Advantages of Second-Order Cartesian Feedback Linearizers for Radio Amplifiers

Cartesian feedback (CF) linearization for radio amplifiers is appealingly simple, but it imposes a trade-off between signal bandwidth and feedback delay. In almost all of the CF literature, it is a first-order circuit. In this paper, we examine two second-order CF linearizers and show that they can...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on circuits and systems. I, Regular papers Regular papers, 2019-11, Vol.66 (11), p.4134-4146
Main Authors: Cavers, James K., Mehrotra, Kishore, Woodward, Graeme K.
Format: Article
Language:English
Subjects:
amplifier linearization
Amplifiers
Analytical models
Bandwidth
Bandwidths
Cartesian coordinates
Cartesian feedback
Circuit stability
Circuits
Computer simulation
Damping
Delay
Delays
Feedback
Intermodulation distortion
intermodulation distortion (IMD)
Linear analysis
Linearization
Nonlinear distortion
power amplifier (PA)
Receivers & amplifiers
Stability analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Cartesian feedback (CF) linearization for radio amplifiers is appealingly simple, but it imposes a trade-off between signal bandwidth and feedback delay. In almost all of the CF literature, it is a first-order circuit. In this paper, we examine two second-order CF linearizers and show that they can provide deeper suppression of intermodulation distortion or greater signal bandwidth at a given loop delay. Twin-loop CF with damping factor 0.707, for example, can provide 20 dB lower intermodulation distortion and half the error vector magnitude (EVM), relative to first-order CF. For twin-loop CF and proportional-integrator CF, we provide the first analyses of their behavior, stability limits, and linearization performance. Comparison with simulations using nonlinear amplifier models shows both the value and the shortcomings of linear analysis. Our new spectral separation approach leads to insights about linearization and EVM.
ISSN:1549-8328
1558-0806
DOI:10.1109/TCSI.2019.2926387