Coding efficiency for different switched-mode RF transmitter architectures

Abstract

Conventional RF transmitters with power amplifiers in e.g. Class-AB operation provide only moderate electrical efficiencies for complex modulated signals. To increase the overall efficiency considerably, we can employ switched-mode power amplifiers (SMPA) with an appropriate baseband signal processing. Such efficient SMPAs generate a lot of spectral out-of-band components, which has to be attenuated by a bandpass filter before the signal is fed to the transmit antenna. Unfortunately, this impacts the electrical efficiency because these unwanted signal components are usually dissipated within the transmitter. A measure for the electrical efficiency degradation for complex modulated signals in a switched-mode operation is the coding efficiency. In this paper we develop different switched-mode transmitter architectures based on baseband PWM coding and analyze the coding efficiencies for different signal statistics.

Concurrent Multiband Digital Outphasing Transmitter Architecture Using Multidimensional Power Coding

“…For a Class-S transmitter (Fig. 4b), which consists of a Class-D PA followed by an LC band-pass filter (BPF), following the definition on coding efficiency given in [11][12][13], we determine the coding efficiency as the in-band channel power divided by the total signal power both at the transmitter output. What makes it different from Class-D operation is the way the input signal is driving the PA; in order to modulate the output envelope amplitude, the input signal duty cycle is modulated.…”

“…The power delivered to the load is given by (8) where is the peak output voltage applied to the load resistance . The average current from the supply is obtained as (9) The power delivered from the supply is determined as (10) The peak PA efficiency is obtained by dividing the power delivered to the load by the power delivered from the supply (11) The average PA efficiency is 100 %, which is obtained from (12) where is the peak PA efficiency and is the coding efficiency. If the PA input is modulated by oversampling such as delta-sigma modulation, although there are no high-order harmonics at the load, the BPF allows in-band quantization noise to pass; consequently coding efficiency drops, which decreases the PA average efficiency.…”

“…The average efficiency of a Class-S outphasing transmitter with Class-D PAs for a modulated signal is obtained by considering the output power and efficiency for each power code that it transmits, (25) where is an outphasing angle associated per power code, is power combiner insertion loss, and is the average coding efficiency, which is defined in [11]- [13], [37] as a ratio of the amount of the transmit power in the band of interests to the total transmit power. As discussed in Section II-E, in order to include the efficiency degradation due to the high-order harmonics, the power efficiency for each power code should include the efficiency degradation factor in Eq.…”

“…The average power efficiency of an ideal Class-S transmitter, as reviewed in Section II-A, can be determined as a product of the PA peak power efficiency and power coding efficiency, which is defined as a ratio of the in-band power to the total output power both at the transmitter output [11]- [13], [37]. Compared to the linear PAs [14]- [16] whose average power efficiency depends on the instantaneous envelope amplitude of the RF signals in transmit, the power efficiency of an ideal Class-D PA in Class-S operation remains constant.…”

Analysis of Non-Uniform Polar Quantisers in a Sigma Delta Transmitter Architecture

“…The latter can be put down to the increased signal level into the quantiser caused by the enhanced noise amplification of the MOD-2 NTF [14]. The reduced edge count is also a factor in the improvement in coding efficiency (defined as the useful signal power to the total signal power, [21]). The results indicate coding efficiency improves with quantisation resolution (N q ), but reduces with filter order.…”