary to repeatedly charge and discharge in accordance with the clock signal CK. However, the capacitance value of the capacitor C01 and the capacitance value of the capacitor Ca1 during the attack period are determined. In order to reduce the ratio {C01 / (Ca1 + Cr1)} between the capacitance value of the capacitor C01 and the total capacitance value of the capacitors Ca1 and Cr1 during the recovery period compared to the ratio (C01 / Ca1) with the following formula (4) It is decided according to the relationship.

  In this way, by setting the capacitance values of the capacitors C01, Ca1, and Cr1, the attack time constant τa1 during the attack period in which the level of the external noise component N rises rapidly is reduced, and the followability to the external noise component N is secured. On the other hand, the recovery time constant τr1 in the recovery period in which the level of the external noise component N decreases is increased to perform the slow discharge process.

  The remaining switched capacitor filters SCF2 to SCFj have the same configuration as the switched capacitor filter SCF1, with the capacitors C02 to C0j being the capacitor C01, the analog switches SA2 to SAj being the analog switch SA1, and the operational amplifiers OP2 to OPj being In the operational amplifier OP1, capacitors Ca2 to Caj correspond to the capacitor Ca1, capacitors Cr2 to Crj correspond to the capacitor Cr1, and analog switches SR2 to SRj correspond to the analog switch SR1, respectively.

  Further, in the above description, for each of the switched capacitor filters SCF1, SF2 to SCFj, the relationship between the capacitance values of the capacitors C01, Ca1, Cr1, the relationship between the capacitance values of the capacitors C02, Ca2, Cr2, and the capacitors C0j, Caj The relationship between the capacitance values of Crj is described, but the relationship between the capacitance values of the capacitors C01, Ca1, Cr1, C02, Ca2, Cr2, and C0j, Caj, Crj among the switched capacitor filters SCF1, SF2 to SCFj is also described. The predetermined relationship is determined and will be described with reference to FIGS.

  FIG. 5A shows the characteristics of the change between the modulated signal component generated in the envelope signal Senv and the broadband noise with respect to the electric field strength of the incoming radio wave shown in FIG. 3, and the external noise component N generated in the envelope signal Senv. The characteristic of the attack time constant τa to be adjusted according to the electric field strength of the incoming radio wave when charging by filtering is shown. FIG. 5B schematically shows the relationship among the broadband noise, the external noise component N, and the attack time constant τa.

  As can be seen from FIG. 5 (a), the automatic gain control (AGC) is performed by the AM detector 1, so that the reception intensity of the broadcast receiver 200 deteriorates in the envelope signal Senv as the electric field strength decreases. The level of the generated broadband noise will increase.

  For this reason, as shown in FIG. 5B, when the external noise component N occurs in the envelope signal Senv due to the external noise that is uncorrelated with the broadband noise, the external noise is generated when the electric field strength is strong. When the difference (voltage difference) between the peak level Lnmax of the component N and the broadband noise level Lzmin is large and charging is performed with a small attack time constant τa, the external noise component N is charged to a level lower than the peak level. Is possible. Therefore, the attack time constant τa when the electric field strength of the incoming radio wave is strong is set to a predetermined small time constant.

  On the other hand, when the electric field strength of the incoming radio wave is relatively weak, the level of the broadband noise increases, so the difference (voltage difference) between the peak level Lnmax of the external noise component N and the level Lzmid of the broadband noise is reduced. With a small attack time constant τa, it becomes difficult to charge the external noise component N to a level lower than its peak level without being affected by broadband noise. Therefore, the attack time constant τa when the electric field strength of the incoming radio wave is relatively weak is set to a value larger than the attack time constant τa when the electric field strength is strong.

  In addition, when the electric field strength of the incoming radio wave is weak, the level of the broadband noise greatly increases, so the difference (voltage difference) between the peak level of the external noise component N and the level Lzmax of the broadband noise is further reduced to a small value. Therefore, it is difficult to charge the external noise component N to a level lower than its peak level without being affected by broadband noise. Therefore, the attack time constant τa when the electric field strength of the incoming radio wave is weak is difficult. Is a value larger than the attack time constant τa when the electric field strength is relatively weak.

  Furthermore, the charging voltage may reach the peak level Lnmax of the external noise component N only by variably adjusting the value of the attack time constant τa according to the strength of the electric field strength. The charging voltage is made to reach an intermediate level Lp that is lower than the peak level Lnmax of the high level external noise component N generated due to the external noise and higher than the level Lzmax of the wideband noise generated when the electric field strength is weakened. The attack time constant τa is determined.

  The attack time constant τa determined in this way is shown in FIG.

  Then, the range EW of the electric field intensity when the electric field strength of the incoming radio wave is strong (when the broadband noise level is Lzmin) and when the electric field strength of the incoming radio wave is weak (when the broadband noise level is Lzmax) is described above. , Ej divided into j stages in association with the j stage field strengths E1 to Ej of the time constant switching signal SEL output from the switching signal generation circuit 9b (see FIG. 4B). , .Tau.aj corresponding to, and the time constants .tau.a1, .tau.a2,..., .Tau.aj are set as attack time constants .tau.a that should be varied in response to changes in the electric field strength. That is, when the electric field strength E1 is strong, the attack time constant τa1 is switched to the smallest, and when the electric field strength E2 is smaller than the electric field strength E1, the attack time constant τa2 is switched to an attack time constant τa2 that is larger. At Ej, the maximum attack time constant τaj is switched.

  Then, in order to generate the attack time constant τa1 by the switched capacitor filter SCF1 shown in FIG. 4C, the capacitance values of the capacitors C01 and Ca1 are determined based on the relationship of the above formulas (2a) and (2b). .

  In order to generate the attack time constant τa2 by the switched capacitor filter SCF2, the capacitance values of the capacitors C02 and Ca2 are determined based on the relationship of the following equations (5a) and (5b). In order to generate the attack time constant τaj in the filter SCFj, the capacitance values of the capacitors C0j and Caj are determined based on the relationship of the following equations (6a) and (6b).

  In this embodiment, the capacitors C01 to C0j are determined to have the same capacitance value, and the capacitance values of the capacitors Ca1 to Caj are proportional to the values of the attack time constants τa1 to τaj on the basis of the same capacitance value. Is decided.

  Next, the capacitance values of the capacitors Cr1 to Crj provided in the switched capacitor filters SCF1 to SCFj will be described.

  As described above, the capacitors Cr1 to Crj are provided to perform low-speed discharge in the recovery period when external noise generated over a long period of time is mixed into the broadcast receiver 200. Therefore, external noise generated at high density in time is analyzed by experiments, etc., and the average of the time intervals at which high-level external noise is generated that is longer than the average value of the time intervals at which individual external noise occurs and gives a sense of incongruity to hearing The recovery time constant τr that completes the discharge within a shorter time than the value is determined, and the capacitance values of the capacitors Cr1 to Crj are determined so that all the switched capacitor filters SCF1 to SCFj perform a low-speed discharge with the recovery time constant τr. It has been.

  That is, the recovery time constant τr1 of the switched capacitor filter SCF1, the high cut-off frequency fr1, and the capacitance value of the capacitor Cr1 are determined from the relationship of the following equations (7a) and (7b).

  Further, the recovery time constant τr2 of the switched capacitor filter SCF2, the high cut-off frequency fr2, and the capacitance value of the capacitor Cr2 are determined from the relationship of the following equations (8a) and (8b). The recovery time constant τrj of the filter SCFj, the high cut-off frequency frj, and the capacitance value of the capacitor Crj are determined from the relationship of the following equations (9a) and (9b).

  Next, the switching circuit SW is formed by an analog switch or the like having j input terminals and one output terminal individually connected to the output terminals of the switched capacitor filters SCF1 to SCFj, and the time constant switching signal SEL. Is switched according to the level of the electric field strength indicated by the above, the output signals of the switched capacitor filters SCF1 to SCFj are exclusively selected and output as the charge / discharge signal R.

  That is, when the time constant switching signal SEL indicating the electric field strength E1 is supplied, the switching circuit SW is connected to the switched capacitor filter SCF1 and outputs the output signal of the switched capacitor filter SCF1 as the charge / discharge signal R. When the time constant switching signal SEL indicating E2 is supplied, the switch capacitor filter SCF2 is connected to output the output signal of the switched capacitor filter SCF2 as the charge / discharge signal R, and similarly, when the electric field strength Ej is indicated. When the constant switching signal SEL is supplied, it is connected to the switched capacitor filter SCFj and the output signal of the switched capacitor filter SCFj is output as the charge / discharge signal R.

  When the switching circuit SW performs the switching operation according to the time constant switching signal SEL, the switched capacitor filters SCF1, SCF2 to SCFj are selected in association with the electric field strengths E1, E2 to Ej, and therefore the selected switched capacity is selected. An output signal charged / discharged based on the attack time constant τa and the recovery time constant τr set in the filter is output as the charge / discharge signal R.

  Therefore, a variable time constant low-pass filter for switching the high-frequency cutoff frequency is realized by the low-pass filter 3 having the switched capacitor filters SCF1, SCF2 to SCFj and the switching circuit SW shown in FIG.

  Next, the operation of the noise canceller circuit 100 will be described with reference to FIGS.

  FIG. 5C shows an external noise component N generated when external noise generated over a long period of time with a high density in time is mixed in the broadcast receiver 200 in a state where the electric field strength of the incoming radio wave is strong, The waveforms of the charge / discharge signal R and the threshold signal Rth are schematically shown. FIG. 5D shows an external noise component N generated when external noise generated over a long period of time with a high density in time in a state where the electric field strength of the incoming radio wave is weak, and charging / discharging. The waveforms of the signal R and the threshold signal Rth are schematically shown.

  First, when the electric field strength of the incoming radio wave is strong, the electric field strength detector 8 detects the electric field strength and outputs a detection signal Se indicating, for example, the electric field strength E1. Further, the time constant switching unit 9 outputs a time constant switching signal SEL indicating the electric field strength E1, and the switching circuit SW in the low-pass filter 3 performs the switching operation, so that the switched capacitor filter SCF1 is selected.

  Furthermore, when the external electric noise is mixed for a long time with a high density in time when the electric field strength of the incoming radio wave is strong, the external signal output from the high-pass filter 2 is exemplified as shown in FIG. The noise component N is input to the switched capacitor filter SCF1 and further input to the detection circuit 9a in the time constant switching unit 9.

  The detection circuit 9a detects the attack period by comparing the voltage of the external noise component N and the charge voltage of the charge / discharge signal R output from the low-pass filter 3, and further detects a logical value “H” in the attack period. In accordance with the attack period signal Sar, the analog switch SR1 of the switched capacitor filter SCF1 is turned off. Therefore, in the attack period, the external noise component N is charged based on the attack time constant τa1 in which the switched capacitor filter SCF1 is set to the smallest value.

  Further, when the electric field strength of the incoming radio wave is strong, the level of broadband noise such as white noise supplied to the switched capacitor filter SCF1 is low, so the base level BLV of the broadband noise is also low. Therefore, when the switched capacitor filter SCF1 charges the external noise component N based on the attack time constant τa1 set to the smallest value with reference to the low base level BLV, the external noise component N having a high charge voltage is obtained. A level lower than the peak level is reached.

  Further, when the attack period signal Sar becomes the logical value “L” and the recovery period is reached, the analog switch SR1 of the switched capacitor filter SCF1 is turned on. For this reason, in the recovery period, low-speed discharge is performed based on the recovery time constant τr1 with respect to the charging voltage charged in the attack period.

  In this way, the switched capacitor filter SCF1 charges / discharges the external noise component N based on the attack time constant τa1 and the recovery time constant τr1, so that the sawtooth charge / discharge signal R illustrated in FIG. And output via the switching circuit SW.

  Further, the sawtooth charge / discharge signal R is input to the amplifier 4 and amplified with the adjusted predetermined amplification factor g, thereby generating the threshold signal Rth whose level is adjusted to the high level side.

  Then, the comparator 5 compares the levels of the threshold signal Rth and the external noise component N, and generates a signal Sa indicating the noise generation period T of the external noise component N at a higher level than the threshold signal Rth.

  Here, as described above, when the switched capacitor filter SCF1 performs charge / discharge on the external noise component N based on the attack time constant τa1 and the recovery time constant τr1, the charge / discharge signal R becomes a high level external noise component. N is charged and held, and the held voltage is maintained at a level higher than the low level external noise component N. For this reason, the threshold signal Rth whose level is adjusted to the high level side by the amplifier 4 is also adjusted to a level that makes it easy to discriminate between the high level external noise component N and the low level external noise component N. As a result, when the comparator 5 compares the levels of the threshold signal Rth and the external noise component N, the noise generation period T of the high-level external noise component N generated due to the high-level external noise that gives a sense of incongruity to hearing. , And the low level external noise component N caused by the low level external noise that does not cause a sense of incongruity is not detected.

  Then, the gate signal generation circuit 6 shapes the signal Sa indicating the noise generation period T of the high-level external noise component N so that the gate signal Sg is generated and supplied to the interpolation circuit 7. However, during the noise generation period T indicated by the gate signal Sg, the noise component generated in the demodulated signal Smd output from the detector in the broadcast receiver 200 (noise component generated due to high level external noise) Are subjected to interpolation processing such as blocking and mute, and output as a demodulated signal Sx.

  As described above, when the external noise generated over a long period of time is mixed when the electric field strength of the incoming radio wave is strong, the interpolation circuit 7 causes the demodulated signal Smd to be generated due to the high level external noise. The generated noise component is removed, and the noise component generated in the demodulated signal Smd due to the low level external noise is not removed.

  For this reason, it is possible to prevent the stripped signal component from being generated in the demodulated signal Sx, and to remove a high-level noise component that gives an uncomfortable feeling.

  Next, with reference to FIG. 5D, an operation when the electric field strength of the incoming radio wave is weak will be described.

  When the electric field strength of the incoming radio wave is weak, the electric field strength detection unit 8 detects the electric field strength and outputs, for example, a detection signal Se indicating the electric field strength Ej. Further, the time constant switching unit 9 outputs a time constant switching signal SEL indicating the electric field strength Ej, and the switching circuit SW in the low-pass filter 3 performs the switching operation, so that the switched capacitor filter SCFj is selected.

  Further, when the external noise is mixed for a long time with a high density in time when the electric field strength of the incoming radio wave is weak, the external signal output from the high-pass filter 2 is exemplified as shown in FIG. The noise component N is input to the switched capacitor filter SCFj and further input to the detection circuit 9a in the time constant switching unit 9.

  Then, the analog switch SRj of the switched capacitor filter SCFj is turned off according to the attack period signal Sar output from the detection circuit 9a and having a logical value “H” indicating the attack period. Therefore, during the attack period, the external noise component N is charged based on the attack time constant τaj in which the switched capacitor filter SCFj is set to the largest value.

  Further, when the electric field strength of the incoming radio wave is weak, the broadband noise such as white noise supplied to the switched capacitor filter SCFj is at a high level, so that the base level BLV of the broadband noise is also increased. Therefore, when the switched capacitor filter SCFj charges the external noise component N based on the attack time constant τaj set to the largest value with the high base level BLV as a reference, the external noise component N having a high charge voltage is obtained. A level lower than the peak level is reached.

  Further, when the attack period signal Sar becomes the logical value “L” and the recovery period is reached, the analog switch SRj of the switched capacitor filter SCFj is turned on. For this reason, in the recovery period, low-speed discharge is performed based on the recovery time constant τrj with respect to the charging voltage charged in the attack period.

  In this way, the switched capacitor filter SCFj performs charge / discharge on the external noise component N based on the attack time constant τaj and the recovery time constant τrj, so that a sawtooth charge / discharge signal R as illustrated in FIG. And output via the switching circuit SW.

  Further, the sawtooth charge / discharge signal R is input to the amplifier 4 and amplified with the adjusted predetermined amplification factor g, thereby generating the threshold signal Rth whose level is adjusted to the high level side.

  Then, the comparator 5 compares the levels of the threshold signal Rth and the external noise component N, and generates a signal Sa indicating the noise generation period T of the external noise component N at a higher level than the threshold signal Rth.

  Here, as described above, when the switched capacitor filter SCFj performs charge / discharge on the external noise component N based on the attack time constant τaj and the recovery time constant τrj, the charge / discharge signal R becomes a high level external noise component. N is charged and held, and the held voltage is maintained at a level higher than the low level external noise component N. For this reason, the threshold signal Rth whose level is adjusted to the high level side by the amplifier 4 is also adjusted to a level that makes it easy to discriminate between the high level external noise component N and the low level external noise component N. As a result, when the comparator 5 compares the levels of the threshold signal Rth and the external noise component N, the noise generation period T of the high-level external noise component N generated due to the high-level external noise that gives a sense of incongruity to hearing. , And the low level external noise component N caused by the low level external noise that does not cause a sense of incongruity is not detected.

  Then, the gate signal generation circuit 6 shapes the signal Sa indicating the noise generation period T of the high-level external noise component N so that the gate signal Sg is generated and supplied to the interpolation circuit 7. However, during the noise generation period T indicated by the gate signal Sg, the noise component generated in the demodulated signal Smd output from the detector in the broadcast receiver 200 (noise component generated due to high level external noise) Are subjected to interpolation processing such as blocking and mute, and output as a demodulated signal Sx.

  As described above, when the external noise generated over a long period of time is mixed when the electric field strength of the incoming radio wave is weak, the interpolation circuit 7 causes the demodulated signal Smd to be generated due to the high level of external noise. The generated noise component is removed, and the noise component generated in the demodulated signal Smd due to the low level external noise is not removed.

  For this reason, it is possible to prevent the stripped signal component from being generated in the demodulated signal Sx, and to remove a high-level noise component that gives an uncomfortable feeling.

  In the above description of the operation, a typical case where the electric field strength of the incoming radio wave is strong and weak is described. However, when the electric field strength is intermediate between them, the electric field strength detector 8 One of the remaining switched capacitor filters SCF2 to SCFj-1 is selected in accordance with the detected electric field strength, and the external noise component N is determined based on the attack time constant τa set in the selected switched capacitor filter. Since it is charged and discharged based on the recovery time constant τr, even if the level of the broadband noise changes according to the electric field strength, the high level external noise component N is detected while the low level is not detected. A threshold signal Rth can be generated. Then, the interpolation circuit 7 removes noise components (noise components caused by high level external noise) generated in the demodulated signal Smd supplied from the detector in the broadcast receiver 200 by interpolation, and forms a strip shape. The demodulated signal Sx that prevents the occurrence of the signal component can be generated, and a sound that does not give a sense of incongruity to hearing can be reproduced by a speaker or the like.

  As described above, according to the noise canceller circuit 100 of the present embodiment, when external noise generated over a long period of time with high density is mixed, the low-pass filter 3 increases as the electric field strength of the incoming radio wave decreases. Since the charging is performed by increasing the value of the attack time constant τa when charging / discharging the external noise component N to generate the charge / discharge signal R, the wideband noise that increases as the electric field strength of the incoming radio wave becomes weaker. A high level external noise component N caused by a high level external noise is detected without being affected, and a low level external noise component N caused by a low level external noise is not detected. The charge / discharge signal R for generating can be generated.

  In other words, according to the noise canceller circuit 100 of the present embodiment, noise detection for detecting external noise that gives a sense of incongruity to a listener when external noise generated over a long period of time is mixed. The conventional problem that the sensitivity fluctuates depending on the electric field strength of the incoming radio wave can be improved, and the noise detection sensitivity can be stabilized.

  Therefore, it is possible to remove the influence of external noise regardless of whether the electric field strength of incoming radio waves varies and the reception state of the broadcast receiver is good.

  In the embodiment described above, the amplifier 4 is provided and the threshold signal Rth is generated by amplifying the charge / discharge signal R with the amplification factor g. It is provided for finely adjusting the level to generate the threshold signal Rth. That is, it is provided for finely adjusting the noise detection sensitivity. However, for example, in a configuration that does not require fine adjustment, the amplifier 4 may be omitted, and the charge / discharge signal R generated by the low-pass filter 3 may be directly supplied to the comparator 5 as the threshold signal Rth. .

  In the embodiment described above, the effective power is measured based on the intermediate frequency reception signal Scv generated by the frequency converter in the broadcast receiver in the electric field intensity detector 8 (see FIG. 4A). A detection signal Se indicating the effective power as the electric field strength of the incoming radio wave is generated and supplied to the time constant switching unit 9 to generate the time constant switching signal SEL. However, the following configuration may be used. That is, since an S meter signal (a signal indicating reception sensitivity) is generally generated in an IF filter in a broadcast receiver, a switching signal in the time constant switching unit 9 is generated using the S meter signal as a signal indicating the electric field strength of an incoming radio wave. As a configuration that is directly input to the circuit 9b, generates a time constant switching signal SEL according to the level of the S meter signal, and supplies it to the switching circuit SW (see FIG. 4C) provided in the low-pass filter 3. Also good. According to such a configuration, the electric field intensity detection unit 8 is not necessary, and thus the noise canceller circuit 100 can be made a simpler configuration.

  Next, an embodiment of the present invention will be described with reference to FIGS.

  6A is a block diagram showing the configuration of the noise canceller circuit 100 of this embodiment, FIG. 6B is a block diagram showing the configuration of the low-pass filter 3 and the time constant switching unit 9, and FIG. These are diagrams for explaining functions and operations. 6 (a) and 6 (b), the same or corresponding parts as those in FIGS.

  The noise canceller circuit 100 according to the present embodiment is provided in a digital broadcast receiver 200 that performs reception by digital signal processing. Therefore, the noise canceller circuit 100 is also formed of a digital circuit so as to perform noise canceller processing by digital signal processing.

  First, the configuration of the digital broadcast receiver 200 will be described. The broadcast receiver 200 shown in FIG. 4A is configured to generate a demodulated signal Smd or the like by performing analog signal processing on the intermediate frequency received signal Scv generated by the frequency converter. In the digital broadcast receiver 200 shown in FIG. 6A, a system up to a frequency converter that handles a high-frequency received signal is formed by an analog circuit, and a system after that is formed by a digital circuit. An A / D converter for digitally converting the analog intermediate frequency reception signal Scv output from the frequency converter into an intermediate frequency reception signal Dcv of digital data is provided, and an IF filter is applied to the intermediate frequency reception signal Dcv. The detector performs predetermined digital filtering and digital detection by digital signal processing to generate a demodulated signal Smd composed of digital data.

  The A / D converter described above is formed of a ΔΣ modulation type A / D converter, oversamples and quantizes the intermediate frequency received signal Scv, and removes quantization noise generated during the quantization. By performing the noise shaping function, the signal is digitally converted to an intermediate frequency reception signal Dcv with improved S / N. The configuration of the A / D converter using the ΔΣ modulation method is widely known, and thus detailed description thereof will be omitted.

  Further, the IF filter in the digital broadcast receiver 200 generates and outputs an S meter signal Smtr composed of digital data representing the reception sensitivity described later.

  Next, the configuration of the noise canceller circuit 100 will be described.

  As shown in FIG. 6A, the noise canceller circuit 100 includes an AM detector 1, a high-pass filter 2, a low-pass filter 3, an amplifier 4, a comparator 5, a gate signal generation circuit 6, and an interpolation circuit 7. For example, digital signal processing is performed in synchronization with a clock signal having a frequency sufficiently higher than the frequency of external noise such as ignition noise emitted from an automobile, so that it is similar to the noise canceller circuit 100 shown in FIG. In addition, noise canceller processing is performed.

  Here, the low-pass filter 3 and the time constant switching unit 9 have the configuration shown in FIG.

  First, the low-pass filter 3 is formed of a digital filter. The output of the high-pass filter 2 (an output including an external noise component N) is input and the difference signal COMP is subtracted by a charge / discharge signal R described later. A subtractor 3a to be generated; and a multiplier 3b that multiplies the difference signal COMP by a predetermined coefficient value (hereinafter referred to as “attack coefficient value”) α and outputs a multiplication signal Da indicating the multiplication value (α × COMP); A multiplier 3c that multiplies the difference signal COMP by a predetermined coefficient value (hereinafter referred to as “recovery coefficient value”) γ and outputs a multiplication signal Dr indicating the multiplication value (γ × COMP); and multipliers 3b and 3c An adder 3d for adding the multiplication signal D1 which is one of the multiplication signals Da or Dr output exclusively from the charge signal R and the charge / discharge signal R and generating an addition signal D2 indicating the addition value (D1 + R); , A delay circuit 3e that delays the addition signal D2 by one sample time set by the clock signal and outputs it as a delay signal D3; and when the value (level) of the delay signal D3 is 0 or more, the delay signal D3 is satisfied. A limiter circuit 3f that outputs as a discharge signal R and forcibly outputs a charge / discharge signal R adjusted to a value of 0 when the value (level) of the delay signal D3 is less than 0, and an attack period signal Sar described later. When the logic value is “H”, the multiplier 3b is operated to output the multiplication signal Da and the operation of the multiplier 3c is stopped to stop the output of the multiplication signal Dγ, while the attack period signal Sar is the logic value “ When L ″, the multiplier 3c is operated to output the multiplication signal Dγ, and the switching signal (not shown) is generated to stop the operation of the multiplier 3b and stop the output of the multiplication signal Da. A circuit 3g and an inverting circuit 3h are provided.

  The coefficient values α and γ of the multipliers 3b and 3c are set to values in the range of 0 to 1 as shown in the following equation (10), and are further recovered as compared to the attack coefficient value α. The coefficient value γ is set to a smaller fixed value, and the attack coefficient value α is switched to one of j-stage attack coefficient values α1 to αj by a filter coefficient switching unit 9b described later.

According to the low-pass filter 3 having such a configuration, when the delay of one sample time of the delay circuit 3e is z ⁻¹ , the input of the subtractor 3a is N (z), and the charge / discharge signal R is R (z), the multiplier In a state where 3b operates and the multiplier 3c stops operating, the transfer function Ha (z) expressed by the following equation (11a) is exhibited, and the multiplier 3b stops operating and the multiplier In the state in which 3c is operating, the transfer function Hr (z) expressed in z transformation expressed by the following equation (11b) is exhibited.

  Further, when the angular frequency ω and the sampling period τ are applied to the above equations (11a) and (11b) and replaced with z = jωτ (where j is a complex symbol), the respective transfer functions are expressed by equations (12a) (12b) ), The frequency characteristics are represented by Ha (ω) and Hr (ω).

  The high-frequency cutoff frequencies fa and fr of the transfer functions | Ha (ω) | and | Hr (ω) | are expressed by the following equations (13a) and (13b), respectively, and are further expressed by time constants τa and τr. If it represents, it will become following Formula (14a) (14b).

  FIG. 7 is a characteristic diagram showing the step response characteristics of the low-pass filter 3 described above. When the attack coefficient value α of the multiplier 3b changes from a large attack coefficient value αj to a small attack coefficient value α1, the attack time constant τa gradually increases, so that the step response gradually becomes slow and the smallest value is obtained. When the multiplier 3b set to the recovery coefficient value γ operates, the time constant τr becomes the largest and the step response becomes the slowest.

  As described above, the low-pass filter 3 is configured such that the attack coefficient value α of the multiplier 3b is variably adjusted to any one of the attack coefficient values α1 to αj of the plurality of stages j, so that the charging speed in the attack period with respect to the external noise component N is increased. (In other words, the attack time constant τa) changes, and further, the discharge rate in the recovery period (in other words, the time constant τr) is decreased to perform low-speed discharge.

  As described above, the limiter circuit 3f forcibly sets the charge / discharge signal R to the value 0 when the delay signal D3 output from the delay circuit 3e becomes a value of 0 or less, and the delay signal D3. When is a value greater than or equal to 0, the delay signal D3 is output as the charge / discharge signal R, thereby exhibiting the function of preventing the low-pass filter 3 from operating in an unstable manner.

  However, the limiter circuit 3f is provided in consideration of the stability of the low-pass filter 3, and may be omitted. The output D3 of the delay circuit 3e may be used as the charge / discharge signal R.

Next, the configuration of the time constant switching unit 9 will be described.
As shown in FIG. 6B, the time constant switching unit 9 includes a comparator (corresponding to the detection circuit 9a shown in FIG. 4B) 9a and a filter coefficient switching unit (FIG. 4B). 9b) corresponding to the switching signal generation circuit 9b shown in FIG.

  The comparator 9a compares the level of the output of the high-pass filter 2 with the level of the charge / discharge signal R generated with a delay of one sample time by the delay circuit 3e, and the output of the high-pass filter 2 is higher than the level of the charge / discharge signal R. In addition, an attack period signal Sar having a logical value “H” indicating an attack period is output, and when the output of the high-pass filter 2 is lower than the charge / discharge signal R, the logical value “L” indicating a recovery period is output. The attack period signal Sar is output. Then, by supplying the attack period signal Sar to the non-inverting circuit 3g and the inverting circuit 3h, the multiplier 3b is operated in the attack period and the multiplier 3c is operated in the recovery period.

  Therefore, when the external noise component N is generated in the output of the high pass filter 2, the low pass filter 3 causes the multiplier 3b to pass the external noise component N to the multiplier 3b during the attack period in which the external noise component N rises higher than the level of the charge / discharge signal R. The external noise component N is charged according to the time constant τa set based on the set attack coefficient value α. On the other hand, when the external noise component N becomes lower than the charge voltage level of the charge / discharge signal R, By switching from the attack period to the recovery period and starting the operation of the multiplier 3c, the charging voltage is set as the start level, and the low-speed discharge is performed according to the time constant τr set based on the recovery coefficient value γ.

  In this way, the comparator 9a detects the attack period and the recovery period, switches the operations of the multipliers 3b and 3c of the low-pass filter 3, and generates the charge / discharge signal R for the external noise component N. Yes.

  Next, the filter coefficient switching unit 9b is formed with a read-only memory (ROM) in a look-up table format, and an S meter signal Smtr indicating reception sensitivity generated by the IF filter in the digital broadcast receiver 200. Is detected as any one of j levels corresponding to the level of the S meter signal Smtr, and the attack coefficient value α associated with the detected level is output from the read-only memory to the multiplier 3b. And the above multiplication processing is performed.

  That is, in the read-only memory (ROM), the attack coefficient value α1 associated with the attack time constant τa1, the attack coefficient value α2 associated with the attack time constant τa2, and so on are associated with the attack time constant τaj. The data of the attack coefficient value αj is stored in advance.

  As described with reference to FIGS. 5A and 5B, when the level of the S meter signal mtr is high, the electric field strength of the incoming radio wave is strong, and as the level of the S meter signal mtr decreases, the incoming radio wave Based on the relationship that the electric field strength becomes weaker, the attack time constant τaj having a larger value is determined from the attack time constant τa1 having the smallest value, and the attack coefficient value associated with the attack time constant τa1 having the smaller value is further determined. α1 is stored in advance as the largest value data, and the attack coefficient value αj associated with the largest value attack time constant τaj is stored as the smallest value data.

  Then, when the electric field strength of the incoming radio wave is strong, the filter coefficient switching unit 9b, for example, sets the largest attack coefficient value α1 associated with the smallest attack time constant τa1 in accordance with the level of the S meter signal mtr. By setting the low-pass filter 3 as the attack coefficient value α, the low-pass filter 3 performs high-speed charging based on the attack time constant τa1 during the attack period, and when the electric field strength of the incoming radio wave is weak, it depends on the level of the S meter signal mtr. Thus, for example, by setting the smallest attack coefficient value αj associated with the largest attack time constant τaj as the attack coefficient value α of the multiplier 3b, the low-pass filter 3 can reduce the low speed based on the attack time constant τaj during the attack period. Let the battery charge.

  Therefore, when the electric field strength of the incoming radio wave is strong and the level of broadband noise generated at the output of the high-pass filter 2 is low, the filter coefficient switching unit 9b causes the low-pass filter 3 to reduce the external noise component N based on a small attack time constant τa1. By charging at high speed, the external noise N at a high level can be charged. When the electric field strength of the incoming radio wave is weak and the level of broadband noise generated at the output of the high-pass filter 2 is high, the low-pass filter 3 generates the external noise component N. By charging at low speed based on a large attack time constant τaj or the like, a high level of external noise N can be charged.

  Next, the operation of the noise canceller circuit 100 of the present embodiment having the above-described configuration will be described.

  For example, as shown in FIG. 5C, when the electric field strength of the incoming radio wave is strong, the S meter signal Smtr is at a high level, so that the filter coefficient switching unit 9b sends the multiplier 3b with a small attack time constant τa1, for example. Is supplied with a large attack coefficient value α1 to cause the multiplication process. Further, when external noise generated over a long period of time with a high electric field strength of incoming radio waves is mixed through the receiving antenna, the comparator 9a detects the rising level of the external noise component N. Thus, the attack period is set, and the multiplier 3b is in the operating state and the multiplier 3c is in the operation stopped state. Therefore, the multiplication signal Da shown in FIG. 6B becomes D1, the low-pass filter 3 performs the high-speed charging process, and the level (charging voltage) of the charging / discharging signal R increases.

  Next, when the external noise component N falls from the peak level and becomes lower than the charge / discharge signal R, the attack period ends, the multiplier 3c is in the operating state, and the multiplier 3b is in the operating stop state. Therefore, the multiplication signal Dr shown in FIG. 6B becomes D1, the low-pass filter 3 performs the low-speed discharge process, and the level of the charge / discharge signal R gradually decreases.

  Further, the amplifier 4 amplifies the sawtooth charge / discharge signal R generated in this way with an amplification factor g, thereby generating a sawtooth threshold signal Rth whose level is adjusted to the high level side. The threshold signal Rth and the level of the external noise component N are compared, the high level of the external noise component N is detected, and the low level of the external noise component N is not detected. For this reason, the signal Sa indicating the noise generation period T of the high-level external noise component N is output from the comparator 5, and the gate signal generation circuit 6 generates the gate signal Sg by shaping the signal Sa. And supplied to the interpolation circuit 7.

  Therefore, in the interpolation circuit 7, when a noise component caused by high level external noise is generated in the demodulated signal Smd output from the detector in the digital broadcast receiver 200, the above-mentioned signal is designated by the gate signal Sg. In the noise generation period T, interpolation processing such as removal or attenuation is performed, and a demodulated signal Sx from which the influence of high level external noise is removed is generated and output.

  As a result, it is possible to prevent the stripped signal component from being generated in the demodulated signal Sx, and to remove a high level noise component that gives an uncomfortable feeling.

  Next, with reference to FIG. 5 (d), an explanation will be given of the operation in the case where external noise generated over a long period of time with a high density is mixed through the receiving antenna when the electric field strength of the incoming radio wave is weak. To do.

  When the electric field strength of the incoming radio wave is weak, the S meter signal Smtr is at a low level, and therefore, for example, a small attack coefficient value αj for setting a large attack time constant τaj is supplied from the filter coefficient switching unit 9b to the multiplier 3b. To perform multiplication processing. Furthermore, when external noise generated over a long period of time with a high electric field intensity of incoming radio waves is mixed through the receiving antenna, the comparator 9a detects the rising level of the external noise component N. Thus, the attack period is set, and the multiplier 3b is in the operating state and the multiplier 3c is in the operation stopped state.

  Accordingly, the multiplication signal Da shown in FIG. 6B becomes D1, and the low-pass filter 3 performs the low-speed charging process from the level of the broadband noise whose level is rising. As a result of performing low-speed charging according to a large attack time constant τaj, charging is performed so that the charging voltage reaches a level lower than the peak level of the external noise component N even if the level of broadband noise is high.

  Next, when the external noise component N falls from the peak level and becomes lower than the charge / discharge signal R, the attack period ends, the multiplier 3c is in the operating state, and the multiplier 3b is in the operating stop state.

  Then, the multiplication signal Dr shown in FIG. 6B becomes D1, the low-pass filter 3 performs the low-speed discharge process, and the level of the charge / discharge signal R gradually decreases.

  Further, the amplifier 4 amplifies the sawtooth charge / discharge signal R generated in this way with an amplification factor g, thereby generating a sawtooth threshold signal Rth whose level is adjusted to the high level side. The threshold signal Rth and the level of the external noise component N are compared, the high level of the external noise component N is detected, and the low level of the external noise component N is not detected. For this reason, the signal Sa indicating the noise generation period T of the high-level external noise component N is output from the comparator 5, and the gate signal generation circuit 6 generates the gate signal Sg by shaping the signal Sa. And supplied to the interpolation circuit 7.

  Therefore, in the interpolation circuit 7, when a noise component caused by high level external noise is generated in the demodulated signal Smd output from the detector in the digital broadcast receiver 200, the above-mentioned signal is designated by the gate signal Sg. In the noise generation period T, interpolation processing such as removal or attenuation is performed, and a demodulated signal Sx from which the influence of high level external noise is removed is generated and output.

  As a result, it is possible to prevent the stripped signal component from being generated in the demodulated signal Sx, and to remove a high level noise component that gives an uncomfortable feeling.

  As described above, according to the noise canceller circuit 100 of the present embodiment, when the external noise generated over a long time with high density is mixed, the low-pass filter 3 reduces the electric field strength of the incoming radio wave. Since charging is performed by increasing the value of the attack attack time constant τa when charging / discharging the external noise component N to generate the charge / discharge signal R, the broadband noise that increases as the electric field strength of the incoming radio wave becomes weaker. A high level external noise component N caused by a high level external noise is detected without being affected, and a low level external noise component N caused by a low level external noise is not detected. The charge / discharge signal R for generating can be generated.

  In other words, according to the noise canceller circuit 100 of the present embodiment, noise detection for detecting external noise that gives a sense of incongruity to a listener when external noise generated over a long period of time is mixed. The conventional problem that the sensitivity fluctuates depending on the electric field strength of the incoming radio wave can be improved, and the noise detection sensitivity can be stabilized.

  Therefore, it is possible to remove the influence of external noise regardless of whether the electric field strength of incoming radio waves varies and the reception state of the broadcast receiver is good.

  Further, since the low-pass filter 3 for charging / discharging the external noise component N is formed by a digital filter having an extremely simple configuration, the noise canceller circuit 100 can be reduced in size.

It is a block diagram and function explanatory drawing for demonstrating the structure and function of the conventional noise canceller circuit. Furthermore, it is a block diagram and a function explanatory diagram for explaining a configuration and a function of another conventional noise canceller circuit. It is explanatory drawing for demonstrating the problem of the conventional noise canceller circuit shown in FIG. It is a block diagram showing the structure of the noise canceller circuit which concerns on this embodiment. FIG. 5 is an explanatory diagram for explaining functions and operations of the noise canceller circuit shown in FIG. 4. It is a block diagram showing the structure of the noise canceller circuit which concerns on an Example. It is explanatory drawing for demonstrating the function of the noise canceller circuit shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 ... Noise canceller circuit 200 ... Broadcast receiver 1 ... AM detector 2 ... High pass filter 3 ... Low pass filter 4 ... Amplifier 5 ... Comparator 6 ... Gate signal generation circuit 7 ... Interpolation circuit 8 ... Electric field strength detection part 9 ... Time constant switching Section SCF to SCFj: Switched capacitor filter 3a ... Subtractor 3b, 3c ... Multiplier 3d ... Adder 3e ... Delay circuit 9a ... Detection circuit, Comparator 9b ... Switching signal generation circuit, Filter coefficient switching section

Claims (5)

A noise canceller for removing noise components generated in a demodulated signal output from a detector provided in the receiver due to pulsed external noise mixed in the receiver with a high density in time. A circuit,
Interpolation means for performing interpolation processing on the demodulated signal output from the detector;
AM detection means for generating a signal including an external noise component generated in the intermediate frequency reception signal due to the external noise by envelope detection of the intermediate frequency reception signal generated by a frequency converter in the receiver; A high-pass filter,
A low-pass filter that generates a charge / discharge signal by charging an external noise component with a predetermined attack time constant and discharging at a low speed with a predetermined recovery time constant;
Amplifying means for adjusting the level of the charge / discharge signal to generate a threshold signal;
Comparing means for comparing the threshold signal with an external noise component and detecting a noise occurrence period of the external noise component at a higher level than the threshold signal;
Gate signal generation means for interpolating the noise generated in the demodulated signal due to the external noise by supplying a gate signal indicating the noise generation period to the interpolation means;
Detecting means for detecting electric field strength of radio waves coming from a broadcasting station to the receiver;
When the electric field strength detected by the detection means is strong, the attack time constant of the low-pass filter is set to a small value, and the attack time constant of the low-pass filter is set to a larger value as the electric field strength becomes weaker. Time constant switching means for charging the external noise component,
A noise canceller circuit comprising: The low-pass filter is
Formed of a switched capacitor filter that variably sets the attack time constant according to an instruction of the time constant switching means;
The noise canceller circuit according to claim 1. The low-pass filter is
Formed of a digital filter that variably sets the attack time constant in accordance with instructions of the time constant switching means;
The noise canceller circuit according to claim 1. The digital filter is:
A subtractor for inputting the external noise component;
A first multiplier that multiplies the difference signal output from the subtractor by a predetermined attack coefficient value;
A second multiplier that multiplies the difference signal output from the subtractor by a predetermined recovery coefficient value;
An adder for inputting a multiplication signal output exclusively from the first or second multiplier;
The output of the adder is delayed by one sample time and output as the charge / discharge signal and supplied to the subtractor and the adder, and the subtractor subtracts the charge / discharge signal from the external noise component. A delay circuit that generates the difference signal and adds the multiplication signal and the charge / discharge signal in the adder;
The level of the external noise component and the charge / discharge signal are compared. When the level of the external noise component is higher, the first multiplier is operated, and when the level of the external noise component is lower, the first A comparator operating two multipliers;
When the electric field strength detected by the detection means is strong, the attack coefficient value of the first multiplier is set to a predetermined large value, and the attack coefficient value is set to a smaller value as the electric field strength becomes weaker. Filter coefficient switching means to be set;
The noise canceller circuit according to claim 3, further comprising: Noise that removes noise components generated in the demodulated signal output from the detector provided in the receiver due to pulsed external noise that occurs in the receiver at a high density for a long time. A removal method,
An interpolation process for interpolating the demodulated signal output from the detector;
An AM detection step of generating a signal including an external noise component generated in the intermediate frequency reception signal due to the external noise by performing envelope detection on the intermediate frequency reception signal generated by a frequency converter in the receiver; A high-pass filtering process;
The low-pass filtering step of generating a charge / discharge signal by charging the external noise component with a predetermined attack time constant and discharging at a low speed with a predetermined recovery time constant;
An amplification step of adjusting a level of the charge / discharge signal to generate a threshold signal;
A comparison step of comparing the threshold signal with an external noise component, detecting a noise generation period of an external noise component at a level higher than the threshold signal, and setting it as a period of an interpolation process in the interpolation step;
A detection step of detecting electric field strength of radio waves coming from a broadcasting station to the receiver;
When the detected electric field strength is strong, the attack time constant in the low-pass filtering step is set to a small value, and the attack time constant in the low-pass filtering step is set to a larger value as the electric field strength becomes weaker. A time constant switching step for setting and charging the external noise component,
The noise removal method characterized by comprising.

Images