This article presents the design and implementation of a novel time-division duplex (TDD) front-end architecture, namely, the Indirectly nonreciprocal load modulated balanced amplifier (INR-LMBA), targeted at emerging massive MIMO communication… Click to show full abstract
This article presents the design and implementation of a novel time-division duplex (TDD) front-end architecture, namely, the Indirectly nonreciprocal load modulated balanced amplifier (INR-LMBA), targeted at emerging massive MIMO communication systems. Unlike conventional solutions that place a circulator at the power amplifier (PA) output to mitigate load mismatches, the proposed architecture relocates the circulator to the low-power control amplifier (CA) path. This significantly reduces power stress on the circulator by nearly an order of magnitude, thus paving the way for compact, nonmagnetic implementations. Combined with a transmit/receive (T/R) switch, INR-LMBA enables standard TDD functionality while offering quasi-isolation during transmission and low loss during reception. Under matched-load conditions, the architecture operates similar to a conventional pseudo-Doherty load modulated balanced amplifier (PD-LMBA). Under mismatched conditions, however, the circulator redirects reflected power to its isolation port, minimizing mismatch effects on the PA. As a proof-of-concept demonstration for this architecture, a 2–2.5-GHz INR-LMBA prototype is implemented, which achieves an efficiency of 62%–73% at peak output power and 53%–61% at 10-dB output back-off (OBO) over the in-band operation at 50- $\Omega $ load. In modulated evaluation with a 20-MHz OFDM signal, the measured average PA efficiencies are around 52% for a matched load with ACPR greater than 34 dB across the band of operation for matched load. The PA also exhibits strong mismatch resilience with higher than 43% average efficiency and >35-dB ACPR across the band for a mismatched load at 2:1 voltage standing wave ratio (VSWR).
               
Click one of the above tabs to view related content.