The disclosure of Japanese Patent Application No. 2009-130299 filed on May 29, 2009 including the specification, drawings and abstract is incorporated herein by reference in its entirety.
The present invention relates to a method of suppressing variation of control voltage-output power characteristic of a multistage bias circuit with regard to high-frequency power of a wireless communication device, and particularly to a method of suppressing variations among products of control voltage-output power characteristic due to temperature variation or process variation.
Traditionally, power amplification modules equipped with a variable gain function that makes the output power variable by changing the bias condition of a multistage power amplifier have been widely applied to portable wireless communication devices. One of the problems with the multistage power amplifier is drastic change of control voltage-output power characteristic nearby the threshold voltage of a transistor.
The three-stage bias circuit includes a first bias circuit 1, a second bias circuit 2 and a third bias circuit 3, which are serially-connected. Matching circuits are coupled between respective bias circuits and at the output side of the third bias circuit 3.
A current mirror circuit is applied to the first bias circuit 1, the second bias circuit 2, and the third bias circuit 3 of the three-stage bias circuit. The bias circuit of each stage is controlled by the electric current output from an APC (automatic power control circuit) 4.
The APC 4 is a control circuit that outputs control current Ibias to the bias circuit of each stage as appropriate, after voltage to current conversion of the input Vapc.
That is, the APC of the three-stage amplifier bias circuit performs a linear current control of the first and second stages from the start of control. On the other hand, a control that increases the drain drive current ID later and more steeply than the first and second stages is performed on the third stage after a voltage exceeding a certain level has been added to the control voltage Vapc (see
Such a control brings about a steep variation at the start-up of the current because mutual conductance (gm) of an FET is proportional to the square root of the drain drive current ID, and the variation becomes approximately linear thereafter. By controlling the output Pout of the three-stage amplifier bias circuit in this linear section, it is possible to provide a high-frequency power amplification module having superior control characteristics.
Japanese Patent Laid-Open No. 2001-102881 (Patent Document 1) discloses a method of providing a multistage, high-frequency power amplification module comprising a plurality of serially-connected MOSFETs, which smoothens the gradient of the linear section to improve control characteristics thereof.
However, there remains a problem even if the control characteristics are improved by improving the linear section as disclosed in the patent document 1.
In other words, there exist variations among products at the control in a section accompanied with a steep variation before reaching the linear section, that is, at the start-up.
As can be seen in
The present invention has been made in view of the above circumstances and provides a bias circuit for gain control that can reduce gain variation at low-power output, facilitate setting of output power, and is unlikely to be affected by variation in element values and variations among products.
The other purposes and the new feature of the present invention will become clear from the description of the present specification and the accompanying drawings.
The following explains briefly the outline of a typical invention among the inventions disclosed in the present application.
A bias circuit according to a representative embodiment of the present invention includes a linear amplifier that amplifies current input from an input end; and a resistor coupled to a connection point between the input end and the linear amplifier and grounded via a diode-connected first FET, wherein the linear amplifier, together with a peripheral circuit thereof, composes a buffer amplifier including a threshold voltage offset, and the output of the buffer amplifier is coupled to the base terminal of a second FET.
In the bias circuit, influence of undesired components of the bias circuit may be excluded by setting current drivability coefficients of the first FET and the second FET.
In the bias circuit, part of current flowing from the connection point of the resistor and the first FET to the first FET may be drawn.
In the bias circuit, a current mirror circuit may draw part of current flowing in the first FET.
A high-power amplifier according to a representative embodiment of the present invention comprises a plurality of serially-connected bias circuits, wherein the bias circuit according to any one of claims 1 to 4 is used as the last-stage bias circuit of the serially-connected bias circuits.
A portable information terminal using the high-power amplifier falls within the scope of the present invention.
The following explains briefly the effect acquired by the typical invention among the inventions disclosed in the present application.
By applying the bias circuit according to representative embodiment of the present invention to a three-stage bias circuit which is an example of multistage power amplifiers, it becomes possible to vary the voltage added to the third stage gate voltage at low output linearly, and not in a square root manner. Sensitivity to the bias voltage of the output power is thereby suppressed, facilitating output power control using the bias voltage.
In addition, it becomes possible to reduce influence of device variation by modifying the current mirror circuit conventionally used in the bias circuit. Product yield can be improved thereby.
Embodiments of the present invention will be described below, referring to the drawings.
The bias circuit according to this embodiment includes a constant current source 101, an amplifier 102, a diode-connected FET 103, a resistor 104, a resistor 105, a resistor 106, and an FET 107.
The constant current source 101 plays the role of controlling the control bias current Icont to a constant current.
The FET 103 is coupled to the connection point between the constant current source 101 and the amplifier 102 in a branched manner. The FET 103, which is diode-connected, has a purpose of drawing a certain amount of the current flowing in the amplifier 102. The FET 103 and the resistor 104 determine the electric potential V1 at the connection point between the constant current source 101 and the amplifier 102. The electric potential at the connection point between the FET 103 and the resistor 104 is denoted as VG1.
The amplifier 102 includes a high-frequency power amplifier and peripheral circuits thereof, which amplify the signal that drew the current from the output of the constant current source 101 using the FET 103 or the like. The amplifier 102 is a linear amplifier.
The resistors 105 and 106 divide the voltage of the output of the amplifier 102. Based on the electric potential of the connection point between the resistors 105 and 106, the gate terminal of the FET 107 is controlled. The resistors 105 and 106 have a function of attenuating high frequency signals.
A buffer amplifier 10 having a threshold offset includes the amplifier 102, the diode-connected FET 103, the resistor 104, the resistor 105, and the resistor 106.
The FET 107 has an RF amplification function that amplifies and outputs high frequency signals that are input to the gate.
Using the principle diagram, how the third stage bias circuit works will be explained.
VG1 in the circuit can be calculated by the following equation.
In the above expression, Icont denotes the current value from the constant current source, and β1 denotes the current drivability coefficient of the FET 103. In addition, Vth denotes the threshold voltage (temperature-dependent) of the FET 103.
Calculation of VG1 also provides the electric potential V1 at the connection point between the constant current source 101 and the amplifier 102.
Here, R denotes the resistance value of the resistor 104.
The value of V1 is directly applied to the gate of the FET 107 by the buffer amplifier 10 (shown in
Based on the foregoing, ID can be calculated from the current drivability coefficient β2 of the FET 107 as follows.
The above-mentioned β1 and β2 can be calculated by the expression of a general current drivability coefficient.
Here, Cox denotes the oxidized film capacitance per unit area, μ denotes the electron mobility speed, W denotes the gate width, and I denotes the gate length.
Further discussion of (Equation 4) will be provided below.
In (Equation 4), the first term (β2×R2×Icont2) is the square-law characteristic. The second and third terms express undesired terms (undesired components). The undesired components become dominant over ID when Icont is small, in other words, immediately after start-up of the constant current source 101 of the third bias circuit 3-1, which is not desired.
The inventors conceived a method of excluding such undesired components from (Equation 4), and derived the present invention. That is, undesired components can be reduced by increasing the current drivability coefficient β1 of the FET 103 (=increase the size of the FET 103) to increase the denominator of the second and the third terms.
In other words, the current drivability coefficient β1 is increased by increasing each parameter in the numerator of (Equation 5) in the FET 103, and decreasing the gate length I which is a parameter in the denominator. Undesired components can be excluded thereby.
The high-power amplifier (HPA) of
The APC 4-1 supplies Icont according to the above (Equation 4). As for this supply, a current mirror circuit 14 in the APC 4-1 plays the role of the constant current source 101.
As has been already described above, this HPA has its characteristic in the third bias circuit 3-1. Additionally, the third bias circuit 3-1 includes, on as-is basis, the content of the principle diagram shown in
The steepness of the gain against the control voltage of the third bias circuit 3-1 can be suppressed thereby. An ideal bias circuit can be provided by increasing the current drivability coefficient β1 of the FET 103 in this manner.
A second embodiment of the present invention will be described next.
The object of the first embodiment is to exclude undesired components of the bias circuit by increasing the current drivability coefficient β1 of the FET 103 as described above. Attention has been paid on its feasibility by increasing each parameter in the numerator and decreasing the parameter in the denominator of (Equation 5).
However, these parameters depend on the physical size of devices such as the gate width W or the gate length I. Difficulty in design exists in that characteristic of the bias circuit largely depends on the physical size of the device.
Therefore, it is an object of the second embodiment to exclude undesired components of the bias circuit without depending on device size.
The adjustment circuit 201 draws current which is proportional to the input current from the current flowing in the diode-connected FET 103. Accordingly, the adjustment circuit 201 is coupled to the connection point between the diode-connected FET 103 and the resistor 104 in a branched manner.
A specific configuration of the adjustment circuit 201 will be described below.
The ratio of the drawn current to be drawn by the adjustment circuit 201 from the input current flowing in the FET 103 is defined as a (0<a<1).
First, operation of the third stage bias circuit will be explained using the principle diagram, similarly as the first embodiment.
VG1 in this circuit can be calculated by the following equation.
When VG1 is calculated, the electric potential V1 at the connection point between the constant current source 101 and the amplifier 102 can also be calculated, as with the first embodiment.
In this case, the voltage applied to the gate terminal of the FET 107 is given as follows.
Based on the foregoing, ID can be calculated as follows from the current drivability coefficient β2 of the FET 107.
Also in the second embodiment, the first term of the right side is the ideal component, whereas the second and the third terms are undesired components.
Now attention should be paid to variation of the entire bias circuit caused by the value of a of the adjustment circuit 201.
In each drawing, three parameters I, Ierror, and ID are described. I denotes the ideal component of the first term of the right side of (Equation 9). Ierror denotes the sum of the second and the third terms of the right side of the same equation. ID denotes ID of the left side of (Equation 9).
The smaller the numerator of the second and the third terms of (Equation 9), the smaller undesired components of ID become. Therefore, it is obvious from
Comparing the HPA of this embodiment with that of the first embodiment, the APC 4-2 and the third bias circuit 3-2 are different.
The third bias circuit 3-2 has two input terminals (here, the amplifier 102 drive voltage supply terminal is excluded) so that aIcont, as well as Icont from the constant current source 101 (included in the APC 4-2) can be input.
The third bias circuit 3-2 has a current mirror circuit 11 that processes the input aIcont. Using the current mirror circuit 11, aIcont can be drawn from the current flowing in the FET 103.
Furthermore, increase of processing speed can be expected by adding a phase compensation capacity 41 to the gate terminal of the FET included in the current mirror circuit 11. Responsiveness in the time axis actually required for a power amplifier can be secured thereby.
The APC 4-2 differs from the APC 4-1 in that the former outputs aIcont. The method of defining a uses pMOSFET21 and pMOSFET22 in the APC 4-2. The gate widths of these pMOSFETs are made completely the same. The ratio of current mirrors coupled below pMOSFET31 in the same APC 4-2 is set to be 1:1. Accordingly, a=1 holds because the subtracted current becomes equal to Icont. This causes the current to stop flowing in the FET 103, and therefore the value of a is determined by narrowing the gate width of the left pMOS by several percents, or setting the current mirror ratio to 1:0.95 or the like.
Influence of undesired components of the bias circuit can thereby be excluded without depending on the physical size of the device.
A third embodiment of the present invention will be described next.
In the second embodiment, undesired components of ID are excluded by drawing the current flowing in the FET 103 by the adjustment circuit 201.
In contrast, undesired components are made constant in the third embodiment of the present invention by further drawing an offset current from the adjustment circuit 201. It is an object to realize a more ideal square-law characteristic thereby.
The adjustment circuit 301 differs from the adjustment circuit 201 in that the input current and the drawn current are not controlled by the ratio, but by drawing current by a value obtained by subtracting a fixed offset value Ioffset from Icont.
Operation of the third stage bias circuit will be explained using this principle diagram.
VG1 in the circuit can be calculated by the following equation.
The electrical potential V1 at the connection point between the constant current source 101 and the amplifier 102 can also be calculated from this.
The voltage applied to the gate terminal of the FET 107 can be similarly calculated.
ID can be calculated from (Equation 12) and the current drivability coefficient β2 of the FET 107 as follows.
As can be seen, although influence of the second term cannot be completely excluded because Icont remains, influence of the second term can be minimized by the setting of Ioffset, and influence of the third term can also be handled. Handling of Ioffset during the control improves significantly because Ioffset is a constant.
Now we consider how to implement this processing in the circuit.
The APC 4-3 and the third bias circuit 3-3 are different in this embodiment. Addition of a constant current source 5 is also characteristic.
The constant current source 5 plays the role of generating IOffset that appears in the above (Equation 10) to (Equation 13).
The third bias circuit 3-3 includes current mirror circuits 12 and 13 in order to realize the principle of
In the current mirror circuit 12, IOffset from the constant current source 5 is input to the constant current source side (where the gate and the drain are connected) of the current mirror circuit. Because the current of the other terminal (output terminal) is set to IOffset by the current mirror circuit 12, input of the current mirror circuit 13 can be reduced.
Icont generated in the APC 4-3 is input to the input terminal of the constant current source side (where the gate and the drain are coupled) of the current mirror circuit 13. The terminal at the output side of the current mirror circuit 12 is also coupled to the connection terminal to which Icont is input. Accordingly, the effective current that is input to the current mirror circuit 13 is given by Icont−Ioffset. Therefore, the current mirror circuit 13 can draw the current of Icont−IOffset from a signal connected to the output terminal.
(Icont−IOffset) in the square root of (Equation 10) can thereby be generated so that this embodiment can be implemented.
Finally, a case that applies the HPA according to the above-mentioned embodiment to a cellular phone will be described.
The present power amplifier module is provided with two power amplifiers, namely, a Low Band in the 1 GHz band and a High Band in the 2 GHz band. In the drawing, the input terminal of the Low Band is TXin-Low Band, and the input terminal of the High Band is TXin-High Band. In addition, the High Band and the Low Band respectively have, within the antenna switch circuit, an LPF (Low Pass Filter) that removes the high harmonics, a switch that switches transmission and reception, and a diplexer circuit that integrates the signals of both the High Band and the Low Band.
The input provided to these terminals is first amplified by one of the above HPAs including the first bias circuit 1, the second bias circuit 2, and the third bias circuit 3. After amplification by the HPA, the input can be fed back to the control voltage Vapc of the power amplifier via a coupler 6, a detector circuit 50, and an error amplifier 51 to accurately output power to the output control voltage Vramp.
Although the detector circuit 50 and the error amplifier 51 are illustrated to process both the Low Band and the High Band collectively, there is no problem with dividing the process into two. Accordingly, two inputs may be provided for the output control voltage Vramp of the High Band and the output control voltage Vramp of the Low Band.
The HPA according to the present invention can thus be applied to a power amplifier.
Although the invention made by the inventors has been specifically described based on its embodiments, the present invention is not limited to the foregoing embodiments and it is needless to say that a variety of modification is possible without deviating from its spirit.
The present invention is expected to be applied to power amplifiers for portable information terminals. It is particularly expected that, through application of the present invention, a preferable bias circuit is provided to a power amplifier for GSM that controls the output using control voltage.
However, application of bias circuits of the embodiments for purposes other than those described above is not prohibited. In addition, although it has been assumed in the foregoing description to serially-connect bias circuits in three stages, there is no problem with using a single bias circuit alone. There is also no problem with serially-connecting bias circuits in more stages.
Number | Date | Country | Kind |
---|---|---|---|
2009-130299 | May 2009 | JP | national |
Number | Name | Date | Kind |
---|---|---|---|
6492872 | Fujioka et al. | Dec 2002 | B1 |
7345547 | Wang et al. | Mar 2008 | B2 |
Number | Date | Country |
---|---|---|
2001-102881 | Apr 2001 | JP |
Number | Date | Country | |
---|---|---|---|
20100301935 A1 | Dec 2010 | US |