The present invention relates to a transconductance stage with improved linearity, and more particularly, to a transconductance stage with low distortion for providing an output signal that is free from parasitic components linked to third order intermodulation products. The invention has applications in transmitters and receivers, and in particular, in communication equipment such as portable phones.
A transconductance stage, also called a transconductor, is an electronic device that converts an input voltage into an output current. The voltage can be a voltage referenced relative to a potential or to a differential voltage. In the same way, the current can be a differential current.
By naming the variations of the voltage of the transistor base v, and the variations in the collector current they produce i, the transconductance stage of
In this equation gm is the transconductance of the isolated transistor in the absence of degeneracy resistance. This is such that:
gm=IO/VT (2)
where IO is the quiescent current of the transistor, and VT is the thermal voltage. The thermal voltage VT is such that VT=kT/q, where T is the absolute operating temperature (expressed in Kelvin), k is Boltzmann's constant, and q is the electron charge.
When the transconductance stage is connected by its collector to a load impedance ZOUT, that is, by terminal 14 in the figure, a voltage gain GV is obtained such that:
GV=gmeq*ZOUT (3)
In a receiver or transmitter, the transconductance stage transmits the frequency components of a signal applied to it, but also other components among which include the intermodulation product components. These components, generated by the transconductance stage, are due in particular, to a linearity defect.
As an illustration, if a signal received by the stage comprises two frequency components F1 and F2, the output signal comprises the fundamental components F1 and F2, and also their harmonics 2F1,2F2, 3F1, 3F2, etc., of the second order intermodulation components of the type F1−F2 and F1+F2, as well as the third order intermodulation components of the type 2F1−F2 or 2F2−F1, for example.
The components of the intermodulation products, which are the parasitic components of the output signal, generally have low amplitudes compared to the components of the frequencies from which they are derived. Nonetheless, they are undesirable when their frequency coincides with the frequency of the desired signal.
For example, a frequency component F of low amplitude risks being in competition with a parasitic component of the identical 2F1−F2 type. If F−F2=F2−F1, these difficulties appear, in particular, in the domains such as that of Hertzian (i.e., radio waves) telecommunications, where certain channels, whose reception is very weak, risk being distorted by neighboring strong channels.
To augment the linearity of the transconductance stages, and thus reduce the amplitude of the parasitic components which may be generated, the stages are equipped with feedback. The feedback includes, for example, an emitter degeneracy resistor such as resistor 20 described in reference to
The gain in linearity is obtained at the expense of an equivalent transconductance or a lower voltage gain. Concerning this subject, one can refer to equations (1) and (3) above. Thus, to obtain an output signal of the same amplitude as that which would be obtained without feedback, the supply power has to be raised. This requires raising the quiescent current crossing the transistor or the supply voltage. However, it turns out that for portable communication equipment, such as cellular phones operating on a portable energy source (an electrical battery, for example), increased energy consumption has a very negative influence on autonomy.
Another feedback possibility directed at improving the linearity of a transconductance stage is shown in FIG. 2.
A feedback branch 22, connected between the input and output terminals, makes it possible to extract a fraction α of the output voltage from the input voltage. The equivalent transconductance gmeq of the stage of
The device described in the referenced U.S. patent has the advantage of reducing the third order components of the intermodulation product by significant proportions. However, a common base structure provides the device with a very low input impedance. Thus, means for adapting the impedance to a value of 50 Ω, normal for high frequency transmitters-receivers, would consequently reduce the transconductance significantly relative to an assembly of the same type as shown in
An object of the invention is to provide a transconductance stage which has none of the limitations of the devices described above.
Another object of the invention is to provide a transconductance stage with good linearity and low distortion, free from third order intermodulation product components, and having a high transconductance.
A further object of the invention is also to provide such a transconductance stage with low energy consumption.
Another object of the invention is to provide such a stage with a reduced influence on third order intermodulation components.
Yet another object of the invention is to provide a transconductance stage with an input impedance capable of being adapted easily to a value approaching 50Ω.
These and other objects, advantages and features of the invention are provided by a transconductance stage comprising at least one principal bipolar transistor having a base linked to an input terminal, a collector linked to an output terminal, and an emitter linked to a supply terminal through the intermediary of a degeneracy resistor.
At least one compensation bipolar transistor is connected in parallel to the principal transistor and linked to the supply terminal without going through the degeneracy resistor. The value RE of the degeneracy resistance of the principal transistor is chosen such that RE*I0>VT/2, where VT is the thermodynamic voltage and I0 is the quiescent current of the principal transistor. The choice of the degeneracy resistance RE is preferably made such that RE>>VT/2I0, for example, RE>10VT/2I0.
According to the invention, the compensation transistor is without degeneracy resistance when the electrical liaison resistance in the emitter of this transistor at the supply terminal is sufficiently weak to be neglected compared to the degeneracy resistance of the principal transistor. In other words, with r representing the value of a degeneracy resistance of the compensation transistor, the value r should be such that r<<VT/2I′0*I′0 is the quiescent current of the compensation transistor.
Moreover, it is understood that supply terminal means a terminal used for the polarization of transistors, that is, for setting their quiescent currents. The supply terminal can be a supply source potential, for example, or ground.
Based upon the choice of the degeneracy resistance of the principal transistor indicated above, the phase of the third order intermodulation product components, generated by the principal transistor and the compensation transistor, have opposite signs and oppose each other. The resulting amplitude of the third order components is thus lower than that of the third order components which each of the transistors considered separately would have generated.
A suitable polarization of the compensation transistor, and an adjustment of its quiescent current, makes it possible to generate third order harmonics with this transistor which are also equal in amplitude to those generated by the principal transistor. In this case, the third order harmonics of the two transistors not only oppose each other but are cancelled.
In a particular embodiment of the transconductance stage, an inductor links the principal transistor and the compensation transistor to the supply terminal. The inductor is connected in series with the degeneracy resistor between the emitter of the principal transistor and the supply terminal.
This inductor raises the input impedance of the stage. Its value can be chosen as a function of a desired input impedance, in such a way as to adjust this impedance closer to the usual value of 50Ω. An impedance adaptation can be made by associating a resistor or other suitable passive components at the base of the transistor.
According to the invention, the transconductance stage can further comprise an inductor, called a parallel inductor, connected in parallel to the degeneracy resistor of the principal transistor. The parallel inductor has a value LE such that:
LE<<RE/2πΔF and LE>>RE/2πF
F is a central operating frequency of the transconductance stage, and ΔF is the width of a band of frequencies capable of containing third order intermodulation product components generated by the stage.
The first condition indicated for choosing the value of the parallel inductance makes it possible for the inductor to operate like a short-circuit towards the supply terminal to filter the frequency components whose value corresponds to the chosen frequency band ΔF. These frequencies correspond to second order intermodulation components, of the type F1−F2 or F2−F1, with reference to the example chosen in the introductory part of the description.
The second order intermodulation components combine with the fundamental components to generate new third order components. The filtering carried out by the parallel inductor makes it possible to limit or to eliminate this phenomenon. The second condition for choosing the value LE of the parallel inductance makes it possible to provide the inductor with an impedance very much higher than that of the degeneracy resistance such that it does not disturb the value of this resistance at the operational frequencies around the value F.
The transconductance stage of the invention can be a simple stage or a differential stage. In the second case, it comprises first and second principal transistors and first and second compensation transistors connected in parallel respectively to the first and second principal transistors. The bases of the first and second principal transistors are linked respectively to the first and second input terminals forming a differential input. The collectors of the first and second principal transistors are linked respectively to the first and second output terminals. The emitters of the first and second principal transistors are linked respectively to a supply terminal through the intermediary of a first and second degeneracy resistor.
Moreover, the first and second compensation transistors are linked without degeneracy resistance to the supply terminal. The first and second degeneracy resistors of the principal transistors have the values RE1 and RE2 such that:
RE1*I1>VT/2 and RE2*I2>VT/2
The terms I1 and I2 refer to the quiescent currents of the first and second principal transistors and where VT refers to the thermodynamic voltage.
The criteria for selection of the degeneracy resistances for each part of the differential stage are the same as for the single stage described above. Preferably:
RE1*I1>>VT/2 and RE2*I2>>VT/2.
In the same way, the transconductance stage can be equipped with inductors for facilitating impedance adaptation of the inputs. The transconductance stage then comprises first and second inductors linking respectively the first and second principal and compensation transistors to the supply terminal. The first and second inductors are connected in series with the first and second degeneracy resistors between the emitters of the principal transistors and the supply terminal.
Furthermore, the transconductance stage can comprise an inductor of value L connected between the emitters of the principal transistors. The value of the inductance is chosen such that it presents a high impedance for the signal corresponding closely to the working frequency, so that it does not distort the operation for these frequencies. It is also chosen so that it has a low impedance for the components of the second order intermodulation product in order to filter them.
Considering that the degeneracy resistances of the first and second principal transistors are equal, both having the same value RE, and L can be chosen such that:
The values ΔF and F are the same as those taken into consideration above.
The invention relates not only to a transconductance stage but also to a transmission or reception stage comprising, between an antenna and a modulator or demodulator, a low noise amplifier and a frequency translation device equipped with a mixer, in which at least one of the mixers and amplifiers comprises a transconductance stage as described above. The invention also concerns the use of a transconductance stage in a portable phone.
Other characteristics and advantages of the invention will be understood from the following description. This is provided as a purely illustrative and non-limiting example.
In the following description, identical, equivalent or similar elements of the different figures are marked with the same reference numbers. The transconductance stage of
The transistor bases are connected to an input terminal 112 to which an input voltage Vin is applied. The transistor collectors are linked to an output terminal 114 for connecting to a load (not shown) for the stage. The current crossing this load is called Iout. The quiescent currents of the compensation transistor and the principal transistor are called, respectively, I1 and I0. These currents are fixed by the specifications of the transistors and possibly by polarization resistors (not shown).
The emitters of the transistors are linked to a supply terminal 116 which, in this figure, corresponds to ground. The emitter of the compensation transistor is connected directly to the supply terminal in such a way that it is not degenerate. The emitter of the principal transistor is connected to the supply terminal through the intermediary of a degeneracy resistor 120, of value RE. The resistor 120 can be formed from a single resistive component or can comprise several resistive components. As stated above, the value of the resistance RE is chosen such that RE*I0 is greater than VT/2, and may even be very much greater.
The phase of the harmonic of the third order intermodulation product, as far as the principal transistor is concerned, depends on the value of the degeneracy resistance. This phase reverses around a value of RE which is exactly VT/2I0. As an example, if the phase of the third order harmonic is 180° for a value zero or close to zero for the degeneracy resistance, it is 90° for a value RE=VT/2I0 and zero (0°) for a high value of RE compared with VT/2I0. Thus, the case of a phase equal to 180° corresponds to the compensation transistor whose emitter is not degenerate, whereas the case of a phase of 0° corresponds to the principal transistor.
Since the phases of the components of the third order intermodulation products are opposed, these components, coming from the principal transistor and the compensation transistor, cancel each other. When the amplitude of the third order components is almost the same for the two transistors, the compensation can attain complete elimination of these components. This ideal case can be approached, for example, by using transistors with almost identical specifications and by adjusting the quiescent current I1 of the compensation transistor.
TABLE I below provides, for comparison, the output amplitudes of a desired signal at a frequency of 2 GHz, and the amplitude measured in dBc relative to the amplitude of the fundamental, called Imd3, of the components of the third order intermodulation products for a transconductance stage according to
It can be seen from consulting TABLE I, that for almost identical quiescent currents (close to 26.8 μA), that is, for almost identical electrical consumption, the components of the intermodulation products undergo very high attenuation in the transconductance stage according to the invention (−103 dB instead of −65 dB).
In comparison, to obtain such an attenuation with the transconductance stage of the prior art, the value RE would have had to of been raised to 27.5Ω and the quiescent current I0 of the degeneracy resistor would have had to have been raised to 14.8 mA. These measures would thus have led to a significant increase in the consumption of electrical energy.
The specifications corresponding to the device of
The transconductance stage has two output terminals 114a and 114b which deliver the output currents Iout and Ixout. The dynamic currents must not be confused with the currents I1a, I1b, I0a and I0b shown in the figure. The currents I1a, I1b, I0a and I0b are the quiescent currents of the principal and compensation transistors. The stage input comprises two input terminals which, in
Although it is not described in detail here, the symmetrical transconductance stage can also be produced from PNP transistors. Concerning this, reference can be made to FIG. 3B and to the corresponding description.
When the transconductance stage is to be used in a transmitter or receiver, its input is adapted to a real impedance on the order of 50Ω. The impedance adaptation can take place, for example, by a series connection with the stage input of an appropriate resistance. However, the transconductance stage according to
The value of the inductance 118 can be chosen, for example, as a function of a transition pulse of the stage, in such a way that the real part of the input impedance is on the order of 50Ω. As an example, a value of 0.8 nH can be chosen.
TABLE II below demonstrates the influence of the inductance 118 in the transconductance stage of
In this table I1, I0, RE and L correspond respectively to the quiescent current of the compensation transistor 130, that of the principal transistor 110, the value of the degeneracy resistor 120, and the value of the inductor 118. It is evident that the real part of the input impedance is greatly improved.
As noted in the introductory part of the text, the signal comprises not only third order intermodulation products but also second order intermodulation products. The latter, combined with the fundamental components, are capable of generating supplementary third order components.
The value LE of the parallel inductor 122 is chosen such that it is transparent, that is, it has a very high impedance for the components corresponding to the fundamental frequencies F of the desired signal. It is also chosen to filter, that is, to present a low impedance for a frequency band ΔF corresponding to second order intermodulation. The orders of magnitude of the frequencies F and ΔF are very different. The fundamental frequencies F of the desired signal are on the order of 1 GHz, for example, whereas the intermodulation frequencies ΔF (for example, F2−F1) are on the order of 1 MHz.
As mentioned above, the parallel inductor 122 is thus chosen such that:
LE<<RE/2πΔF and LE>>RE/2πF.
The impedance adaptation inductors 118, 118a and 118b are shown in dotted lines in
Number | Date | Country | Kind |
---|---|---|---|
01 05369 | Apr 2001 | FR | national |
Number | Name | Date | Kind |
---|---|---|---|
4247825 | Berger | Jan 1981 | A |
4287478 | Berger | Sep 1981 | A |
5826182 | Gilbert | Oct 1998 | A |
5903185 | Cargill | May 1999 | A |
6011980 | Nagano et al. | Jan 2000 | A |
6556082 | Wang et al. | Apr 2003 | B1 |
Number | Date | Country |
---|---|---|
0613248 | Aug 1994 | EP |
Number | Date | Country | |
---|---|---|---|
20030006836 A1 | Jan 2003 | US |