JPH11275012A - External optical modulation system provided with advance distortion device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate nonlinearity in an input output characteristic of a nonlinear stage. SOLUTION: A signal is fed to at least one distortion circuit 2, a current is fed to the distortion circuit 2 along a first conduction path, the distortion circuit 2 provides distortion to the signal and at the applying of a current to the distortion circuit 2, the current is supplied to a 2nd conduction path in parallel with the first current conduction path to provide advance distortion to an electric signal.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and a circuit for removing non-linearity of input / output characteristics of a non-linear stage, particularly, third-order non-linearity, by giving a pre-distortion to a signal. About. More particularly, the present invention relates to a method and circuit for providing predistortion for an external light modulator.

[0002]

BACKGROUND OF THE INVENTION Optical signals can be modulated directly by acting on a light source, usually a laser, or indirectly using an external light modulator independent of the light source.

[0003] Amplitude of an optical signal by a radio frequency (RF) modulated signal having a very high frequency (eg, the frequency of a carrier of a television channel, such as in the case of CATV equipment typically in the range of 40 to 860 MHz). Optical modulators that can be used for modulation include, for example, devices based on a Mach-Zehnder interferometer built on lithium niobate (LiNbO ₃ ).

[0004] A characteristic required of an external modulator is modulation linearity. This is the C with analog type transmission.
It is very important for use in ATV equipment, especially if the modulated signal is not a single television channel but consists of a number of television channels, for example 40 to 80 channels. is there.

The electro-optical properties (output optical power as a function of input radio frequency voltage) of a modulator of the indicated type (Mach-Zehnder interferometer) are typically non-linear. In order to limit the distortion of the signal, it is preferable to operate the modulator as close to the part of the characteristic that is as linear as possible.

For this purpose, radio frequency (RF)
Is applied to the RF electrodes of a Mach-Zehnder interferometer-type electro-optic modulator, and a continuous voltage that determines the operating point of the modulator is applied to the same electrode or a second electrode.

[0007] An example of this type of modulator is the one marketed by the applicant under the name PIR PIM1510.

[0008] The modulated signal applied to the RF input comprises, for example, the entire set of carriers modulated by the television channel to be allocated to the user.

In the case of a Mach-Zehnder modulator, the variation in characteristics is similar to that of a sine wave, and when the modulator is close to the point of inflection of the sine wave, It is effective to operate with the applied operating point voltage VQ.

The modulation characteristic of a Mach-Zehnder modulator is related to the operating point and has the following relationship: Pu = Kz
It can be expressed by sinβ. Where Pu
Is the output light power, and Kz is a coefficient that depends on the characteristics of the Mach-Zehnder modulator. β = πV / Vπ is
Modulation index of the modulated signal in radians
dex). Here, V is a fluctuation of the applied voltage with respect to the operating voltage VQ, and Vπ is a constant.

This characteristic for sinusoidal variations is distinguished by two values. That is, (1) a voltage value called Vπ, which represents a voltage variation applied to an RF (radio frequency) electrode and changing the optical power from a maximum value to a minimum value;
(2) The value of the voltage VQ applied to the supply voltage that makes the operating point equal to the inflection point of the characteristic with sinusoidal variation, in other words, has an odd symmetry. In this case, the even-order distortion (consisting of the second harmonic of the applied signal) is cancelled, and the odd-order distortion takes a value that is closely defined.

For example, PIR PIM1510 manufactured by the applicant
In the case of a Mach-Zehnder modulator of the type, the above-mentioned voltages take the values Vπ = 4.3V and VQ = 0.7V.

The value of the operating point voltage VQ is not constant but varies with time (eg, due to static charge build-up in LiNbO ₃ ) and temperature.

The value of the operating voltage can therefore be determined, for example, by the presence and size of even-order distortion, in other words, CSO (composite
e Second Order) must be adjusted continuously by using as information the second order intermodulation products indicated as a whole.

Even when the modulator is operating at the above-mentioned operating point for minimizing even-order distortion, odd-order residual distortion (mainly 3 The following occurs and the intermodulation product or CTB (Co
mposite Triple Beats and cross modulati
on) and degrades the quality of the signal reaching the user. Indeed, the quality of a television channel with carrier amplitude modulation is greatly affected by the presence of the intermodulation products described above. Therefore, its overall level must be kept sufficiently low, for example, below the level of the vision carrier of each channel by more than 65 dB, thereby reducing the user High quality of the signal distributed to the

In order to limit these distortions to some extent, for example, per channel,
Choose a modulation depth that is not too large, approximately 3.5% to 4% (where "modulation depth"
Means the maximum value expressed as a percentage of the modulation index β), it is necessary to make the operation as close as possible to the linear part of the characteristic.

To limit the distortion caused by the non-linearity of the modulator characteristics, the maximum percentage of use of the modulator is typically around 40%. Allowing the above requirements relating to noise characteristics, the maximum number of channels that can be applied is calculated based on the assumption that various uncorrelated carriers are statistically added in quadrature. . Therefore, the sum of the channels can be considered as power. Consequently, the maximum number of channels that can be provided to the modulator is close to 100 in order not to exceed the percentages mentioned above.

In order to reduce residual distortion, especially third-order distortion, a predistortion is provided to the modulated signal using a non-linear element, so that this predistortion compensates for the subsequent distortion of the modulator. It has been proposed to. For this purpose, it is possible to connect before the modulator a distortion circuit whose input / output characteristics for radio frequency signals are inverse functions of the input / output characteristics of the modulator.

For the purposes of the present invention, a "distortion circuit" or "predistortion circuit" is an electrical circuit, especially a non-linear circuit, which, when an electrical signal is applied to its input, provides a predetermined signal to the input signal. A signal that is a non-linear function occurs at its output.

This method and examples of circuits used for this purpose are described, for example, in M. Nazarathy et al., "Progres.
ss in Externally Modulated AM CATV Transmission Sy
stem ", Journal of Lightwave Technology, vol. 11, n
o. 1, 01/93, pp. 82-104. In particular, this document shows a circuit using a diode biased by a current source as a non-linear element.

European Patent Application EP 0 620 661 and US Pat. No. 5,172,068 describe a distortion circuit comprising diodes connected in parallel and thus having opposite polarity to the input signal. The diode operates as a non-linear element at a predetermined bias value that is considered to be the operating point of the characteristic.

[0022]

The applicant has found that the reference circuit described above has several disadvantages.

In particular, Applicants have noticed that the amount of distortion caused by the reference circuit is, in some cases, insufficient to completely compensate for the distortion caused by the modulator.

Applicants have also observed that the performance of these circuits varies with variations in the amplitude of the input signal.

In particular, applicants note that as the amplitude of the applied signal increases, the distortion caused by this type of distortion circuit decreases relative to the theoretical value and the distortion caused by the modulator cannot be compensated. Noticed.

Applicants believe that this is due to the fact that the operating point of the diode varies with variations in the amplitude of the applied signal.

In particular, it is believed that even-order distortion produces a continuous component that modifies the operating point of the diode.

Applicants have also noted that for the same bias current in the distortion circuit, there is a significant discrepancy between the CTB and XMOD measurements.

[0029]

SUMMARY OF THE INVENTION In accordance with the present invention, by allowing the voltage across a diode to vary in a controlled manner, the voltage required in response to wide variations in the amplitude of an input signal is required. It has been found that it is possible to produce a certain amount of distortion.

In a first aspect, the invention relates to a circuit for pre-distorting an electrical signal. The circuit includes an input terminal suitable for receiving an electrical input signal of variable amplitude, an output terminal suitable for emitting an electrical output signal in response to the input signal, and an input terminal connected to the input terminal and the output terminal. Interposed and connected in series with at least a first
A first branch having a non-linear element and a first capacitance, at least a second non-linear element inserted between the input terminal and the output terminal and connected in series, and a second capacitance A second branch having:
Wherein the first branch and the second branch are connected in parallel, and a bias circuit electrically connected to the first nonlinear element and the second nonlinear element. A bias circuit, wherein the first direct current flows along the first conduction path through the first and second non-linear elements. And a second conduction path for a second direct current, which is in parallel with the second direct current.

In a second aspect, the invention relates to a method for predistorting an electrical signal. Providing the signal to at least one distortion circuit; supplying current to the distortion circuit along a first conduction path; and distorting the signal in the distortion circuit; Supplying the current to the distortion circuit comprises supplying a current to a second conduction path that is parallel to the first conduction path.

In a third aspect, the invention relates to a light emitter having external modulation. The light emitter comprises at least two biased non-linear elements, and is suitable for providing a modulated optical signal at its output, a predistortion circuit to which a modulated signal is applied, and for providing a predistorted signal. An optical-to-optical modulator having an electrical input to which a modulated signal is applied and an optical input; and a radiation source connected to the optical input of the modulator.
And at least one resistive element in parallel with the two non-linear elements.

In a fourth aspect, the invention relates to a method for predistorting an electrical input signal. Providing the signal to first and second distortion circuits;
Providing the first and second distortion circuits along a first conduction path; and applying the first and second distortion circuits to the signal.
Applying a distortion in the distortion circuit of
Combining the signal distorted by a second distortion circuit and providing the first and second distortion circuits, wherein the step of providing the first and second distortion circuits comprises: , And each time the electrical input signal increases by 1 dB, the ratio D3 /
C is increased by substantially 2 dB.

For the purpose of the present invention, the first and second
Supplying the distortion circuit means supplying an electric value of voltage or current to the first and second distortion circuits so as to bias them at a desired operating point.
In a fifth aspect, the invention relates to a method for transmitting an optical signal. The method comprises: predistorting a modulated signal comprising at least two non-linear elements; providing the two non-linear elements; providing light radiation to an electro-optic modulator; Modulating the light emission with the modulated signal provided, the step of providing the two non-linear elements controlling a voltage between terminals of the two non-linear elements; The step of obtaining certain CTB and XMOD values.

[0035]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, a light emitter having an analog external modulation will be described with reference to the block diagram of FIG.

Reference numeral 1 indicates an input for a radio frequency electric modulation signal included in a given frequency band. The input 1 is connected to the distortion circuit 2. The input / output characteristic (output voltage as a function of the input voltage) selected as described later is given to the distortion circuit 2. After the distortion circuit 2, an amplifier 4 is preferably provided,
The amplifier 4 operates in a frequency band of an electric modulation signal having a radio frequency. This is, for example, in the case of CATV systems in the range of 40 to 860 MHz.

The output 5 of the amplifier 4 is connected to the electrical input of an electro-optical modulator 7. The radiation source 9 is suitable for generating a continuous optical signal, but is connected via an optical fiber 8 to the optical input of the modulator 7. This radiation source consists of a laser, in particular of the semiconductor type. The modulator 7 sends the modulated optical signal to the output 6.

The input / output characteristics of the distortion circuit 2 are appropriately selected based on the input / output characteristics of the electro-optic modulator so as to compensate for the nonlinearity. This, in other words, is such that the relationship between the power of the optical signal at the output of the modulator and the voltage of the modulated signal at input 1 is as linear as possible. In particular, it is appropriately selected so as to minimize the third-order distortion of the modulator 7. The distortion device 2 generates a third-order distortion having the same amplitude and opposite sign as that generated by the nonlinear characteristic of the modulator 7 in a wide range.

The amplifier 4 is used to match the amplitude of the signal processed in the predistortion circuit with the amplitude of the signal required by the modulator to obtain the appropriate modulation depth.

To obtain the input / output characteristics of this type of distortion device 2, a diode or a transistor is used as a non-linear element.

In the example described, and preferably, the frequency band of terrestrial television broadcasting (40-860 MHZ)
In the case of the operation in z), the distortion circuit 2 uses a diode as a nonlinear element.

FIG. 2 shows a simplified circuit diagram of the distortion circuit 2. An input signal to be distorted is applied to input terminal 1. This signal is sent via capacitor C1 to two parallel branches. These branches each consist of one diode D1 and D2 and one capacitor C2 and C3. These diodes D1 and D2 are arranged such that their polarities are opposite to each other. Diode D
1 has a cathode connected to the capacitor C1, and an anode connected to one terminal of the resistor R1 and the capacitor C2. The other terminal of the resistor R1 is connected to a positive power supply voltage + V. The anode of the diode D2 is connected to the capacitor C1, and the cathode is connected to one terminal of the resistor R2 and the capacitor C3. The other terminal of the resistor R2 is connected to the negative power supply voltage -V.

The resistance Rp in series with the capacitance is
Preferably, it is connected in parallel with the two branches described above. However, these are not shown. This resistance Rp is
Provides greater freedom in circuit design. In particular, this resistor increases the level of the input signal (attenuated by the circuit including diodes D1 and D2) so that it is possible to use a subsequent amplifier stage 4 having a gain not exceeding 15-18 dB. It becomes possible.

The capacitor C2 and the capacitor C3 are connected to the output terminal 3, and a signal whose output terminal is distorted is generated.

The capacitors C1 to C3 are used for flowing a bias current only in the diode.

In FIG. 2, each branch has only one diode, but there may be multiple diodes depending on the required role.

The power supply voltages + V and -V must be selected such that their current / voltage characteristics bias the diodes D1 and D2 at the operating point where they have appropriate nonlinearities. In particular, by varying the operating point of the diode, it is possible to vary the size of the resulting distortion.

The circuit shown in FIG. 2 can mainly generate third-order distortion, but can also generate higher-order distortion. The size of these distortions is controlled by the values of the power supply voltages + V and -V. However, as a result of the symmetry of this circuit, second-order distortions are cancelled. For example, it is possible to use only one diode by removing the diode D2, the capacitor C3, the resistor R2 and the power supply voltage -V. In that case, this asymmetric circuit will also have second order distortion. In the text below, reference is made to circuits of interest of the type shown in FIG. However, those skilled in the art will be able to apply the principles disclosed in the following text to asymmetric circuits using only one diode.

Applicants have constructed a light emitter with external modulation, which is described in the following text with reference to the block diagram of FIG.

The electro-optical modulator 7 used is a model PIR PIM1510 Mach-Zehnder modulator manufactured by the applicant.

The radiation source 9 is driven by a laser,
It is composed of B type. The group constituted by the distortion circuit 2 followed by the amplifier 4 is shown in FIG. Compared to the simplified electrical schematic of the distortion circuit shown in FIG. 2, the complete electrical schematic of the distortion circuit shown in FIG. 3 also has an interface with the remaining elements of the light emitter. The connecting circuits described in the text below are shown.

The input signal applied to input 1 is sent to a T-type resistive attenuator 30, in other words, to a T configuration (10Ω, 120Ω, 10Ω) with about 3.5 dB of attenuation.
Next, a π-type resistive attenuator 31, in other words, about 9d
Π configuration with B attenuation (100Ω, 82.5Ω, 10Ω
0Ω). Then, it is sent to a T-type resistive attenuator 32 (5Ω, 270Ω, 5Ω) with an attenuation of about 1.5 dB.

The resistive attenuators 30, 31, 32 are used to properly adapt the impedance and signal levels between the various elements of the circuit.

In this case, the nonlinear circuit 33 includes six diodes D arranged in two branches connected in parallel.
1 to D6. Each branch is composed of three diodes D1, D2, D3 and D4, D5, D6, the polarity of the diodes D1, D2, D3 arranged in one branch being the diodes D4, D4, The polarity is opposite to that of D5 and D6. The diode used is preferably a MA4E976L type Schottky diode marketed by Macon, but other types of diodes can also be used, for example diodes with a low threshold voltage.

Each branch has one capacitor C
2 (100 nF) and C3 (100 nF) are in series with a diode that decouples the continuous component. The first terminal of the capacitor C2 is connected to the anode of the diode D1, the first terminal of the capacitor C3 is connected to the cathode of the diode D6, and the second terminals of the capacitors C2 and C3 are connected to each other.

The nonlinear circuit 33 is connected in parallel with the resistive attenuator 31 for the AC component of the signal. The capacitor C1 (100 nF) connects the junction between the anode of D4 and the cathode of D3, which forms the input terminal of the nonlinear circuit 33, to one end of the π-type resistive attenuator 31. The second terminals of C2 and C3 are connected to the other end of a π-type resistive attenuator 31 forming the output terminal of the nonlinear circuit 33. The above-described resistance Rp (82.5Ω) is
In this case, the resistor is connected in series with the input signal and includes a resistor of a π-type resistive attenuator 31. With the resistor Rp having an appropriate value, the level of the input signal 1 becomes
This allows it to rise above the level that would be required for a circuit that does not have this resistance. Further, the distortion circuit 33
Is considered equal to the non-linear resistance whose value depends on the bias current, the resistive attenuator 31
Is connected in parallel with this nonlinear resistor. Thus, the presence of Rp increases the linear component of the signal sent to the output of the distortion circuit as compared to that carried by the diode of the distortion circuit 33. The size of the non-linear component (primarily the third order) relative to the size of the linear component must be such that the modulator distortion is compensated (in other words, the ratio of the non-linear component to the linear component is applied Must have a particular value that depends on the amplitude of the input signal) (if Rp is present).
Need to be increased to obtain the ratio required by the modulator. In fact, for example, a third-order distortion component is much more pronounced (3) than a linear component that simply increases in a proportional manner.
It should be recalled that the next component increases (as a function of the amplitude of the input signal) in such a way that it increases in proportion to the cube of the amplitude of the applied signal. As the amplitude of the input signal increases, the output signal increases equally, the need for amplification by the amplifier 4 decreases, and the noise component also decreases.

A first variable potentiometer P1 (10k) having two lateral terminals and a variable central terminal
Ω), one side terminal is connected to the power supply voltage V + (12 V), and the other side terminal and the center terminal are connected to one side terminal of the second potentiometer P2 (50 kΩ). It is connected. The other side terminal of the potentiometer P2 is connected to a ground terminal. Potentiometers P1 and P2 have a function of dividing power supply voltage V + and generating a voltage Vp (about 2.7 V) at a connection point therebetween. Capacitor C4 (100 nF) is connected between this point and the ground terminal to stabilize voltage Vp. The voltage Vp is applied to the anode of the diode D1 via the resistor R1 (33 kΩ). The cathode of the diode D6 is connected to the ground terminal via the resistor R2 (33 kΩ).

The terminal at the center of potentiometer P2 is 2
It is connected between the two diode branches D1 to D6, in particular, between the cathode of the diode D3 and the anode of the diode D4 via a resistor R4 (33 kΩ).

Diode bias current (about 30 μA)
Are two large resistors R1 and R2 (about 33k
Ω). Capacitors C1, C2, C3
Have a value of about 100 nF each.

The diode bias current is supplied to the distortion circuit 2
The amount of distortion caused by the voltage is determined in the calibration stage by varying the voltage (Vp) using potentiometer P1.

Balancing the current in the two diode branches and minimizing the second-order distortion is regulated by potentiometer P2.

The circuit is optimized by appropriately selecting the values of the resistance of the π-type resistive attenuator 31 and the diode bias current to obtain the desired distortion.

Suitably, the amplifier 4 has suitable amplification values and suitable dynamic, linear and noise characteristics, but these are designed at the design stage by known methods not considered here. Should be considered. To equalize (equalize) the frequency response of the amplifier 4 and improve both the frequency response and the variation in distortion with frequency, RL
A C network (10 nH, 10Ω, 22 pF) is connected to its output.

The amplification is about 17 dB and is obtained using a CA922 amplifier marketed by Motorola.

The signal level to be applied to the input is 4%
-17 dB to obtain a modulation depth of 3.5%
Is -18 dB.

For CTB (according to the measurement method described in IEC Publication 728-1) and XM
With respect to OD (according to the NCTA measurement method), measurements were made on an optical modulator linearized using the distortion circuit described above, using 80 television channels, using a modulation index of 3.5%. . These measurements are shown in the following table.

[0067]

[Table 1]

The results of the measurements relate to the case where the value of CTB is minimized by varying the diode bias current. This value is clearly better than for XMOD.

It has been found that an improvement in the value of XMOD can be obtained by slightly (about 10%) reducing the value of the diode bias current. However, in this case, the value of CTB deteriorates.

The value of CTB and the value of XMOD are both 3
That basically depends on the size of the next distortion,
The applicant has found. The amount of pre-distortion that occurs, especially the amount of large third-order distortion, depends on the distortion device 2
By varying the bias current of the diodes D1 to D6. Therefore, CTB and XMO
Applicants predicted that the measurement of D would have a minimum for the same size of the resulting predistortion and, consequently, for the same diode bias current. However, the operating point determined by potentiometers P1 and P2 and minimizing the value of CTB is different from the operating point minimizing the value of XMOD.

Therefore, the distortion of the third harmonic of the distortion circuit is
As shown in FIG. 4, at a frequency of 200 MHz, the input signal level was varied from -15 dBm to +10 dBm, and the ratio D3 / C was measured. The input / output characteristics of the nonlinear device can be analytically approximated by, for example, the following type of expansion in a power series. That is,

[0072]

[Number 1] ^{Vu = K1Vi + K2Vi 2 + K3Vi} 3 + K4Vi 4 (1) where, Vu is the output signal, Vi is an input signal, K1, K2, K3, K4 are constants.

When the input signal is a sine wave, for example, in the following (2), the output is as shown in the following (3). However, in (3), the series is expanded up to the third order term.

[0074]

## EQU2 ## Vi = Vcosωt (2)

[0075]

[Number 3] Vu = K1Vcosωt + (1/2) K2V 2 cos2ωt + (1/4) K3V 3 cos3ωt + (3/4) K3V 3 cosωt + ·· (3) from the end of the equation (3), the third-order harmonic The distortion of
It is proportional to the cube of the amplitude of the input signal.
Each time the dB increases, the input signal increases by 3 dB.
It can be seen that when the dB increases, the ratio D3 / C increases by 2 dB.

The measurement of the ratio D3 / C caused by the distortion device 2 shown in FIG. 4 shows that the distortion is when the input signal is 1 dB.
When it does, it does not increase by 2 dB as expected, indicating that the amount of increase is slightly less.

In particular, the variation in distortion indicates that when a signal having a level higher than 0 dBm is applied, an increase of less than 2 dB occurs even if the level of the applied signal increases by 1 dB.

Such a phenomenon is due to the presence of any order, but mainly second order, even order distortion (because the amplitude is greater than for higher order distortions). From) that is the (1/2) K2V in equation (3)
Applicant believes that it produces a continuous component as described in section ² .

Referring to the circuit diagram of FIG. 3, according to the symmetric structure of the nonlinear circuit 33, at the connection point between the capacitors C2 and C3, in other words, the diodes D1-D
At the output of 6, the even-order signal components resulting from the two parallel branches compensate each other by canceling each other, while the continuous component is a diode D1-
Applicants have found that it algebraically adds to the bias current of D6.

Next, in the circuit shown in FIG. 3, the operating point of the diode fluctuates due to the presence of a continuous current component caused by even-order distortion, and in particular, the operating point voltage of the diode decreases. Applicant has found that By varying the operating point in this manner, the size of the generated odd-order distortion, particularly the magnitude of the third harmonic distortion, is reduced.

Consequently, as the level of the signal applied to the distortion circuit increases, the size of the operating point variation increases, and thus the increase in third harmonic distortion is shown by the curve in FIG. Is smaller than what would be expected theoretically.

Furthermore, since the television signal is amplitude-modulated with a modulation depth that varies with time, this is applied because, despite the fact that it is actually necessary, the above-mentioned effects exist. Applicants have found that it is not possible to keep the compensation for odd order distortions constant as the signal level varies.

In the circuit of FIG. 3, and in the circuit described in the above-cited document, diodes D1-D6
Are current biased. The battery voltage + V is applied to the diodes D1-D6 via the resistor network (P1, P2, R1, R2).

When diodes D1-D6 are biased by current or voltage generators and resistors (P1, P2, R1, R2), the voltages at the operating points of diodes D1-D6 will vary with changes in the input signal. Can fluctuate. For example, the current resulting from the second harmonic component is algebraically added to the bias current, and fluctuates the voltage at the operating points of the diodes D1 to D6.

According to the present invention, it is necessary to bias the diodes such that their operating points fluctuate in a controlled manner in order to eliminate the undesirable behavior as described above. Applicant has discovered.

One embodiment of a distortion device according to the present invention is shown in FIG. An additional resistor R3 connected between the anode of D1 and the cathode of D6 is provided in the non-linear circuit 33 to provide a conduction path for the continuous component. The resistor R3 is therefore connected to the diodes D1-D
6 are connected in parallel with each other.

The value of the resistor R3 can be determined as follows. A multi-carrier signal having an amplitude such that a modulation depth of 4% per carrier is obtained (for example, 40
To 860 MHz band and 80 carriers).
Is applied to the input of the circuit shown in FIG. In general, R
3 has a modulation depth greater than or equal to the desired maximum modulation depth and is determined using a signal composed of multiple carriers equal to or greater than the expected maximum number of carriers.

Potentiometer P1 is adjusted to bias diodes D1-D6 in a manner that substantially minimizes the value of CTB. If necessary, potentiometer P2 is also adjusted to minimize second order distortion.

Voltage V between terminals of diodes D1-D6
d (CTB) and the current I flowing through the diodes D1-D6
d (CTB) is measured.

The same operation is performed to minimize the value of XMOD. Then, again, the voltage Vd (XMOD) between the terminals of the diodes D1-D6 and the diode D1-D6
The current Id flowing through D6 (XMOD) is measured.

ΔI = I, the difference between the two measured currents
d (CTB) -Id (XMOD) and the difference between the two measured voltages, ΔV = Vd (CTB) −Vd (XMO
D) is calculated.

The value of the resistor R3 is given by the ratio between the absolute values of ΔV and ΔI. In the example shown in FIG.
Since V = 40 mV and ΔI = 20 μA, R3
Is 2 kΩ.

If the variation does not exceed 10% of this value,
It is not expected to cause unacceptable performance degradation for CTB and XMOD.

The values of the resistors R1 and R2 of the bias circuit of the diodes D1-D6 are set to 2 kΩ. These values must be chosen such that the bias voltage Vd is on the order of 2.5V to 3V. The reason is that the diode bias current and the current through R3 flow through these resistors.

These two resistance values are not important and only affect the final value of the voltage Vp required for the optimum bias operation. It has been found that the optimal bias voltage for Vp is about 2.7 V when the above resistance values are selected.

By measuring D3 / C by varying the level of the signal applied to the distortion circuit, a variation that is practically equal to a typical variation is obtained. In other words, as shown in FIG. 6, even for a signal level around +5 dBm or higher, if the input signal increases by 1 dB, D3 / C increases by about 2 dB (± 0.2 dB).

The CTB and XMOD measurements were performed on a complete circuit using 80 channels and a modulation index of 3.5%, yielding the results shown in the following table. The table shows how the two measurements for a single diode bias condition are in good agreement. For CTB values of −60 dB or less, especially −64 dB or less, in the frequency range considered, the XMOD measurement value is −
Less than 55 dB, preferably -60 dB or less, especially -64
Note that it is less than dB.

[0098]

[Table 2]

In the above example, the resistance R3 is preferably
Although connected in parallel with diodes D1-D6, modifications will be apparent to those skilled in the art based on the principles disclosed herein. For example, the resistor R in FIG.
Instead of 3, it is also possible to use a resistor connected in series with one or more diodes. Other modifications include, for example, a resistor connected in parallel with the diodes D1-D3 (or a series of one resistor and one or more diodes) and a resistor connected in parallel with the diodes D4-D6 (or (A series of one resistor and one or more diodes).

As still another modification, the resistor R3 may be replaced with a series resistor and an inductance or a variable potentiometer.

In these modified examples, similarly, there is a conduction path for the bias current in parallel with the diode, and the diode D1-D1 caused by the fluctuation of the amplitude of the input signal.
Voltage fluctuations between the terminals of D6 are to be compensated.

In the above embodiment, only one power supply voltage was preferably used, but by modifying the circuit in a manner obvious to those skilled in the art, a plurality of, for example, one positive and one negative, voltages could be used. It is possible to use a bias voltage. Thus, the performance of the two non-linear branches can be adjusted independently.

The principles of the present invention are also valid for other distortion circuits, such as those based on transistors (or transistors and diodes) arranged in a push-pull configuration as non-linear elements.

Further, the circuit shown in FIG. 2 further changes the operating points of the diodes D1 and D2, which are considered to be caused by the detection of the peak of the signal component by the diodes D1 and D2 combined with the capacitors C2 and C3. Applicants have noticed that Capacitor C2 having a high capacitance greater than the commonly used 100 nF, and preferably about 1 μF
And by using C3 to reduce or even avoid this problem by reducing the self-impedance to the AC component of the signal,
The applicant has found.

[Brief description of the drawings]

Further details of the present invention can be obtained from the foregoing description by referring to the following accompanying drawings.

FIG. 1 is a diagram of a light emitter with external modulation.

FIG. 2 is a simplified circuit diagram of a distortion circuit.

FIG. 3 is a complete circuit diagram of a distortion circuit.

FIG. 4 is a graph showing a variation in a measured value of the third harmonic distortion of the distortion circuit shown in FIG. 3;

FIG. 5 is a circuit diagram of a distortion circuit according to one embodiment of the present invention.

FIG. 6 is a graph showing a variation in a measured value of the third harmonic distortion of the distortion circuit shown in FIG. 5;

 ──────────────────────────────────────────────────の Continuation of front page (71) Applicant 591011856 Pirelli Cavies System e. p. A (72) Inventor Massimo Notaljacomo Biella, Italy, 13787 Candero, Via Albo 26 (72) Inventor Giuseppe Lavasio Bergamo, Italy, 24042 Capri Arte San Gervasio, Via de Gasperi 20 (72) Inventor Claudio Zanmarchi Milan, Italy 20100 Milan, Via U Ceva 29

Claims (23)

[Claims] 1. A circuit for predistorting an electrical signal, comprising: an input terminal suitable for receiving a variable amplitude electrical input signal; and a circuit suitable for emitting an electrical output signal in response to the input signal. An output terminal, a first branch inserted between the input terminal and the output terminal, and having at least a first nonlinear element and a first capacitance connected in series; and the input terminal and the output terminal. And a second branch having at least a second nonlinear element and a second capacitance connected in series and having a second capacitance, wherein the first branch and the second branch are connected in parallel. And a bias circuit electrically connected to the first nonlinear element and the second nonlinear element, wherein the first direct current is:
A bias circuit flowing through the first and second non-linear elements along a first conduction path, wherein the second direct current is in parallel with the first conduction path. Circuit comprising a second conduction path for: 2. The circuit for predistorting an electrical input signal according to claim 1, wherein said first nonlinear element comprises at least one diode. 3. The circuit for predistorting an electrical input signal according to claim 1, wherein said second nonlinear element comprises at least one diode. 4. The circuit for predistorting an electrical input signal according to claim 2, wherein said diode has an opposite polarity and is in series with said first direct current. A circuit characterized by the following. 5. The circuit according to claim 1, wherein said first direct current flows through said first nonlinear element and said second nonlinear element. . 6. The circuit for applying a predistortion to an electric input signal according to claim 1, wherein the electric input signal is electrically connected to the input terminal via a capacitance. 7. The circuit of claim 1, wherein said second conduction path comprises at least one resistor. 8. The circuit for predistorting an electrical input signal according to claim 7, wherein said second conduction path comprises at least one diode. 9. The circuit for applying a predistortion to an electric input signal according to claim 1, wherein the bias circuit has a positive bias terminal and a negative bias terminal. 10. The circuit for predistorting an electrical input signal according to claim 9, wherein said positive bias terminal is connected between said first nonlinear element and said first capacitance. Features circuit. 11. The circuit for predistorting an electrical input signal according to claim 9, wherein said negative bias terminal is connected between said second nonlinear element and said second capacitance. Features circuit. 12. The circuit for applying predistortion to an electric input signal according to claim 1, wherein said first and second nonlinear elements are connected in series with said first direct current. Circuit to do. 13. The circuit for applying a predistortion to an electrical input signal according to claim 12, wherein said first and second nonlinear elements form said first conduction path. 14. A method for predistorting an electrical signal, comprising: providing the signal to at least one distortion circuit; and supplying current to the distortion circuit along a first conduction path. Distorting the signal in the distortion circuit, wherein supplying an electric current to the distortion circuit supplies an electric current to a second conduction path that is parallel to the first conduction path. A method comprising the steps of: 15. The method of providing a predistortion to an electrical signal according to claim 14, wherein said signal is supplied to first and second distortion circuits; and a current is supplied along a first conduction path. Supplying the signal to first and second distortion circuits; distorting the signal in the first and second distortion circuits; and the signal distorted by the first and second distortion circuits Synthesizing the first conductive path and supplying a current to a second conductive path that is in parallel with the first conductive path. 16. The method for applying a predistortion to an electric input signal according to claim 14 or 15, wherein
Wherein the conductive path comprises a resistive path. 17. The method of providing predistortion to an electrical input signal according to claim 15, wherein said second conduction path is parallel to a series of said first and second distortion circuits. how to. 18. An optical emitter having an external modulation, comprising at least two biased nonlinear elements, a predistortion circuit to which a modulation signal is applied, and providing a modulated optical signal at its output An electro-optical modulator having an electrical input to which a predistorted modulated signal is applied and an optical input, and a radiation source connected to the optical input of the modulator. A light emitter, comprising: at least one resistive element in parallel with said two nonlinear elements. 19. The optical emitter according to claim 18, further comprising a power supply unit for supplying power to said two nonlinear elements and said resistive element. 20. The light emitter with external modulation according to claim 18, wherein said electro-optical modulator comprises a Mach-Zehnder interferometer modulator. 21. A method of providing predistortion to an electrical input signal, the method comprising: providing the signal to first and second distortion circuits; and providing the first and second distortion circuits with a first conduction path. Supplying the signals along the following directions; providing the signals with distortion in the first and second distortion circuits; combining the signals distorted by the first and second distortion circuits; Providing the first and second distortion circuits, controlling the voltage between the terminals of the first and second distortion circuits, wherein the electrical input signal is 1d
The method comprising, for each B increase, the ratio D3 / C increasing by substantially 2 dB. 22. A method of transmitting a modulated optical signal corresponding to an electrical modulation signal, wherein the modulation signal is predistorted by at least two non-linear elements, and thus the predistorted electrical signal is provided. Generating; providing the two non-linear elements; providing light radiation to an electro-optic modulator; and using the electro-optic modulator to modulate the light with the pre-distorted modulation signal. Modulating the radiation and thus providing a modulated signal, controlling the voltage between the terminals of the two non-linear elements, and controlling the −60 d in the modulated signal
Obtaining a CTB and XMOD value that is less than or equal to B. 23. The method of transmitting a modulated optical signal according to claim 22, wherein controlling the voltage comprises generating a current in parallel with the two nonlinear elements. .