Patent Application
20070014382
The invention relates to a transmission device and method for use in a transmission system, such as a cellular radio transmission system.
A power amplifier (PA) is a critical part of any radio transmitter. It amplifies the information-bearing RF (Radio Frequency) signal to a suitable power level for transmission. It is usually the last active section in the transmitter (TX) chain before the antenna. It typically also has the highest power consumption of any single part of the transmitter.
There are many different classes of power amplifiers. They can be distinguished from each other in terms of topology, or in the way in which they are driven or matched.
Most power amplifiers currently used in modern wireless communications are linear. This means that the input signal to the power amplifier is a fully modulated RF signal, containing all amplitude and phase modulation, already applied earlier in the transmitter. The power amplifier just provides gain, producing a ‘faithful copy’ of the input at the output, just at increased power.
“Class-A” refers to the most linear class of power amplifier, where the amplifier output follows the input waveform throughout the entire cycle of the RF input. This leads to the least distortion, but results in the least efficient class of power amplifiers—the power amplifier's bias current must be high enough so that the input RF signal never forces the transistor into a non-linear region, e.g., in the case of a bipolar type transistor, causes the device to go into saturation or cut-off.
By decreasing the conduction angle, through re-biasing the device so that the transistor is off for part of the input cycle, the efficiency of the power amplifier can be increased, but at the expense of non-linear distortion. Full conduction is Class-A. When conduction is 50% (i.e. only half of the input cycle is reproduced at the output) of the input cycle, the amplifier is in “Class-B”. When the amplifier is operating between these two classes then it is said to be “Class-AB”. Power amplifier designers try to achieve a trade-off between efficiency and non-linear distortion. The designer wishes the power amplifier to be as efficient as possible, while still meeting the wireless system spectrum requirements, e.g., adjacent channel leakage ratio, spectrum due to modulation, etc.
Designers also use various techniques to allow a linear power amplifier to operate with higher efficiencies, but with acceptable distortion. These include measures such as for example predistortion, adjustment of PA power supply with output power level and envelope tracking.
When conduction is at less than 50% of the input cycle then the amplifier is said to be operating in “Class-C”. This is an example of a fully non-linear amplifier. In the most efficient power amplifiers, the transistor operates as a switch. Amplifiers in this switched-mode category are “Class-D”, “Class-E” and “Class-F”, although Class-C and hard-driven or saturated Class-B amplifiers are also often placed in this group.
Non-linear, or switched-mode power amplifiers are unable to pass any signal containing amplitude modulation (AM) without massive distortion and spectral regrowth. However, if a constant-envelope RF signal without AM is used as an input, no distortion occurs. The output amplitude of these amplifiers is also, in the ideal case, directly proportional to the power supply. Thus, AM can be imposed onto the power amplifier supply in order to obtain complex modulation containing AM and phase modulation (PM) at the output of the power amplifier. Non-linear amplifiers are also very efficient, with theoretical efficiencies approaching 100%.
One form of transmitter using the switched-mode power amplifier of
In recent years, especially since the advent of fast, delta-sigma fractional-N phase-locked loops (PLLs), the EER concept has been developed and refined further. Envelope elimination and restoration is no longer necessary, but rather the amplitude and phase signals can be created in the digital baseband. The amplitude signal is then fed to a digital-to-analog converter (DAC) and on to the non-linear power amplifier power supply. The phase signal is differentiated to obtain a signal describing frequency and then this is used to modulate a PLL synthesizer. This is often a fractional-N PLL with the frequency data put into a sigma-delta modulator to obtain FM modulation.
The most efficient way to implement the fast power supply in the AM path is with a switched-mode power supply (SMPS). The bandwidth of the SMPS is however limited by the achievable switching speed.
In a polar transmitter architecture, I and Q signals are transformed from Cartesian coordinates (sine and cosine) into polar coordinates (amplitude and phase). The amplitude and phase information are separated and sent down separate paths until they are recombined in the switched-mode power amplifier. As already mentioned above, the phase information extracted from the original signal (either constant envelope or non-constant envelope) is transformed into a constant envelope signal. This is achieved by phase modulating a phase-locked loop designed to output the desired transmit frequencies. The resulting signal may now be amplified by compressed amplifiers without concern of distorting amplitude information. The extracted amplitude information is used to modulate the power supply of the power amplifier.
However, switched-mode transmitters are also limited in terms of their dynamic range. This is a function not just of the switched mode power amplifier, which exhibits extreme amplitude and phase non-idealities at low voltage, but also of the switched-mode power supply—the lowest available output voltage is limited both by the available switching duty cycle within the SMPS and the ripple present from the switching action.
This dynamic range issue may be the most difficult problem to address in switched-mode transmitters, such as polar transmitters. Systems built around various versions of Code Division Multiple Access (CDMA) schemes (e.g. 3GPP WCDMA (3rd Generation Partnership Project Wideband CDMA) or CDMA2000) have very large power control ranges, in excess of 70 dB. However, the power control range that is available from a polar transmitter might only be around 30 dB. This may be enough for GSM (Global System for Mobile communication) or GSM-EDGE (Enhanced Date for GSM Evolution) type systems, but not for CDMA type systems where large power-control ranges are required.
It is an object of the invention to provide a highly efficient transmission device and method, by means of which flexible use in all type of transmission systems can be ensured.
This object is achieved by a transmission device comprising:
Furthermore, the above object is achieved by a transmission method comprising the step of controlling an amplification of a transmission signal so as to selectively amplify said transmission signal either in a switched operation mode or in a linear operation mode based on a transmission system through which said transmission signal is transmitted.
Accordingly, power efficiency of the transmission can be increased through selective use of the switched-mode approach whenever possible, e.g., if the power control range is sufficient. Moreover, the ability to switch to linear mode for wide dynamic range systems opens the possibility of using the same hardware for different systems and thus leads to an increased flexibility.
Power supply means may be provided for supplying power to the amplifier means, wherein the power supply means are controlled in response to the switching means so as to generate a power supply with an amplitude modulation if the switched operation mode is selected, and to generate a constant power supply if the linear operation mode is selected. Hence, in the switched operation mode, the amplitude modulation can be selectively reintroduced through the power supply signal.
Additionally, at least one of predistortion, adjustment of supply voltage with output power and envelope tracking may be applied in the linear operation mode, so that a limited amplitude modulation of the supply power is obtained in the linear operation mode. Thereby, efficiency can be improved.
Furthermore, signal processing means may be provided for generating an amplifier input signal supplied to the amplifier means, wherein the signal processing means may be controlled in response to the switching means so as to generate said amplifier input signal with a constant envelope if said switched operation mode is selected and to generate said amplifier input signal with an amplitude modulation if said linear operation mode is selected. As an example, the switching means may comprise first switching means for selectively connecting either an envelope signal corresponding to the amplitude modulation or a constant power control signal to the power supply means.
Additionally, extraction means may be provided for extracting the amplitude modulation from an input signal of the transmitter device. In particular, the extraction means may comprise conversion means for converting an in-phase component and a quadrature component of the input signal into an amplitude signal and a phase signal, and wherein the amplitude modulation is derived from the amplitude signal. Thereby, a reconfigurable polar transmitter is provided which can be driven by a Cartesian I/Q signal. Variable delay means may be configured to selectively adjust a relative delay between the extracted amplitude modulation and the phase modulation of the input signal in response to the switching means.
The signal processing means may comprise amplitude modulation means controlled in response to the switching means. The amplitude modulation means can be set to a constant output state if the switched operation mode is selected. As an example, the switching means may comprise second switching means for selectively connecting either an envelope signal corresponding to the amplitude modulation or a constant power control signal to a modulation input of the amplitude modulation means.
As an additional measure, predistortion means may be provided for applying selective predistortion to a carrier input signal of the amplitude modulation means in order to selectively compensate for characteristics of the amplitude modulation means if the linear operation mode is selected.
The operation mode may be selected or set by using biasing means for changing a bias signal of the amplifier means in response to the switching means. The biasing means may comprise at least one of a programmable current source for generating a variable bias current and a programmable voltage source for generating a variable bias voltage.
Further advantageous developments are described below.
In the following, the present invention will be described based on an embodiment with reference to the accompanying drawings in which:
The embodiment of the present invention will now be described in connection with a reconfigurable polar transmitter as shown in
According to the embodiment, the polar transmitter can be changed between switched-mode operation (switched operation mode) and a linear-mode operation (linear operation mode) as desired, depending on which mode of operation best meets the needs of the radio system in use.
When operating in switched-mode as shown in
When operating in linear mode as shown in
Efficiency-improving techniques associated with linear transmitters can be used when the power amplifier 4 is in linear mode, e.g. predistortion, adjustment of supply voltage with output power and envelope tracking. A power supply unit 30 for supplying power to the power amplifier 4 can be variable in bandwidth, switching between static power control mode, envelope tracking and full amplitude modulation depending on the circumstances.
By controlling switching states of a first switching unit 40 and a second switching unit 42, the transmitter can be selectively set to linear (
For example, if a transmitting systems with low dynamic range requirements (e.g. GSM) is detected or determined to be used by the transmitter, the power amplifier 4 can be operated in the switched operation mode as shown in
The output signal of the power control circuit 26 is also used for controlling a variable gain amplifier (VGA) 2 to set the required dynamic range and maximum gain for driving the power amplifier 4.
When the transmitter is being used in a system requiring high dynamic range (e.g. WCDMA), the power amplifier 4 can be operated in the linear operation mode as shown in
The power amplifier 4 must be designed so that it can operate in both switched operation mode and linear operation mode with acceptable performance. Specifically, a bias signal supplied to the power amplifier 4 by a biasing circuit 34 can be set e.g. by programmable current and/or voltage sources. These bias voltages and/or bias currents are set to bias the power amplifier 4 to bias values suitable for either the linear operation mode or the switched operation mode depending on the transmission system, i.e. the switching state of the first and second switching units 40, 42. Thus, the biasing circuit 34 may have a control input (not shown) which is controlled by the same control signal or information supplied to the first and second switching unit 40, 42.
As an additional measure, it may be necessary to apply a predistortion to the transmission chain (lower branch in
The power supply unit 30 supplies the power signal via a first low pass filter 32 for removing unwanted high frequency components or spurious signals and may typically be a switched mode power supply, although it could also be implemented as a linear regulator, a combination of a switched mode power supply and a linear regulator, a linear amplifier, a switched-capacitor supply or the like.
In the upper branch or amplitude path used for supplying the amplitude information or envelope signal, a digital-to-analog converter (DAC) 24 is provided if the transmitter receives digital I and Q data streams at its input. In some implementations the DAC 24 in the amplitude path could be eliminated and a digital or PWM (pulse width modulation) signal passed to the switched-mode power supply unit 30. The DAC 24 is followed by a second low pass filter 28 for removing unwanted high frequency components or spurious signals.
A back end RF-IC 20 will take the digital I and Q data streams after pulse shaping and convert them to amplitude and phase signals. One way to do this is with some kind of Cordic algorithm applied by a Cordic processor. The Cordic processor transforms the Cartesian coordinates (sine and cosine) of the I and Q data streams into polar coordinates (amplitude and phase). The amplitude and phase information are separated and supplied to separate paths, i.e., the upper amplitude branch and the lower transmission chain, respectively.
The amplitude information is fed to the DAC 24. In the switched operation mode of
A variable delay unit 22 is provided to have the capability to adjust the relative delays between the upper amplitude path and the lower phase path in the transmission chain so that these modulation signals arrive at either the power amplifier 4 (when the transmitter is running in the switched operation mode) or the amplitude modulator 36 (when the transmitter is running in the linear operation mode) at the same time.
The required power control dynamic range can be provided by the VGA 2 or a VGA line-up after the amplitude modulator 36. Furthermore, it may be necessary, for some systems, to add a bandpass filter (not shown) before the power amplifier 4 in order to filter noise and/or spurious signals.
The phase information or phase modulation is differentiated and then fed to a PLL synthesizer modulator 38, which can be implemented either as a single-point FM modulator (e.g. fractional N synthesizer) or with a two-point modulation, as desired. A voltage-controlled oscillator (VCO, not shown) provided in the PLL synthesizer modulator 38 may be running at the channel frequency or multiples of the channel frequency (e.g. 2× or 4×). When the VCO runs on a multiple of the channel frequency, the PLL synthesizer modulator 38 has a frequency divider to convert the VCO frequency (e.g. divided by 2 or 4) to the actual channel frequency. This will depend on the number of bands to be supported and their frequency allocations as supplied from a channel information provided as a control information or stored in a (programmable) channel unit or memory 44. Additionally, the characteristics of the PLL synthesizer modulator 38 may be measured and characterized, i.e. for single-point PLL modulation with pre-emphasis or two-point PLL modulation.
The proposed transmitter according to the above-described embodiment provides advantages relative to a traditional IQ modulator approach in that power amplifier efficiency is improved by using the switched-mode approach whenever possible. Moreover, advantages relative to “pure” switched-mode transmitters are achieved by the ability to switch to the linear operation mode for wide dynamic range systems, which opens the possibility of using the same hardware for different systems.
In summary, a transmission device and method have been described based on the embodiment of
It is to be noted that the present invention is not restricted to the above embodiment and can be implemented in any transmitter architecture having an amplifier circuit or device which can be configured to selectively operate either in a linear operation mode or in a switched operation mode. The first and second switching units 40, 42 may be implemented by using any kind of switching element, e.g. active or passive semiconductor elements or switching circuits. Furthermore, the only one or more than two switching units may be provided to achieved the selective supply of the amplitude information to either the amplitude modulator 36 or the power supply input of the power amplifier 4. In general, the present invention is intended to cover any embodiment or modification where an amplifier can be selectively switched between a linear mode of operation and a switched mode of operation. The preferred embodiments may thus vary within the scope of the attached claims.
