JPH05129860A - Power amplifier and transmitter - Google Patents

Abstract

PURPOSE:To execute the burst control of the linear power amplifier without incurring the spread of spectrums. CONSTITUTION:The envelope signal of an input signal RFin is smoothed through a waveform shaping circuit 14 of a changeover circuit 11, then inputted to a comparator 4 in feed-back loop, thus comparative reference itself is smoothed and smooth burst control can be realized during the transition period of burst transmission. In the case of the transmitter, of which an input signal to the power amplifier is a QAM signal, smooth burst control can be realized by modifying transmission data during the transition period of burst transmission to those data which minimizes the amplitude fluctuation of the QAM signal.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a linearized power amplifier used in a radio transmitter for mobile communication and the like and a transmitter equipped with the linearized power amplifier, and more particularly to a burst control method thereof.

[0002]

2. Description of the Related Art For example, QAM (quadrature amplitude m
In the case of power amplification of an odulation signal, its amplitude is also an important factor, and linear amplification is required.

FIG. 2 shows the configuration of a conventional linearized power amplifier using such envelope feedback.

In FIG. 2, a saturated power amplifier (PA)
1 for amplifying power, the saturation power amplifier 1 has a control terminal 15 for controlling its output signal (output power), and the voltage applied to this terminal (control signal) is The output power is changed accordingly. However, since the power amplifier 1 is a saturated type, the control signal and the output signal have a non-linear relationship. Therefore, in order to form a linear amplifier using the saturated power amplifier 1, envelope feedback is used as shown in FIG.

The structure for realizing this linear amplification will be described below. Output signal RFo of the saturation type power amplifier 1
A part of ut is taken out by the output coupler (for example, a coupling strip line) 13 and given to the output detector 2. The output detector 2 obtains the envelope signal of the extracted signal and supplies it to the inverting input terminal of the comparator 4 via the gain varying device 3 of the gain β1.

On the other hand, most of the input signal RFin is input to the saturation type power amplifier 1 via the input coupler 12, part of which is extracted by the input coupler 12 and input to the input detector 5. The input detector 5 obtains the envelope signal of the input signal and then supplies it to the non-inverting input terminal of the comparator 4 via the gain varying device 6 having the gain β2.

The comparator 4 obtains the difference between the envelope signal in the output signal and the envelope signal in the input signal, and applies this difference signal to the control terminal 15 as a feedback voltage.

Next, the operation of the above configuration will be described to clarify that the amplitude component of the input signal RFin is faithfully (linearly) amplified.

The amplitude component changes every moment, for example, QAM
A part of the signal is taken out by the input coupler 12 and detected by the input detector 5 to generate an amplitude component (envelope) of the QAM signal. This envelope signal is amplified by an appropriate gain β2 and becomes a reference signal for the comparator 4.

The output signal RFout is similarly detected by the output detector 2 to generate its amplitude component (envelope), amplified by an appropriate gain β1 and given to the comparator 4.

Here, the gains β1 and β2 are the power amplifier 1
Is selected in consideration of the amplification factor of. The comparator 4 applies the error component of the envelope of the output signal RFout to the control terminal 15 as a feedback signal with respect to the reference signal obtained from the input signal RFin, and thus, by the operation of this feedback loop, the power amplifier 1 Even if the non-linear amplification operation is performed, the amplitude component of the input signal RFin is always reproduced (linearly amplified) in the output signal RFout.

That is, if the loop gain has a sufficient loop gain in the band of the amplitude component of the QAM signal, the non-linearity of the applied voltage to the control terminal 15 and the output power is also improved by the loop gain.

By the way, in a digital automobile telephone system, TDMA (carrier sharing by time division multiplexing) is generally adopted. Therefore, the wireless transmitter needs to perform transmission at every specific time, and the linearized power amplifier also needs to perform on / off control, that is, burst control.

For example, in the above-mentioned digital cellular car telephone system, since 3-channel TDMA is adopted, as shown in FIG. 3, it is necessary to carry out burst transmission for every 6.66 msec of 20 msec.

The burst control will be described below.

FIG. 4B is a transmission burst definition for the digital cellular car telephone system described above. The first three-symbol period is a guard time, and is set to rise to a specified power level in the next three-symbol period, and is lowered in the last three symbols.

The following matters are important for raising and lowering the power level. That is, the rise and fall should be performed as smoothly as possible. This is because the abrupt rising and falling waveforms include high frequency components, and the transmission spectrum spreads at the moment of falling and rising, causing adjacent channel interference. That is, a waveform change having a steep slope as shown by a broken line in FIG. 4A and an angular waveform change as shown by a one-dot chain line are not preferable, and a smooth waveform change as shown by a solid line is preferable.

Conventionally, as a method for performing such smooth burst control, AM modulation is performed by using a variable attenuator with a PIN diode on the output side of the power amplifier, and baseband I and Q signals are smoothly started up. There was a method of going down or down.

FIG. 5 is an explanatory view of the method of FIG. Two PIN diodes 19 and 20 whose cathodes are grounded are provided at positions where the phases of the transmission lines 18 provided on the output side of the power amplifier differ by a predetermined amount. These PIN diodes 19 and 20 can be used as variable attenuators because their impedance changes according to the bias current during forward bias. Therefore, the output signal can be burst-modulated by the bias currents of the PIN diodes 19 and 20.

FIG. 6 is an explanatory view of the method of FIG. In the digital car telephone system, a quadrature modulator shown in FIG. 6 is provided immediately before the power amplifier, and the multiplier 21 multiplies the I signal and the carrier signal, and the Q signal and π /
The multiplier 22 multiplies the carrier signal from the two-phase shifter 23 and the combined signal is combined by the combiner 24 to form the input signal (QAM signal) for the power amplifier. By raising or lowering the input I and Q signals with a desired waveform, the input power level to the power amplifier also changes smoothly, and smooth burst control is realized.

[0021]

However, each of the above burst control methods has the following drawbacks.

In the case of the method of (1), in practice, a plurality of PIN diodes are required to obtain the required carrier off level for the output signal of the power amplifier, and the bias control circuit thereof is also required. Therefore, the number of parts increases, and there is a problem in terms of configuration and cost.

In the case of the method (1), the I and Q signals to the quadrature modulator are changed. Generally, the carrier leak when the carrier of the quadrature modulator is off is about 30 dB, and the output level of the modulator is sufficient. It is impossible to narrow down to. Therefore, it is impossible for the power amplification device to reduce the output power to the carrier off level.

Further, in the above both burst control methods, no consideration is given to the signal waveforms at the time of rising and the time of falling. Therefore, even if an attempt is made to perform smooth rising and falling, the signal waveforms There was a limit.

The first aspect of the present invention has been made in consideration of the above points, and it is an object of the present invention to provide a power amplification device which can realize burst control without spectrum spread with a simple structure.

A second aspect of the present invention is to provide a power amplifying device which can sufficiently limit the output power to the carrier off level during the transmission stop period.

A third aspect of the present invention is to provide a transmitter capable of supplying a signal waveform suitable for burst control at the time of start-up and at the time of fall to an internal power amplifier.

[0028]

A first invention is a variable attenuator having a variable gain terminal for attenuating an input signal, and a saturation type power amplifier having a control terminal for power amplifying the output of the variable attenuator. A feedback loop in which the DC component of the error signal of the envelope signal of the output signal from the power amplifier is input to the control terminal and the AC component is input to the gain variable terminal, based on the envelope signal of the input signal. A power amplifier for a transmitter using envelope feedback for linearly amplifying the input signal,
A switching circuit having a waveform shaping circuit that smoothes the envelope signal of the input signal is provided, and the switching circuit changes the envelope signal of the input signal through the waveform shaping circuit during the transient period of burst transmission. The input is made to the comparator in the feedback loop.

According to a second aspect of the present invention, in the power amplification device, a carrier off for forcibly setting the voltage at the control terminal and the gain variable terminal to zero for a certain period after a certain period of time has elapsed from the start of the fall of burst transmission. It is provided with a circuit.

According to a third aspect of the present invention, in a transmitter including the power amplification device, wherein the input signal to the power amplification device is a QAM signal, transmission data in a transient period of burst transmission is changed in amplitude of the QAM signal. Is the minimum data.

[0031]

In the first aspect of the present invention, during the transient period of burst transmission, the envelope signal of the input signal smoothed through the waveform shaping circuit is input to the comparator in the feedback loop for comparison. The standard itself becomes smooth, and smooth burst control can be realized.

According to the second aspect of the present invention, the carrier off circuit forcibly holds the control terminal of the power amplifier and the like at the zero voltage at the time of the burst fall, so that the residual voltage or the vibration in the waveform shaping circuit of the switching circuit is caused. It prevents the influence and quickly narrows down the output power of the power amplifier to the off level.

In the third aspect of the present invention, the smoothness of burst control can be realized by setting the transmission data in the transient period of burst transmission to the data in which the amplitude fluctuation of the QAM signal is minimized.

The QAM signal is a π / 4 shift DQP.
If the signal is an SK modulated signal, the transmission data whose phase transition is +45 degrees (or -45 degrees) in the rising period of burst transmission and whose phase transition is -45 degrees (or +45 degrees) in the falling period is amplitude fluctuation. Will be the minimum.

[0035]

1 is a circuit diagram of a band division type linearized power amplifier according to an embodiment of the present invention. In FIG.
7, 9 are low-pass filters (LPF), 8 are high-pass filters (HP
F), 10 is an RF variable attenuator (RF variable ATT) having a variable gain terminal 16 and a variable attenuation amount, 11 is a switching circuit described later, and others are the same as those denoted by the same reference numerals in FIG. ..

The input signal RFin is input to the input coupler 12, R
Saturation type power amplifier (PA) 1 through F variable ATT10
Output power R from the output coupler 13 after power amplification by
It is output as Fout. A part of the output signal RFout of the saturation type power amplifier 1 is taken out by the output coupler 13 and given to the output detector 2. This output detector 2 obtains an envelope signal of the output signal, which is gain β.
It is given to the inverting input terminal of the comparator 4 via the gain variable device 3 of 1. On the other hand, a part of the input signal RFin is taken out by the input coupler 12 and given to the input detector 5, and the input detector 5 obtains the envelope signal of the input signal. It is applied to the non-inverting input terminal of the comparator 4 via a switching circuit 11 described later. The comparator 4 obtains the difference between the envelope signal in the output signal and the envelope signal in the input signal, and outputs this as a feedback control signal to the LPF 7 and LPF 8.

This feedback control signal is a DC component (low frequency component) that determines the average output power level of the saturated power amplifier 1.
And an AC component (high frequency component) that compensates for distortion of amplitude fluctuation. The feedback control signal input to the LPF 7 has its AC component removed and is applied to the control terminal 15 to control the gain of the saturated power amplifier 1. Therefore, the power control by the saturated power amplifier 1 mainly functions as control of the average power level. On the other hand, the feedback control signal input to the HPF 8 has the DC component removed, and is applied to the gain variable terminal 16 through the LPF 9 that limits the band in a band that can sufficiently follow the data amplitude variation, and the attenuation amount of the RF variable ATT 10 is reduced. Control. Therefore, the control of the attenuation amount of the RF variable ATT 10 mainly functions as the control of the amplitude distortion compensation.

As described above, in the linearization power amplifier of FIG. 1, the feedback control voltage is separated into the DC component and the AC component, and D
Since the average power level is controlled by the C component and the amplitude distortion compensation is performed by the AC component, the power control range in the saturation power amplifier 1 can be reduced, and the AM-PM conversion distortion can be reduced. Is obtained.

The RF variable ATT 10 may be a high frequency variable attenuator using a PIN diode or the like as shown in FIG. 5, and the LPFs 7, 9 and HPF.
8 is a resistor R and a capacitor C as shown in FIG.
The following simple circuit is enough. Also LPF7, HPF8, LP
The frequency characteristics of F9 are set as shown in A, C, and B of FIG. 7, and the cutoff frequencies are f.
₀ (for example, 1 kHz), f ₁ (for example, 100 kHz),
It is f ₀ .

Next, the operation of the switching circuit 11 will be mainly described. In burst control, it is necessary to smoothly raise and lower the output signal RFout from the viewpoint of spurious diffusion prevention.
Considering that it can respond to the change of the falling edge sufficiently quickly, the input voltage to the inverting input terminal of the comparator 4, that is, the output voltage of the switching circuit 11 may be smoothly raised and lowered.

As shown in FIG. 1, the switching circuit 11 has two routes RA and RB for giving the envelope signal of the input signal from the gain varying device 6 to the comparator 4. Root RA
A switch SW1 is inserted in the. A switch SW2, a waveform shaping circuit 14 whose detailed configuration is shown in FIG. 9, and a switch SW3 are sequentially inserted in the other route RB.

Each switch SW1, SW2, SW3 is a simple switch which is turned on / off by a burst control signal from a timing control circuit (not shown), and is composed of an IC such as a CMOS.

The waveform shaping circuit 14 is for smoothing the envelope signal so as to shape the waveform at the rise and fall of the burst control as described later. The waveform shaping circuit 14 is realized by, for example, as shown in FIG. 9, a second-order low-pass filter (hereinafter referred to as a second-order circuit) having an RC filter section in two stages, which is configured by an active filter using an operational amplifier. To be done.

In FIG. 9, a resistor Rg is a ground resistor for setting the DC input level to 0V, and is given to the main body of the low-pass filter via the buffer amplifier H. The low-pass filter body includes an operational amplifier OP, two resistors R1 and R2 inserted between the output end of the buffer amplifier H and the non-inverting input end of the operational amplifier OP, and a resistor R2 and RC.
One end of the filter section is connected to the capacitor C1 whose one end is connected to the non-inverting input end of the operational amplifier OP, and one end is connected to the connection midpoint of the resistors R1 and R2 and the other end is connected to the operational amplifier OP.
Of the operational amplifier OP, and a resistor R3 connected to the inverting input terminal and the output terminal of the operational amplifier OP.

Such a secondary circuit (waveform shaping circuit) 14
As shown in FIG. 10, the output waveform of the step input response can be arbitrarily set by the sharpness Q value. In addition, the change is smoother than that of a primary circuit using one RC filter unit, and a sufficient output waveform can be obtained as compared with a tertiary circuit using three or more RC filter units. Nevertheless, the configuration is simple. In this example,
As described above, one feature is that the secondary circuit is used as the waveform shaping circuit 14.

Next, the burst control operation in the linearized power amplifier will be described.

FIG. 11 shows the transmission burst and the switching timing of the switches SW1, SW2 and SW3 of the switching circuit 11, and is a chart corresponding to the above-mentioned digital cellular car telephone system.

FIG. 11A shows the on (transmission) off (stop) timing of the output power, which is basically represented by a rectangular wave shape, but has rising and falling edges in order to suppress the spread of the spectrum. It needs to be smooth. Switches SW1, SW2, S shown in FIGS. 11B to 11D.
In the operation of W3, ON means connection and OFF means open. Further, the open / close transition time of each switch SW1, SW2, SW3 is assumed to be sufficiently fast.

Below, the periods are set as follows: T1, T2, T
The operation will be described by dividing it into five types of T3, T4 and T5.

The period T1 (~t _0): In this period to turn on only the switch SW3 (FIG. 11 (D)). At this time, since the switches SW1 and SW2 are opened,
The output from the switching circuit 11 (FIG. 11E) becomes 0V (ground) regardless of whether or not an input signal arrives. Therefore, the linearization loop operates with 0V as the reference voltage, and the output power becomes 0. For example, even if a carrier signal leaked from the quadrature modulator (see FIG. 6) is input, the output power is 0.
Therefore, the carrier off level can be surely achieved.

The period T2 (t ₀ ~t _1): In this period the switch SW2 is turned on in addition to the switch SW3. Therefore, the envelope signal of the input signal can pass through the route RB. During this period, the input signal itself rises due to the transition to the transmission operation, but since the envelope signal is input to the secondary circuit 14, the output signal of the switching circuit 11 is as shown in FIG. As shown, it smoothly rises toward the average value (integral value) of the input envelope signal. Since the reference voltage to the comparator 4 rises smoothly in this way, the output power also rises smoothly.

[0052] period T3 (t ₁ ~t _2): In this period,
Switches SW1 and SW2 are turned on, and switch SW3
To turn off. Therefore, the envelope signal of the input signal passes through the route RA as it is and becomes the output signal of the switching circuit 11.
At this time, since the input signal itself has risen completely and the envelope signal is given to the comparator 4 as a reference signal, linear amplification is performed by a normal loop operation.

Since the switch SW3 is opened, the output from the secondary circuit 14 is not given to the comparator 4. However, the envelope signal is input to the secondary circuit 14, and The average level voltage of the input envelope is output from the next circuit 14. This is intended to be executed with a smooth waveform change in the fall period described later.

[0054] period T4 (t ₂ ~t _3): In this period,
The switches SW1 and SW2 are turned off, and the switch SW3
To turn on. Therefore, at this time, the output of the secondary circuit 14 is given to the comparator 4 as the output signal of the switching circuit 11. During this period, the input envelope signal is not input to the secondary circuit 14, but according to the time constant of the secondary circuit 14, the average level of the envelope signal in the normal operation changes from 0V to 2V.
The step-down response waveform of the next circuit 14 causes a smooth fall. Therefore, the output power level also falls smoothly.

[0055] period T5 (t ₃ ~): In this period, to turn on only the switch SW3. Therefore, the period T described above
Since the same operation as that of No. 1 is performed, the description is omitted.

Although the band of the linearization loop is sufficiently wide, the control range that can be followed at high speed depends on the variable range of the RF variable ATT 10 in FIG. 1, and therefore, up to a sufficient carrier off level (-60 dBm or less). It is difficult to reduce the output power. Therefore, only the fall and rise times are
7. Switching the time constant of HPF8, L of the linearization loop
By speeding up both the route including the PF7 and the route including the HPF8, both the attenuation amount of the RF variable ATT10 and the gain of the saturated power amplifier 1 are rapidly controlled, and the carrier off level is secured.

Specifically, the LPF 7 and the HPF 8 are configured as shown in FIGS. 12A and 12B, respectively.
The switch SW shown in FIG. 12 is switched by a burst control signal having the same timing as the timing of switching the switch SW2 shown in FIG.
That is, the band of the LPF 7 is changed from f _o to f ₁ (wide band) during the period of t _o to t ₁ and t ₂ to t ₃ shown in FIG.
The bandwidth of the F8 to through from f _o. As a result, both the route including the LPF7 and the route including the HPF8 of the linearization loop become a wide band at the time of burst rise and fall, and both the RF variable ATT10 and the saturation type power amplifier 1 function to cause the carrier off level. It is possible to start up from (-60 dBm) and fall to the carrier off level. Note that FIG. 13 collectively shows the above-mentioned control method for the pass band of the LPF 7 and the HPF 8.

In the case of the digital cellular car telephone system adopted in North America, since three symbol periods (one symbol period is about 41 μsec) are assigned to the rising and falling edges, the rising period is set in consideration of the settling period, for example. 82 μsec. At this time, assuming that the sharpness Q of the secondary circuit 14 is 0.7 that can realize a good rising with little overshoot at the time of rising (see FIG. 10).
The values of the respective circuit elements constituting the secondary circuit 14 shown in are, for example, R1 = R2 = 1 kΩ, C1 = 0.038 μF, C2 =
It may be 0.019 μF, R3 = 10 kΩ, and Rg = 10 kΩ.

Although it has been described that the rise and fall of the burst control are smoothed by providing the switching circuit 14, in the case of the digital cellular car telephone system adopted in North America, in addition to this. hand,
By doing the following, the rising and falling waveforms can be made smoother. That is, in the digital cellular car telephone system, π / 4 shift DQPSK
A signal (a type of QAM signal) amplitude-phase modulated by the (differential quadrature phase shift keying) method is linearized and amplified, but the rise and fall times are each 3
By devising the data for each symbol period as follows, a smoother transient response of burst control can be obtained.

This will be described below.

FIG. 14 is a spatial phase transition diagram of a π / 4 shift DQPSK modulated wave. In FIG. 14, the vertical axis represents the I signal component input to the quadrature modulator (see FIG. 6), and the horizontal axis represents the Q signal component input to the quadrature modulator. The input audio signal is converted into a digital signal and then converted into an I signal component and a Q signal component through a mapping process for each predetermined bit. The I signal component and the Q signal component are respectively
It takes ± 1, 0, ± 2 ^1/2 , and the amplitude is the same among these combinations (the distance from the origin O is the same).
Eight phase points b1 to b8 that can transit are set.
The diameter (radius) of the spatial phase transition diagram represents the amplitude. Further, in the case of the π / 4 shift DQPSK modulated wave, it is determined that a certain phase point is shifted to any one of four phase points that are ± 45 degrees and ± 135 degrees. For example, the phase point b1 may be shifted to the phase points b2, b4, b6 or b8.

The phase point changes every 41 μsec.

Here, such a π / 4 shift DQPS
The relationship between the phase transition and the amplitude waveform of the K modulated wave will be considered for the three-symbol period that is the rising and falling periods.

FIG. 15 shows the case where the phase transition is every 45 degrees, and FIG. 16 shows the case where the phase transition is every 135 degrees.

When the phase transition is 45 degrees each,
As shown in (A), since the phase transitions at almost the same distance from the origin O, during the normal period, there is little change in the amplitude waveform as shown in FIG. 15 (B), and even if it is applied to the rising period. The variation is small as shown in FIG.

On the other hand, when the phase transition is 135 degrees at a time, the transition is made to the next phase point passing near the origin O as shown in FIG. For example, the fluctuation of the amplitude waveform is large as shown in FIG. 16B, and even if it is applied to the rising period, the fluctuation is large as shown in FIG. 16C.

Such a relationship between the phase transition and the amplitude change holds in the fall period as well.

Therefore, in the case of this embodiment, the phase is changed by +45 degrees (or -45 degrees) in the rising period and is changed by -45 degrees (or +45 degrees) in the falling period. This allows a very smooth transient response.

By the way, when a secondary circuit as shown in FIG. 9 is used as the waveform shaping circuit 14, due to the nature of the step response of the secondary circuit, the output power is turned off within the specified time at the falling edge of the burst.・ Level (eg -60
There is a problem that it does not attenuate to dBm).

That is, the time constant and Q of the secondary circuit rise,
The response at the time of falling is as smooth as possible to the final value,
Although it is selected to move without vibration, there is some residual voltage and vibration. At the time of start-up, there is a certain allowable range (± 3 dB) in the final output power value, so there is no problem. However, at the time of the fall, minute residual voltage, vibration, etc. exist until reaching the final value, which prevents these from reaching the carrier off level.

For example, in the digital cellular car telephone system adopted in North America, FIG.
3 symbols at the time of falling shown in (B) (= 120 μse
It is necessary to lower the output power to −60 dBm or less in c). However, in the fall response of the secondary circuit, several% of the amplitude remains even after passing 3 symbols, and it is not possible to fall to −60 dBm or less.

In such a case, it is preferable to provide a carrier-off circuit for forcibly setting the voltage to the control terminal 15 and the gain variable terminal 16 to 0V after a lapse of a certain time from the start of the fall.

FIG. 17 is a circuit diagram showing another embodiment of the present invention, which is the control terminal 1 of the linearized power amplifier shown in FIG.
5 and the variable gain terminal 16 are connected to carrier-off circuits 18 and 19, respectively.

FIG. 18 is a circuit diagram showing an embodiment of the carrier-off circuits 18 and 19, and its operation will be described below.
In FIG. 18, the control signal (SW2) has an inverse relationship with the timing of switching SW2 shown in FIG.
_{At 0} , the H level is changed to the L level, and at t ₂ , the L level is changed to the H level. First, consider the operation in the steady state. When the transmission burst is on, that is, the control signal (SW
When 2) is L level, the transistor (Tr) is turned off, and the collector voltage of Tr becomes H level. When the transmission burst is off, that is, when the control signal is at H level, Tr is turned on, so that the collector voltage of Tr is at L level, that is, 0V.
(There is strictly a residual voltage). When the collector voltage of Tr becomes 0V, the voltages of the control terminal 15 and the variable gain terminal 16 shown in FIG. 17 become 0V, and carrier off is executed.

Next, consider the operation in the transient state. Figure 1
As shown in FIG. 9 (A), when the burst rises (t ₁ ),
That is, when the control signal (SW2) changes from the H level to the L level, the base voltage of Tr discharges with a time constant τ ₁ determined by (R1 + Di on resistance) and C.
As shown in (B), it decreases from 0.6 V (V _{BE of} Tr) according to the time constant τ ₁ . When the voltage at the base of Tr falls below the base-emitter voltage V _BE , Tr is immediately turned off and the collector voltage becomes H level (open).
Therefore, the output power is 2 due to the action of the original linearization loop.
It rises according to the step-up curve of the next circuit 14.

On the other hand, when the burst falls (t ₂ ), that is, when the control signal changes from L level to H level, the base voltage of Tr is the time constant determined by (R1 + R2) and C shown in FIG. 19B. At τ ₂ , it changes toward the final target voltage value of 5V. However, when the base voltage of Tr reaches the voltage at which Tr is turned on (V _BE = 0.6V), Tr is turned on at that moment, and the collector voltage becomes L as shown in FIG. 19C.
Become a level. Note that the time t ₀ (delay time) from t ₂ to turning on Tr is t ₀ = 0. ₀ because the rise of the base voltage can be approximated by a linear line (V = 5t / CR).
It can be calculated by 6 × CR / 5. Therefore, (R1
By selecting + R2) and the value of C, the collector voltage of Tr can be set to ₀ V at the desired delay time t ₀ from the time when the burst falls (t ₂ ).

For example, the delay time t ₀ is 100 μs.
If it is selected to be about ec, the transmission output will fall smoothly from t ₂ of FIG. 11 by the step-down curve of the secondary circuit 14, and at t ₂ +100 μsec, the carrier-off circuit 18,
The above-mentioned operation of 19 causes the carrier to be rapidly and compulsorily lowered to the carrier-off level. That is, it can be said that the falling operation is divided into two steps. Main circuit is secondary circuit 1
4 is a function of the step-down response and the linearization loop, and after 100 μsec has passed, the carrier-off circuit 1
By 8 and 19, they are forcibly and rapidly lowered to the carrier-off level.

As described above, by using the carrier-off circuit together, it is possible to avoid the problem due to the residual of the step-down response of the secondary circuit.

In FIG. 17, the RF variable ATT1 is used.
Return route including 0, HPF8, LPF9 and LPF7
Even when the above is excluded, that is, when the band division type is not adopted, the same effect as described above can be obtained by the cut-off circuit connected to the control terminal 15.

Although the above-mentioned embodiment has been described with respect to the digital cellular car telephone system adopted in North America, the power amplifier device applied to other systems (not limited to the car telephone system), The present invention can also be applied to a transmission device (regardless of wireless cable) provided with a power amplification device.

The power-amplified QAM signal is also π /
It is not limited to the 4-shift DQPSK modulation signal. Even for other QAM signals, it is important to determine the data of the rising and falling periods so that the amplitude fluctuation due to the phase transition is minimized.

Furthermore, the data selection for the rising and falling periods of the QAM signal is effective even when the power amplifier device does not have a switching circuit.

[0083]

As described above, according to the first aspect of the present invention,
A switching circuit is provided so that the waveform-shaped input envelope signal is supplied to the comparator during the rising and falling periods of the transmission period, which enables smooth burst control at the rising and falling edges, and the spectrum It is possible to prevent the spread and prevent interference with other channels.

Further, according to the second aspect of the present invention, since the cut-off circuit is provided and the control terminal of the power amplifier and the like are forcibly held at zero voltage at the time of the burst fall, the power amplifier is brought to the carrier off level. Can be brought down quickly.

Further, according to the third aspect of the present invention, since the contents of the input signal at the time of rising and the time of falling are determined so that the amplitude fluctuation of the input signal given to the power amplifier is minimized, Burst control can be performed smoothly during downlink, spectrum spread can be prevented, and interference with other channels can be prevented.

[Brief description of drawings]

FIG. 1 is a circuit diagram of an embodiment of the present invention.

FIG. 2 is a circuit diagram of a conventional linearized power amplifier.

FIG. 3 is a diagram showing an example of control timing of a transmission burst.

FIG. 4 is an explanatory diagram of burst control.

FIG. 5 is an explanatory diagram of smoothing of conventional burst control.

FIG. 6 is an explanatory diagram of smoothing of conventional burst control.

FIG. 7 is a diagram showing frequency characteristics of HPF and LPF.

FIG. 8 is a circuit diagram of HPF and LPF.

FIG. 9 is a circuit diagram of a waveform shaping circuit.

FIG. 10 is a characteristic diagram of a step response of the waveform shaping circuit.

FIG. 11 is an operation timing chart of the switching circuit.

FIG. 12 is a circuit diagram of LPF7 and HPF8.

FIG. 13 is an explanatory diagram of pass band control of LPF7 and HPF8.

FIG. 14 is a spatial phase transition diagram of the π / 4 shift DQPSK modulation method.

FIG. 15 is an explanatory diagram of phase transitions every 45 degrees.

FIG. 16 is an explanatory diagram of phase transitions every 135 degrees.

FIG. 17 is a circuit diagram of another embodiment of the present invention.

FIG. 18 is a circuit diagram of a carrier-off circuit.

FIG. 19 is a diagram showing waveforms at various points in the carrier-off circuit.

[Explanation of symbols]

 1 Saturation type power amplifier 2 Output detector 3,6 Gain variable device 4 Comparator 7,9 LPF 8 HPF 10 RF variable ATT 11 Switching circuit 14 Waveform shaping circuit 25,26 Cut-off circuit

Claims (6)

[Claims] 1. A variable attenuator having a variable gain terminal for attenuating an input signal, a saturation type power amplifier having a control terminal for power amplifying the output of the variable attenuator, and an envelope signal of the input signal as a standard. And a feedback loop for inputting a DC component of the error signal of the envelope signal of the output signal from the power amplifier to the control terminal and an AC component to the variable gain terminal, and linearly amplifying the input signal. A power amplifying device for a transmitter using envelope feedback, wherein a switching circuit having a waveform shaping circuit for smoothing the envelope signal of the input signal is provided, and the switching circuit is provided in a transient period of burst transmission. The power amplifier device is characterized in that the envelope signal of the input signal via the waveform shaping circuit is input to a comparator in the feedback loop. . 2. The power amplification device according to claim 1, wherein a second-order low-pass filter is applied as the waveform shaping circuit. 3. A carrier-off circuit for forcibly setting the voltage at the control terminal and the variable gain terminal to zero for a certain period of time after a certain period of time has elapsed from the start of the fall of burst transmission. The power amplification device according to 1. 4. The device according to claim 1, further comprising means for inputting both the DC component and the AC component of the error signal to the control terminal and the gain variable terminal only during a burst transmission transient period. Power amplifier. 5. The power amplification device according to claim 1,
In a transmitter in which an input signal to the power amplifier is a QAM signal, transmission data in a transient period of burst transmission is data in which an amplitude fluctuation of the QAM signal is minimized. 6. The previous QAM signal is a π / 4 shift DQPS.
It is a K-modulated signal and is characterized in that the transmission data is such that the phase transition is +45 degrees (or -45 degrees) in the rising period of burst transmission and the phase transition is -45 degrees (or +45 degrees) in the falling period. The transmitting device according to claim 5.