WO2008107833A1

WO2008107833A1 - A method to improve a linearity of a rf transfer characteristic of a modulating rf amplifier

Info

Publication number: WO2008107833A1
Application number: PCT/IB2008/050758
Authority: WO
Inventors: Paulus T. M. Van Zeijl
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2007-03-07
Filing date: 2008-03-03
Publication date: 2008-09-12

Abstract

A method to improve the linearity of a transfer characteristic (6) of an envelope signal (2) to the amplitude of an output signal (3) of a modulating RF amplifier (10). The modulating RF amplifier (10) is arranged to provide an output signal (3) of which the amplitude is being modulated under control of the envelope signal (2). The modulating RF amplifier (10) is operable in a calibration mode (30) in which with a combination of a DC and RF measurement and an adjustment of a gain of the modulating RF amplifier the transfer characteristic (7) is determined. The modulating RF amplifier (10) is further operable in an operation mode (36) in which the envelope signal (2) is provided in dependence of the inverse of the transfer characteristic (7).

Description

A method to improve a linearity of a RF transfer characteristic of a modulating RF amplifier

FIELD OF THE INVENTION

The invention relates to a method to improve a linearity of a transfer characteristic of a modulating RF amplifier. The invention further relates to a modulating RF amplifier, a polar transmitter comprising a modulating RF amplifier and a device comprising a polar transmitter. Examples of such a device are mobile phones and wireless interfaces.

BACKGROUND OF THE INVENTION.

In 'A Novel Wideband Digital Power Amplifier and Transmitter Architecture for Multimode Handsets', Pierce Nagle et. al., Radio and Wireless Conference, 2004 IEEE a novel digital transmitter architecture is described comprising a digital power amplifier. The digital power amplifier is arranged to convert a received phase signal and amplitude signal to a modulated RF waveform. The digital power amplifier may therefore be referred to as a modulating RF amplifier. In the digital power amplifier a method to improve a linearity of a RF transfer characteristic is applied using digital optimization. In said method digital amplitude states of the amplitude signal corresponding with a required RF output current are mapped. With the mapping for a required RF output power the corresponding digital amplitude state of the amplitude signal is obtained. With said mapping it is possible to correct for a non-linearity in the transfer characteristic.

It is a problem that the non-linearity of the transfer characteristic will generally vary from modulating RF amplifier to modulating RF amplifier limiting the achievable improvement of the linearity of the transfer characteristic.

SUMMARY OF THE INVENTION

The invention is based on the insight that with the known method the achievable improvement of the linearity of the transfer characteristic is limited by the use of one mapping to correct for the non-linearity in the transfer characteristics of a plurality of modulating RF amplifiers. The invention is further based on the insight that because of cost it is undesirable to determine during production of a modulating RF amplifier the mapping corresponding to the transfer characteristic of said modulating RF amplifier. It is an object of the invention to provide a method to further improve the linearity of the transfer characteristic of a modulating RF amplifier.

This object is achieved with the method as defined in claim 1. The method according to the invention comprises steps in which for the modulating RF amplifier in the calibration mode measurements of the transfer characteristic of the envelope signal to the output signal with a DC and RF signal provided at the phase input are combined to determine the inverse of the transfer characteristic of said modulating RF amplifier. In the operation mode of the modulating RF amplifier the envelope signal is provided to the modulating RF amplifier in dependence of the inverse of the transfer characteristic of said modulating RF amplifier. For each modulating RF amplifier of a plurality of modulating RF amplifiers the inverse of its transfer characteristic is determined and used to correct for the non-linearity in the transfer characteristic. Thus the linearity of the RF transfer characteristic of said modulating RF amplifier is further improved.

An embodiment of the method as defined in claim 2 has the advantage that the sequence of steps in which the inverse transfer characteristic is determined will be repeated after detection of a predetermined change in a condition. After detection of said change in said condition the inverse of the transfer characteristic is updated. By updating the inverse of the transfer characteristic possible changes in parameters of the modulating RF amplifier in response to a change in said condition are taken into account. Examples of a condition that may influence the transfer characteristic are supply voltage or temperature. Another example of a condition that may impact the transfer characteristic is the load being coupled to the output of the modulating RF amplifier.

A further embodiment of the method as defined in claim 7 has the advantage that the transfer characteristic is stored in a conversion table providing for each one of the codes of the digital coded envelope signal a corresponding digital code for the value of the amplitude of the output signal. With the conversion table the inverse transfer characteristic is easily determined since the conversion table reveals what the digital code of the envelope signal must be in order to achieve a predetermined value of the amplitude of the output signal. In a further embodiment as defined in claim 8 the conversion table is stored in a history table after the predetermined change in the condition has been detected. Next in calibration mode the sequence of steps of the method of claim 1 is repeated resulting in an updated conversion table. By comparing the updated conversion table with the history table the sensitivity of the transfer characteristic as a function of said predetermined change in the condition may be determined. This has the advantage that knowing the sensitivity for said predetermined change in the condition a new content of the conversion table may be calculated in response to another change in said condition making a repeating in calibration mode of the sequence of steps of the method according to claim 1 after said another change unnecessary.

An embodiment of a modulating RF amplifier as defined in claim 11 has the advantage that the linearity of the transfer characteristic is improved. Unlike prior art the modulating RF amplifier as defined in claim 11 takes into account that the non-linearity will vary from modulating RF amplifier to modulating RF amplifier. To improve the linearity of the transfer characteristic said characteristic is measured and the inverse of said characteristic is determined for each modulating RF amplifier.

A further embodiment of the modulating RF amplifier as defined in claim 12 has the advantage that the instantaneous amplitude of the output signal is determined by a plurality of amplitude switches and a plurality of current sources under control of the envelope signal. The current delivered by the current sources is in dependence of the modulated phase signal. The envelope signal may be a digital signal comprising a plurality of bits. The plurality of amplitude switches are under control of one or more bits out of the plurality of bits of the digital coded envelope signal thereby simplifying the design. The impedance network may be optimized for the transfer characteristic of the modulating RF amplifier in operation mode whereas the transistor provides a simple means to adjust the gain of the transfer characteristic in the calibration mode.

The polar transmitter according to the invention is defined by comprising the modulating RF amplifier according to the invention and comprising a circuit for generating a phase/frequency signal and the envelope signal and further comprising an oscillator for receiving the phase/frequency signal and for generating the modulated phase signal.

An embodiment of a polar transmitter according to claim 14 or 15 comprising the modulating RF amplifier according to any one of claim 11 to 13 has the advantage of a reduced EVM (Error Vector Magnitude) and an improved ACPR (Adjacent Channel Power ratio) in the transmit signal.

In an embodiment as claimed in claim 16 the device comprises the polar transmitter as defined in claim 14 or 15. Examples of such a device are mobile phones and wireless interfaces. These and other aspects of the invention are apparent from and will be elucidated with reference to the embodiments described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS In the drawings:

Fig. 1 shows schematically a polar transmitter architecture, Fig. 2 shows schematically a transfer characteristic of a digital coded envelope signal to the amplitude of the output signal,

Fig. 3 shows schematically a transfer characteristic with a RF and a transfer characteristic with a DC signal provided at the phase input,

Fig. 4 shows schematically the steps of a method to improve a linearity of a RF transfer characteristic of a modulating RF amplifier

Fig. 5 shows an embodiment of a modulating RF amplifier, Fig. 6 shows schematically a further embodiment of a modulating RF amplifier,

Fig. 7 shows schematically a device according to the invention comprising a polar transmitter according to the invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS There are several known architectures for the transmission of signals, one of them being polar transmission. With polar transmission a signal to be transmitted is represented in the form of polar signals being an envelope signal r(t) 2 and a modulated phase signal phi(t) 1. The envelope signal r(to) provides an instantaneous amplitude at t=t₀, and the modulated phase signal phi(to) gives an instantaneous phase Re(e Λ^ωOtO+Phl(tOΛ^ _at t=t₀. The transmitted signal may be written as

s(t) = r(t) * Re(e^jla»^+pl"^W] )

Fig. 1 shows a simplified schematic diagram of the polar transmitter architecture. The polar transmitter architecture comprises a Voltage Controlled Oscillator and/or Phase Locked Loop 100 and a modulating RF amplifier 10. By modulation of a phase signal 4 provided to the Voltage Controlled Oscillator and/or Phase Locked Loop 100 the modulated phase signal phi(t) 1 is obtained and coupled to a phase input of the modulating RF amplifier 10. The envelope signal r(t) 2 is coupled to an envelope input of the modulating RF amplifier 10. The amplitude of the modulated phase signal phi(t) 1 is modulated under control of the envelope signal r(t) 2 resulting in an amplitude modulated output signal 3 that is radiated at an antenna 130.

A known problem of a modulating RF amplifier 10 relates to the non-linearity of a transfer from the envelope signal 2 to the amplitude of the output signal 3. The envelope signal 2 may be a digital signal comprising a plurality of bits. Fig. 2 shows schematically a transfer characteristic 6 of a modulating RF amplifier 10 from the envelope signal 2 to the amplitude of the output signal 3. On the horizontal axis the envelope signal comprising 8 bits is shown resulting in a range of codes from 000 to 256. On the vertical axis the amplitude of the output signal 3 is shown. The value of the amplitude of the envelope signal 2 is represented by a code value out of said range of codes. With increasing code values the amplitude of the output signal 3 increases but as shown in Fig. 3, for large code values the amplitude of the output signal 3 is no longer proportional to the code value.

The performance of a polar transmitter may be characterised by parameters known as EVM and ACPR. EVM or Error Vector Magnitude is a direct measure of modulation accuracy and transmitter performance, embodying all the transmitter impairments in a single number. ACPR or Adjacent Channel Power Ratio is a measure of the amount of interference or power in an adjacent channel related to the power in a transmitted frequency channel. The non-linearity in the transfer characteristic from code value to the amplitude of the output signal limits the minimal achievable EVM and contributes to the amount of interference or power in the adjacent channel. Clearly to achieve a minimal EVM and a maximal ACPR the transfer characteristic from code value to amplitude of the output signal 3 should be linearised.

A known method to correct a non-linear characteristic from an input signal to an output signal employs pre-distortion. Knowing the non- linear characteristic of a typical modulating RF amplifier the inverse characteristic is determined. Then, for each modulating RF amplifier said inverse characteristic is used to determine for a desirable output signal the corresponding input signal.

The inventor has recognized that the achievable improvement in the linearization will be limited due to variation in the non-linear characteristic from modulating RF amplifier 10 to modulating RF amplifier 10. Thus, to achieve an optimal linearization of the transfer characteristic from code value of the envelope signal 2 to the amplitude of the output signal 3 said transfer characteristic should be measured for each modulating RF amplifier 10 to determine its corresponding inverse characteristic instead of using the inverse characteristic of a typical modulating RF amplifier. Because of cost it is however undesirable to determine during production of a modulating RF amplifier for each modulating RF amplifier its corresponding inverse characteristic.

The inventor has further recognized that an inverse transfer characteristic of a modulating RF amplifier may be determined for each modulating RF amplifier by implementing a calibration mode 30 in each modulating RF amplifier. To prevent that such a calibration mode 30 is time consuming and to achieve an optimal linearization of the transfer characteristic a RF and a DC measurement should be combined. An advantage of a combination of a RF and a DC measurement is that the means to perform these measurements may be implemented on chip, allowing an integration of a method to improve a linearity of a RF transfer characteristic of a modulating RF amplifier.

Fig. 4 shows two flow diagrams of a method to improve a linearity of a transfer characteristic 6 of a modulating RF amplifier 10 as defined by claim 1. In said method the modulating RF amplifier 10 comprises a phase input and an envelope input and is arranged to provide an output signal 3. An amplitude of said output signal 3 is being modulated under control of an envelope signal 2 coupled to the envelope input. The left flow diagram in fig. 4 shows that the modulating RF amplifier 10 is operable in a calibration mode 30 in which a signal is provided at the phase input, and a value of a slope of the transfer characteristic 6, 7 of the envelope signal 2 to the amplitude of the output signal 3 is determined, an operation mode 36 in which the envelope signal 2 is provided in dependence of an inverse of the transfer characteristic.

When the modulating RF amplifier 10 is initialized 37, e.g. after power on, the modulating RF amplifier runs through the calibration mode 30 before entering the operation mode 36.

The right flow diagram of fig. 4 shows that the method comprises in the calibration mode 30 a first step 31 in which a RF signal is provided at the phase input and a first value of the slope of the transfer characteristic 6 of the envelope signal 2 to the amplitude of the output signal 3 is determined; a second step 32 in which a DC signal is provided at the phase input and a second value of the slope of the transfer characteristic 7 of the envelope signal 2 to the amplitude of the output signal is determined; a third step 33 in which the second value of the slope of the transfer characteristic 7 determined with the DC signal provided at the phase input is matched to the first value of the slope of the transfer characteristic 6 determined with the RF signal provided at the phase input by adjusting a gain of the amplifier 10; - a fourth step 34 in which a DC signal is provided at the phase input and the transfer characteristic 7 of the envelope signal 2 to the amplitude of the output signal 3 is determined; a fifth step 35 in which the inverse of the transfer characteristic 7 is determined. To explain the operation of the method to improve the linearity of the transfer characteristic of a modulating RF amplifier 10 as defined by claim 1 as an example a modulating RF amplifier 10 as shown in Fig. 5 is assumed. The modulating RF amplifier 10 of Fig. 5 comprises an impedance network 60 coupled between a power supply 30 and an amplifier output 5 , a plurality of amplitude switches 70 coupled between the amplifier output 5 and a plurality of current sources 80, wherein each one of the amplitude switches 70 is under control of the envelope signal 2 and each one of the current sources 80 is arranged to provide a current in dependence of the modulated phase signal 1. In the first step 31 of the method a RF or Radio Frequency signal is provided at the phase input and the value of the slope assumes a first value, as will be discussed later. The output signal 3 of the modulating RF amplifier 10 with a RF signal provided at the phase input may be written as:

V₁^ ₀₁₁, = R_load .f(code.I_ref )

In this formula is:

V_RF out the output signal 3 of the modulating RF amplifier 10

Rioad a resistance value of the impedance network 60 under the condition that a RF signal is provided at the phase input

I_ref a value of a current of the current source 80 code a digital code representation of the envelope signal 2 The function VRF out = Rioad-f(code.I_ref) represents the transfer characteristic 6 of the modulating RF amplifier 10 and may be approximated by:

V_RF __ou, = ^Ri_oad f(code.I_ref ) = R_load .code.I_ref + R_load .b.(code.I_ref f + R_load .c.(code.I_ref )³ + ...

In the second step 32 of the method a DC or Direct Current signal is provided at the phase input and the value of the slope assumes a second value, as will be discussed later. The output signal 3 of the modulating RF amplifier 10 with a DC signal provided at the phase input may be written as:

V_{DC out} = R_DC.f(code.I_ref )

R_DC is the resistance of the impedance network 60 under the condition that a DC signal is provided at the phase input.

In the third step 33 of the method the second value of the slope of the transfer characteristic 7 determined with the DC signal provided at the phase input is matched to the first value of the slope of the transfer characteristic 6 determined with the RF signal provided at the phase input by adjusting a gain of the transfer characteristic 7 of the amplifier 10. The slope of the transfer characteristic 6, 7 is defined as the coefficient of the first order term in the transfer characteristic 6, 7. Thus the second value of the slope of the transfer characteristic 7 determined with the DC signal is given by:

slope DC = R_DC.code.I_ref

And the first value of the slope of the transfer characteristic 6 determined with the RF signal is given by:

slope RF = R_load.code.I_ref

Next the second value of the slope of the transfer characteristic 7 determined with the DC signal must be matched to the first value of the slope of the transfer characteristic 6 determined with the RF signal. This may be realized by changing the gain of the transfer characteristic 7 of the modulating RF amplifier 10. In the modulating RF amplifier 10 of the example the gain may be changed by adjusting the value of R_DC- TO match the second value of the slope to the first value of the slope R_DC must be adjusted to the value of Rioad. As a further example the modulating RF amplifier 10 shown in Fig. 6 is assumed wherein each one of the current sources 80 comprises a first transistor 85 and each one of the amplitude switches 70 comprises a second transistor 75. The first and second transistor 85, 75 each comprise first and second main electrodes. The first main electrode of each one of the second transistors 75 is coupled to the second main electrode of one of the first transistors 85 and the second main electrode of each one of the second transistors 75 is coupled to the amplifier output 5.

In the modulating RF amplifier 10 shown in Fig. 6 one or more of the first transistors 85 may enter the triode region when the output 5 is handling a signal with a large amplitude. In the triode region the amplitude of the output signal is no longer proportional to the code value thereby providing a source of non-linearity in the transfer characteristic 6, 7 from code value to the amplitude of the output signal.

The weighting of the current provided by the first transistors 85 may be binary with the advantage that each one of the second transistors 75 is under control of a bit in the digital envelope signal 2. Also each one of the first transistors 85 may have equal scaling and provide an equal current with the plurality of second transistors 75 being controlled by a thermometer code, the thermometer code being in dependence of the envelope signal 2.

Assume that g_m is the transconductance of the smallest of the first transistors 85 and that said smallest transistor is providing a current I_ref under the condition that a DC signal is provided at the phase input. Under these assumptions the output signal 3 with the DC signal provided to the phase input may be written as:

V_DC__out = R_DC.code.I_ref + b.R_DC(code.I_ref f + c.R_DC.(code.I_ref f + ...

The output signal 3 with the RF signal provided to the phase input may be written as:

V*_{F ou}, = K_ad -code.g_m V_1n + b.R_load (code.g_m V_1n f + c.R_load.(code.g_m V_1n )³ + . To match the second value of the slope of the transfer characteristic 7 determined with the DC signal to the first value of the slope of the transfer characteristic 6 determined with the RF signal the value of R_DC is adjusted according to the formula:

ό m in

^RDC - ^Rload ^■ l ref

In the fourth step 34 a DC signal is provided at the phase input and the transfer characteristic 7 of the envelope signal 2 to the output signal 3 is determined. Once the value of R_DC has been adjusted to match the second value of the slope to the first value of the slope the coefficients b, c, ... of the transfer function 7 may be obtained with a DC measurement.

Fig. 3 shows: the transfer characteristic 6 of a modulating RF amplifier 10 with a RF signal provided at the phase input, the transfer characteristic 7 of a modulating RF amplifier 10 with a DC signal provided at the phase input wherein the value of R_DC has been adjusted to match the value of the slopes.

Fig. 3 further shows that for small values of code the transfer characteristic 6 determined with a RF signal provided at the phase input overlaps with the transfer characteristic 7 determined with a DC signal provided at the phase input. For large values of code there is no overlap indicating a remaining error.

In the fifth step 35 the inverse of the transfer characteristic 7 is determined.

In an embodiment of the method an analogue to digital converter 20, such as for example a sigma delta analogue to digital converter, may be used in the fourth step 34. The analogue to digital converter 20 is coupled to the output 5 of the modulating RF amplifier 10 and used to obtain for each one of the codes of the envelope signal 2 a digital code for a corresponding value of the amplitude of the output signal 3. Next for each one of the codes of the envelope signal 2 the digital code for the corresponding value of the amplitude of the output signal 3 is stored in a conversion table. The inverse of the transfer characteristic 7 is obtained by using the conversion table to provide for a desired amplitude of the output signal 3 a corresponding value of the code of the envelope signal 2.

In another embodiment of the method the analogue to digital converter 20 may be used in the fourth step 34 to obtain for each one of the codes of the envelope signal 2 a corresponding value of the output signal 3. With the value of R_DC adjusted to have matching values of the slopes the value of the coefficients b, c, .. of the transfer characteristic 7

V_DC__out = R_Dc -code.g_m.V_m + b.R_DC (code.g_m.VJ² + c.R_DC. (code. g_m . V_1n f + ...

may be determined in a program running on a microprocessor. In the program the values of the coefficients in the equation for V_{DC OUT} is chosen such that said equation for V_{DC OUT} fits within a predetermined error on the transfer characteristic 7 determined in the fourth step 34 with the analogue to digital converter 20. Once the value of the coefficients b, c, ... has been determined the inverse of the transfer characteristic 7 is obtained using said microprocessor by determining for a desired amplitude of the output signal 3 a corresponding value of the code of the envelope signal 2.

In a method to improve a linearity of a transfer characteristic 6 of a modulating RF amplifier 10 as defined by claim 2 the method according to claim 1 further comprises in operation mode 36 a step in which after detection of a change in a condition 38, said change being larger than a predetermined value the modulating RF amplifier 10 is put in the calibration mode 30, a sequence of the first step 31, second step 32, third step 33, fourth step 34 and fifth step 35 of the method as defined in claim 1 is performed, - the modulating RF amplifier 10 is put in the operation mode 36.

In practice the transfer characteristic 6 may change in response to a change in a condition 38 of the modulating RF amplifier 10. When such a change occurs it may be necessary to repeat a calibration by performing the sequence of the first to the fifth step 31-35 of method claim 1 to prevent that the non-linearity of the transfer characteristic 6 exceeds a certain value. The left flow diagram of fig. 4 shows that when a change in a condition 38, said change being larger than a predetermined value is detected the modulating RF amplifier 10 is put in calibration mode 30.

Examples of conditions that may determine the transfer characteristic 6 of the modulating amplifier 10 are a supply voltage provided to the modulating RF amplifier 10, a temperature of the modulating RF amplifier 10 or a load coupled to the output 5 of the modulating amplifier 10.

A temperature sensor may be used to measure the temperature. In case the modulating RF amplifier is integrated on a die it is advantageous to integrate the temperature sensor with the modulating RF amplifier on the same die. When a change in a condition 38 is detected, said change being larger than a predetermined value, the sequence of the first to the fifth step 31-35 of method claim 1 are performed and an update of the inverse of the transfer characteristic 7 is obtained using one of the previously discussed embodiments. In one embodiment after performing the sequence of the first to the fifth step of method claim 1 for each one of the codes of the envelope signal 2 the digital code for the corresponding value of the amplitude of the output signal 3 is stored in a conversion table. The inverse of the transfer characteristic 7 is obtained by using the conversion table to provide for a desired amplitude of the output signal 3 a corresponding value of the code of the envelope signal 2. In a method as defined in claim 8 after detection of a change in a condition 38, said change being larger than a predetermined value, the content of the conversion table is stored in a history table prior to performing the first step 31 of the calibration mode 30. After performing the sequence of the first to the fifth step 31-35 of method claim 1 a content of the conversion table is updated. A method as defined in claim 9 further comprises a step in which the content of the conversion table is compared with the content of the history table. With the data from the history table and the conversion table the sensitivity of the transfer characteristic 7 to a change in said condition may be determined.

A method as defined in claim 10 further comprises a step in which after detection of another change in said condition with the calculated sensitivity of the transfer characteristic 7 a new content of the conversion table is calculated.

A modulating RF amplifier 10 as defined in claim 11 comprises a phase input and an envelope input. The modulating RF amplifier 10 is arranged to provide an output signal 3 of which the amplitude is being modulated under control of an envelope signal 2 coupled to the envelope input. The modulating RF amplifier 10 is operable in a calibration mode 30 in which a signal is provided at the phase input, and a value of a slope of the transfer characteristic 6, 7 of the envelope signal 2 to the amplitude of the output signal 3 is determined, and an operation mode 36 in which the envelope signal is provided in dependence of an inverse of the transfer characteristic 7.

The modulating RF amplifier 10 as defined in claim 11 has the advantage that the linearity of the transfer characteristic 6 is improved. Unlike prior art the modulating RF amplifier 10 as defined in claim 11 takes into account that the non-linearity will vary from modulating RF amplifier to modulating RF amplifier. To improve the linearity of the transfer characteristic said characteristic is measured and the inverse of said characteristic is determined for each modulating RF amplifier. Therefore the modulating RF amplifier 10 as defined in claim 11 further comprises means to provide a RF signal at the phase input and determine a first value of the slope of the transfer characteristic 6 of the envelope signal 2 to the amplitude of the output signal 3; means to provide a DC signal at the phase input and determine a second value of the slope of the transfer characteristic 7 of the envelope signal 2 to the amplitude of the output signal 3; - means to adjust a gain of the modulating RF amplifier 10 to match the second value of the slope of the transfer characteristic 7 to the first value of the slope of the transfer characteristic 6, means to provide a DC signal at the phase input and determine the transfer characteristic 7 of the envelope signal to the amplitude of the output signal; - means to determine the inverse of the transfer characteristic 7.

Fig. 5 shows an example of a modulating RF amplifier 10 as defined in claim 11 wherein the modulating RF amplifier 10 comprises an impedance network 60 comprising a first terminal and a second terminal, the first terminal being coupled to an amplifier output 5, - a plurality of amplitude switches 70, each amplifier switch 70 being coupled between the amplifier output 5 and a current source 80, each of the amplifier switches 70 being under control of the envelope signal 2, a plurality of current sources 80, each current source 80 being coupled to an amplitude switch 70, each current source 80 being under control of a modulated phase signal 2 provided at the phase input.

Said modulating RF amplifier 10 comprises means to provide a RF signal at the phase input resulting in the value of the slope to assume a first value, as will be discussed later. The output signal 3 of said modulating RF amplifier 10 with a RF signal provided at the phase input may be written as:

V_BF__aΛ = R_bai-f{codeJ^)

In this formula is:

V_RF out the output signal 3 of the modulating RF amplifier 10 Rioad a resistance value of the impedance network 60 under the condition that a RF signal is provided at the phase input

Iref a value of a current of the current source 80 code a digital code representation of the envelope signal 2 The function VRF out = Ri_Oad.f(code.I_ref) represents the transfer characteristic 6 of the modulating RF amplifier 10 and may be approximated by:

V_{RF 0U1} = K_ad f(code.I_ref ) = R_load.code.I_ref + R_load.b.{code.I_reff + R_load.c.(code.I_ref f + ...

Said modulating RF amplifier 10 comprises means to provide a DC signal at the phase input resulting in the value of the slope to assume a second value, as will be discussed later. The output signal 3 of the modulating RF amplifier 10 with a DC signal provided at the phase input may be written as:

V_{DC out} = R_DC.f(code.I_ref )

R_DC provides a means to adjust the gain of the transfer characteristic 7 of the modulating RF amplifier 10 under the condition that a DC signal is provided at the phase input. In the modulating RF amplifier 10 as defined in claim 12 the means to adjust the gain of the transfer characteristic 7 with the DC signal provided at the phase input comprises a transistor 40. The transistor 40 has a main conductive path coupled between the second terminal of the impedance network 60 and a power supply 30. The transistor 40 has a control terminal 41 arranged for controlling a current flow in the main conductive path. An example of said transistor 40 is a MOS transistor or MOST, well known in the art. With a MOST 40 a resistance of the main conductive path between a source and a drain terminal may be controlled with a voltage provided at a gate terminal 41. With said MOST 40 a value of R_DC may be adjusted by controlling the voltage provided at the gate terminal 41.

In another embodiment as defined in claim 13 the means to adjust the gain of the modulating RF amplifier 10 may further comprise at least one resistor 50 coupled between the amplifier output 5 and the power supply 30. In this case the value of R_DC is determined by the parallel combination of said at least one resistor 50 and a resistance of the main conductive path of the transistor 40. In another embodiment the means to adjust the gain of the modulating RF amplifier 10 may further comprise a network of a plurality of resistors and a plurality of switches. The network provides a current path in parallel with the main conductive path of the transistor. With the plurality of switches a value of R_DC may be adjusted. The second value of the slope of the transfer characteristic 7 determined with the DC signal provided at the phase input may be matched to the first value of the slope of the transfer characteristic determined 6 with the RF signal provided at the phase input by adjusting the resistance value of R_DC- The value of the slope of the transfer characteristic 6, 7 is defined as the coefficient of the first order term in the transfer characteristic. Thus the second value of the slope of the transfer characteristic 7 determined with the DC signal is given by:

slope DC = R_DC.code.I ref

slope _RF = R_load.code.l ref

wherein Ri_oad is the value of the impedance Zi_oad for the frequency of the RF signal. In operation mode 36 as well as in calibration mode 30 when the first value of the slope of the transfer characteristic 6 is determined with the RF signal provided at the phase input a voltage is provided at the control terminal 41 of the transistor 40 to obtain a resistance R_DC of the main conductive path of the transistor that is much smaller than Ri_oad. The impedance network 60 may comprise a parallel coupling of a resistor, an inductor having an inductance and a capacitor having a capacitance. The value of the inductance and capacitance are in dependence of a frequency f₀. The inductance value and the capacitance value are chosen such that the impedance network 60 acts as a resistive pull up for a RF signal with frequency f₀ provided at the phase input. The impedance network 60 will attenuate signals with other frequencies than fo. This gives the advantage that the modulated phase signal 1 is filtered and the signal with frequency fo will have the largest amplitude at the output 5. With a DC signal provided at the phase input the inductor causes the resistance value of the impedance network 60 to be low. Thus for a RF signal provided at the phase input of the modulating RF amplifier 10 the gain of the transfer characteristic 6 is determined by the impedance network 60 whereas for a DC signal provided at the phase input of the modulating RF amplifier the gain of the transfer characteristic 7 is determined by the resistance of the transistor 40, the transistor 40 in parallel with the at least one resistor 50 or the network comprising a plurality of switches and resistors coupled between the second terminal of the impedance network 60 and the power supply 30.

The means to determine the inverse of the transfer characteristic 7 may comprise an analogue to digital converter 20 to obtain for each one of the codes of the envelope signal 2 a corresponding value of the amplitude of the output signal 3. The means may further comprise a memory 21 to store for each one of the codes of the envelope signal 2 the corresponding value of the amplitude of the output signal 3 in a conversion table. The inverse of the transfer characteristic 7 is obtained by using said conversion table to provide for a desired amplitude of the output signal 3 a corresponding value of the code of the envelope signal 2. Fig. 7 shows a polar transmitter 110 as defined in claim 14 comprising the modulating RF amplifier 10 as defined in any one of claims 11 to 13. The polar transmitter further comprises a circuit 90 for generating a phase/frequency signal 4 and the envelope signal 2, an oscillator 100 for receiving the phase/frequency signal 4 and generating a modulated phase signal 1. The modulated phase signal 1 is coupled to the phase input. In an embodiment said polar transmitter 110 further comprises an analogue to digital converter 20 and a memory 21 arranged for storing a conversion table. For each one of the codes of the envelope signal 2 a digital code of the corresponding value of the amplitude of the output signal 3 is stored in the conversion table. The inverse of the transfer characteristic 7 is obtained by using the conversion table to provide for a desired amplitude of the output signal 3 a corresponding value of the code of the envelope signal 2.

In a further embodiment the polar transmitter 110 further comprises an analogue to digital converter 20, such as for example a sigma delta analogue to digital converter. The analogue to digital converter 20 is coupled to the output 5 of the modulating RF amplifier 10 and used to obtain for each one of the codes of the envelope signal 2 the digital code for a corresponding value of the output signal 3.

In the discussed embodiments the DC signal may comprise a low frequency signal having a value that remains substantially equal during the calibration mode 30. The DC signal may also be a signal that is active or switched on in the calibration mode 30 and is inactive or switched of in the operations mode. Through the use of the method to improve the linearity of a transfer characteristic 6 of a modulating RF amplifier the output power and efficiency of the polar transmitter may be increased, while still fulfilling EVM and ACPR requirements.

The principle of polar transmission may be used wireless standards such as GSM-EDGE, DCS 1800, CDMA2000, WB-CDMA, WLAN 802.1 la/b/g, WIMAX 802.16, Bluetooth Medium-Rate.

A device 120 as defined in claim 16 comprises the polar transmitter 110 as defined in claim 13 or 14.

Devices 120 may be used in cellular-phones, cordless-phones, laptops and/or desktops using special cards for wireless connection to the internet, wireless base-stations, etc..

A polar transmitter 110 as defined in claim 14 or 15 may further be used in a communication system. Said communication system comprises a transmitter and a receiver and may further comprise at least one device 120 as defined in claim 16. An example of a communication system comprises a laptop having a wireless connection to the internet and a server connected to the internet wherein the laptop and server are exchanging data.

Claims

CLAIMS:

1. A method to improve a linearity of a transfer characteristic (6) of a modulating RF amplifier (10) comprising a phase input and an envelope input, the modulating RF amplifier (10) being arranged to provide an output signal (3) of which an amplitude is being modulated under control of an envelope signal (2) coupled to the envelope input, the modulating RF amplifier (10) being operable in a calibration mode (30) in which a signal is provided at the phase input, and a value of a slope of the transfer characteristic (6, 7) of the envelope signal (2) to the amplitude of the output signal (3) is determined, an operation mode (36) in which the envelope signal (2) is provided in dependence of an inverse of the transfer characteristic (7), the method comprising in the calibration mode (30) a first step (31) in which a RF signal is provided at the phase input and a first value of the slope of the transfer characteristic (6) of the envelope signal (2) to the amplitude of the output signal (3) is determined; - a second step (32) in which a DC signal is provided at the phase input and a second value of the slope of the transfer characteristic (7) of the envelope signal (2) to the amplitude of the output signal (3) is determined; a third step (33) in which the second value of the slope of the transfer characteristic (7) is matched to the first value of the slope of the transfer characteristic (6) by adjusting a gain of the modulating RF amplifier (10); a fourth step (34) in which a DC signal is provided at the phase input and the transfer characteristic (7) of the envelope signal (2) to the amplitude of the output signal (3) is determined; a fifth step (35) in which the inverse of the transfer characteristic (7) is determined.

2. A method according to claim 1 further comprising in operation mode (36) a step in which after detection of a change in a condition (38), said change being larger than a predetermined value, the modulating RF amplifier (10) is put in the calibration mode (30), a sequence of the first step (31), second step (32), third step (33), fourth step (34) and fifth step (35) is performed, the modulating RF amplifier (10) is put in the operation mode (36).

3. A method according to claim 2, the condition being a supply voltage provided to the modulating RF amplifier (10).

4. A method according to claim 2, the condition being a temperature, the temperature being measured by a temperature sensor, the modulating RF amplifier (10) being implemented on a silicon die, the silicon die further comprising the temperature sensor.

5. A method according to claim 2, the condition being a load coupled to an output (5) of the modulating RF amplifier (10).

6. A method according claim 1 or 2, wherein the envelope signal (2) is a digital coded signal and wherein in the fourth step (34) for each one of the codes of the envelope signal (2) a digital code for a corresponding value of the amplitude of the output signal (3) is determined.

7. A method according to claim 6 wherein in the fifth step (35) the inverse of the transfer characteristic (7) is determined using a conversion table, wherein the conversion table stores for each one of the codes of the envelope signal (2) a digital code for the corresponding value of the amplitude of the output signal (3).

8. A method according to claim 7, wherein after detection of the predetermined change in the condition and before performing the first step (31) in the calibration mode (30) a content of the conversion table is stored in a history table.

9. A method according to claim 8 further comprising a step in which the content of the conversion table is compared with the content of the history table and a sensitivity of the transfer characteristic (7) for the predetermined change in said condition is determined.

10. A method according to claim 9 further comprising a step in which after detection of another change in said condition with the determined sensitivity of the transfer characteristic (7) a new content of the conversion table is calculated.

11. A modulating RF amplifier (10) comprising a phase input and an envelope input, the amplifier (10) being arranged to provide an output signal (3) of which an amplitude is being modulated under control of an envelope signal (2) coupled to the envelope input, the modulating RF amplifier (10) being operable in a calibration mode (30) in which a signal is provided at the phase input, and a value of a slope of the transfer characteristic (6, 7) of the envelope signal (2) to the amplitude of the output signal (3) is determined, and an operation mode (36) in which the envelope signal (2) is provided in dependence of an inverse of the transfer characteristic (7), the modulating RF amplifier (10) further comprising - means to provide a RF signal at the phase input and determine a first value of the slope of the transfer characteristic (6) of the envelope signal (2) to the amplitude of the output signal (3); means to provide a DC signal at the phase input and determine a second value of the slope of the transfer characteristic (7) of the envelope signal (2) to the amplitude of the output signal (3); means to adjust a gain of the modulating RF amplifier (10) to match the second value of the slope of the transfer characteristic (7) to the first value of the slope of the transfer characteristic (6), means to provide a DC signal at the phase input and determine the transfer characteristic (7) of the envelope signal (2) to the amplitude of the output signal (3); means to determine the inverse of the transfer characteristic (7).

12. A modulating RF amplifier (10) according to claim 11 further comprising an impedance network (60) comprising a first terminal and a second terminal, the first terminal being coupled to an amplifier output (5), a plurality of amplitude switches (70), each amplifier switch (70) being coupled between the amplifier output (5) and a current source (80), each of the amplifier switches (70) being under control of the envelope signal (2), a plurality of current sources (80), each current source (80) being coupled to an amplitude switch (70), each current source (80) being under control of a modulated phase signal (1) provided to the phase input, and wherein the means to adjust the gain of the transfer characteristic (7) with the DC signal provided at the phase input comprises a transistor (40) having a main conductive path coupled between the second terminal of the impedance network (60) and a power supply (30).

13. A modulating RF amplifier (10) according to claim 12 wherein the means to adjust the gain further comprise at least one resistor (50) coupled between the amplifier output (5) and the power supply (30).

14. A polar transmitter (110) comprising the modulating RF amplifier (10) as defined in any one of claims 11 to 13, wherein the polar transmitter (110) further comprises a circuit (90) for generating a phase/frequency signal (4) and the envelope signal (2), the polar transmitter (110) further comprising an oscillator (100) for receiving the phase/frequency signal (4) and generating a modulated phase signal (1), the modulated phase signal (1) being coupled to the phase input.

15. A polar transmitter (110) according to claim 14 wherein the envelope signal (2) is a digital coded signal, the polar transmitter (110) further comprising a memory (21) arranged for storing a conversion table, the conversion table storing for each one of the codes of the envelope signal (2) a digital code for the corresponding value of the amplitude of the output signal (3).

16. A device (120) comprising the polar transmitter (110) as defined in claim 14 or 15.

17. A communication system comprising the polar transmitter (110) as defined in claim 14 or 15.