The 5G wireless revolution is bringing dramatic changes in the RF design realm, and power amplifiers for handsets and radio base stations are no exception. For a start, power amplifier chips in 5G wireless applications are going to be much different from those employed in 4G networks.
That’s primarily because the broadband modulation for 5G transmissions demands high-power efficiency and stringent linearity from power amplifiers. Moreover, when the 5G networks are going to employ the phased array antennas to focus and steer multiple beams, what will really matter is the ability to divide the transmission task among multiple beams.
Here, with a phased array antenna, for instance, comprising a 4×4 array, it’s imperative that the power amplifier operates at much lower power than those needed to amplify the single-beam, omnidirectional signals currently used in the cellular systems.
It’s worth mentioning here that 5G networks are initially being implemented in the sub-6 GHz frequency range. However, 5G’s real promise comes with the commercial realization of millimeter wave (mmWave) communications in the 24 GHz, 28 GHz, and 39 GHz frequency bands. Therefore, while centimeter-wave (cm-Wave) 5G systems are likely to arrive in the market first on sug-6 GHz systems, it’s their mmWave counterpart that is going to challenge the RF design norms.
As a result, multiple-input multiple-output (MIMO) antennas serving a variety of devices in the densely deployed environments will call for power amplifier chips with high-power efficiency and stringent linearity. The phased-array MIMO antennas with numerous RF front-ends will also demand power amplifiers that offer greater integration at lower costs.
This predicament can be seen at play in the new PA devices that include PA modules, PA-duplexers, switched power amplifier plus duplexer (S-PADs), PA module integrated duplexers (PAMiDs), and total radio modules (TRMs).
New integration milestones
The PA modules, already a cornerstone of integration, are further reducing the parts count in the 5G RF front-end. The 5G networks feature more bands, and that mandates more RF switching, filtering, and power amplification elements in PA modules. Therefore, as the 5G networks evolve, PA modules will continue to increase in complexity.
Already, in the 4G wireless arena, the pressure to put almost the entire RF front-end covering multiple bands and technologies into a few PA modules has forced many smaller suppliers out of business. Inevitably, in the 5G realm, the pressure to pack more circuitry into PA modules is likely to increase.
So, NXP Semiconductors is trying to simplify radio power solutions for MIMO and massive MIMO (mMIMO) systems by combining smaller and lighter active antenna systems with multi-chip modules (MCM) serving the RF power. These RF power amplification solutions, while significantly boosting the integration, span from sub-6 GHz to 40 GHz bands while facilitating power supply from milliwatts to kilowatts.
Qorvo, a supplier of PA modules for 5G designs, is also warming up to the challenges confronting 5G power amplifiers. In 2016, the RF chipmaker joined hands with NanoSemi, a developer of linearization software. The collaboration aims to enhance Qorvo’s PA modules with NanoSemi’s machine learning-based digital pre-distortion (DPD) algorithms and thus ensure ultra-wideband linearization in power amplifiers.
The multi-carrier configurations pose serious challenges for power amplifiers serving the multi-band 5G designs, and NanoSemi’s digital compensation technology helps power amplifiers in tuning to power and capacity requirements according to available resources.
NanoSemi has also entered into a collaboration with National Instruments (NI), a supplier of automated test and measurement solutions, to allow designers to validate and optimize the performance of 5G power amplifiers coping with increasing bandwidth and power efficiency demands. The test solution provides 5G designers with a deep insight into the performance parameters of power amplifiers under extreme linearization conditions.
PA’s underlying technology
Another worthwhile comparison with 4G relates to the underlying technology for power amplifiers. In the 4G domain, gallium arsenide (GaAs) has been a leading technology in the power amplifier chips. That’s because GaAs can easily support the high voltages required for power amplifiers. However, once the wireless industry moves beyond the sub-6 GHz communications, where GaAs devices are likely to dominate, a new breed of semiconductor solutions are vying for a place in the mmWave regime.
For example, a new RF silicon-on-insulator (SOI) technology developed at the University of California at San Diego (UCSD) is making waves for putting silicon-based transistors in a series to achieve higher voltages in power amplifiers. The stacked transistors — four transistors in a serial arrangement — provide the necessary output power for 5G power amplifiers. The stacking of transistors not only boosts the overall voltage handling, it also removes the parasitic problems associated with body effects and substrate capacitance.
Other candidates for 5G power amplifiers include gallium nitride (GaN) and silicon germanium (SiGe). The GaN technology bolsters the PA performance, efficiency, and power by facilitating the transmission of multiple data streams with greater capacity and thermal efficiency. According to Yole Développement, the RF market for GaN devices is expected to grow from $380 million in 2017 to $1.3 billion by 2023.
The 5G design world is in a state of flux, and as this article shows, power amplifier chips are fully part of this transformation. It’s also apparent that the journey toward 5G’s capacity revolution is going to impact all the major facets of power amplifier design: physical size, efficiency, linearity and reliability.