Abstract
The recent allocation of the 6 GHz band for wireless communication systems has opened up new possibilities for high speed and low latency applications. Analog Devices’ 16 nm transceiver family offers a highly integrated solution for this new frequency band, featuring low power consumption and high performance. This article introduces the 6 GHz band and discusses the benefits of the zero-IF architecture employed in ADI’s transceiver family. This article also highlights the key features of this transceiver family and explores their application in various scenarios.
Introduction
With the forward momentum of wireless communication systems, new technologies and new spectrum are continuously being pursued and adopted. For those involved in the wireless industry, it’s no surprise that the 3GPP’s (the 3rd Generation Partnership Project) inclusion of the 6 GHz frequency band in frequency range 1 (FR1) was welcome news. By expanding the original FR1, both lower and higher, from [450 MHz to 6000 MHz] to [410 MHz to 7125 MHz], the industry gains access to a significant amount of new spectrum, opening up fresh opportunities for growth and innovation.
Compared to the legacy FR1 frequency band, the new 6 GHz band supports much wider bandwidths: n96 is 1200 MHz (5925 MHz to 7125 MHz), n102 is 500 MHz (5925 MHz to 6425 MHz), and n104 is 700 MHz (6425 MHz to 7125 MHz). Offering high capacity spectrum with good network energy performance and better propagation than frequency range 2 (FR2) bands ensures that the new 6 GHz band will become an important resource for wireless connectivity. In the U.S., the Federal Communication Commission (FCC) earmarked the band for unlicensed Wi-Fi in 2020, making the 6 GHz band highly competitive in this market.
This article reviews the features and benefits of ADI’s 16 nm transceiver family for these applications. Available in both 8T8R (eight transmitters and eight receivers) and 4T4R (four transmitters and four receivers), the ADI 16 nm transceiver is a highly integrated device with extensive digital front-end functionality, including digital predistortion (DPD), crest factor reduction (CFR), and digital channel upconverters and downconverters (CDDC and CDUC). There are energy saving features as well.
Architecture
As shown in Figure 1, ADI’s 16 nm transceiver series integrates eight differential transmitters (Tx0-7), eight differential receivers (Rx0-7), and two differential observation receivers (ORx0-1). The tunable frequency ranges from 400 MHz to 7125 MHz with two radio frequency (RF) synthesizers as the local oscillator (LO). The tunable bandwidth is as wide as 600 MHz. A high speed JESD204B/JESD204C interface is designed to connect to the baseband processor.
Transmitter
The transmitter applies zero-IF architecture as shown in Figure 1. The in-phase and quadrature (I/Q) baseband signals from the digital-to-analog converters (DACs) are reconstructed and filtered by the baseband low-pass filters (LPFs) and then upconverted by an analog modulator and an LO to RF output. The zero-IF transmitter provides higher linearity and noise performance with relatively lower power consumption than the RF sampling converter.
The transfer function of the DAC takes a general form of sin(x)/x, and its frequency response is not flat as shown in Figure 2. The analog output is attenuated at the higher frequency. Images of the desired signal are generated during the sampling process and need to be filtered out. Otherwise, they pollute the radio spectrum and violate the emission requirements from 3GPP and the FCC.
As a result, the maximum usable DAC output frequency is typically 40% of the sampling clock rate. For RF sampling to function effectively with 6 GHz band (up to 7.125 GHz), the DAC sample clock must operate at above 18 GHz, which costs significant power. Here the advantage of the zero-IF transmitter is readily apparent. It only needs to digitize the baseband I/Q signal, and the DAC sample clock could be as low as 3 GHz to support the 6 GHz band. This provides flatter output power across the whole 6 GHz band (Figure 3) and lower noise spectral density (NSD) performance with relatively low energy consumption. Typically, even with the same process, RF sampling converters consume about 125% more energy than baseband I/Q converters for equivalent noise performance for a typical single band application.
Receiver
On the receiver’s path, the RF input is downconverted to a baseband I/Q signal using an analog demodulator and an LO. The continuous-time delta-sigma ADC is specifically designed to digitize the baseband I/Q signal. This ADC incorporates inherent antialiasing filtering, which significantly relaxes the filtering requirements compared to traditional sampling techniques. At the RF input port, the wideband matching provides a flat frequency response across the 6 GHz band as shown in Figure 4.
The baseband amplifier can be a classic topology using feedback circuitry for good linearity and noise performance. However, the RF sampling receiver needs expensive extra filtering at the RF frequency band. To sample the 6 GHz band, the RF sampling ADC requires an 8 GSPS sampling clock to convert the desired signal from the second Nyquist zone, making it impossible to avoid aliased products without aggressive filtering to mitigate its impact. Alternatively, a higher than 15 GSPS sampling clock can be used to relax the antialiasing requirement, but this approach consumes significantly more energy compared to the zero-IF’s baseband I/Q sampling. In contrast, the zero-IF’s baseband I/Q sampling only requires a low I/Q sampling clock of around 3 GSPS to achieve sufficient performance.
Additionally, the NSD for the zero-IF receiver is generally independent of the frequency band. As shown in Figure 5, the NSD at 6300 MHz and 7100 MHz are almost the same.
Observation Receiver
In this highly integrated transceiver, two observation receivers are designed as the RF sampling architecture, which provides performance for the loopback receiver of the DPD for the power amplifier (PA), the monitoring path for the transmitter output power, or the sniffing receiver for the RF spectrum, etc. with appropriate front-end design.
To support the various applications, the observation receiver can be configured to work at four sampling clock rates, which provides the flexibility to choose between bandwidth, NSD performance, and power. See Table 1 for the NSD performance and power at the different sampling clock rates.
Table 1. NSD Performance and Power at the Different Sampling Clock Rates
| Sampling Clock | Usable Nyquist Bandwidth | NSD | Relative Increased Power to 2949.12 MSPS |
| 2949.12 MSPS | 1274.56 MHz | –144 dBFS/Hz | 0 mW |
| 3932.16 MSPS | 1766.08 MHz | –145 dBFS/Hz | 235 mW |
| 5898.24 MSPS | 2749.12 MHz | –147 dBFS/Hz | 365 mW |
| 7864.32 MSPS | 3732.16 MHz | –148 dBFS/Hz | 780 mW |
Applications
Wireless Massive Multiple-Input Multiple-Output (MIMO) System
ADI’s 16 nm transceiver has achieved extensive deployment in sub-6G massive MIMO systems and millions of base transceiver station (BTS) equipment incorporating this technology successfully operating in real-world applications. It has demonstrated its proficiency as a dependable radio solution within the sub-6G frequency spectrum. As of 2025, with the extended 3GPP FR1, the same transceiver can provide the same performance at the 6 GHz band as well as the following advantages.
Wide Bandwidth Supported
- Support 600 MHz instantaneous bandwidth (IBW) on both transmitter and receiver and 800 MHz synthesis bandwidth for the digital predistortion (DPD) of the PA.
- Two observation receivers can be used as the feedback channel for the PA’s digital predistortion.
- JESD204B/JESD204C digital interface up to 19.66 Gbps/32.44 Gbps to support the wide bandwidth.
Technology to Reduce the Phase Variation over Channels
- Multichip synchronization (MCS): As part of device initialization, the MCS state machine takes the global system reference signal (SYSREF) to reset the data converter clocks and all other clocks at the digital data path to synchronize the clock phase to the device clock (DEVCLK), which aligns the phase from the JESD interface to data converters. Additionally, the MCS state machine resets the RF PLL phase to align with DEVCLK and the dividers on the LO distribution path, which provides the overall phase alignment at the RF input and output ports.
- Phase compensation over transmitter attenuation: The gain or attenuation change on the signal path is another cause for phase variation. To mitigate the phase variation, a pre-characterized phase compensation has been added to each transmitter attenuation index. This ensures that a phase correction is applied whenever the system tunes the attenuation.
Designing these technologies in the transceiver helps reduce the complexity of system antenna calibration by initializing the channels to a more consistent startup condition. This allows the antenna calibration to run less frequently during operation by reducing the temperature dependence of the RF PLL and mitigating the phase variation over gain change.
Power Saving
Discontinuous transmission (DTX) mode: Traditional radio units consume a considerable amount of energy even when there is no user in the cell. This transceiver includes the DTX functionality that can deactivate components in the transmitter data path during the empty transmission time intervals (TTIs). With DTX configured, the transceiver deactivates the power amplifier and other transmitter components when it identifies zero data conditions. The devices are quickly activated when nonzero data is detected. In scenarios using actual mobile network operator data, the technology has reduced RU energy consumption by more than 30%, without impacting quality of service (QOS).
16 nm Transceiver for Unlicensed 6 GHz Band for Wi-Fi System
After the U.S. FCC voted to allow unlicensed wireless LANs to operate in the 6 GHz band in 2020, the Wi-Fi Alliance allocated the spectrum from 5925 MHz to 7125 MHz into Wi-Fi 6E channels.3 This adds 14 additional 80 MHz channels or seven additional 160 MHz channels on top of the traditional 2.4 GHz band and 5 GHz band. See Table 2 for the unlicensed bands at the 6 GHz band.
Table 2. Unlicensed NII Bands at 6 GHz Band
| U-NII Bands | Frequency Range (GHz) | Bandwidth (MHz) |
| U-NII-5 | 5.925 to 6.425 | 500 |
| U-NII-6 | 6.425 to 6.525 | 100 |
| U-NII-7 | 6.525 to 6.875 | 350 |
| U-NII-8 | 6.875 to 7.125 | 250 |
The 6 GHz band is covered by ADI’s 16 nm transceiver series, providing good performance and the flexibility to trade-off between energy consumption and bandwidth, which also benefit from the zero-IF architecture as discussed in the previous sections.
Single Radio Chip to Support 1200 MHz with Spatial Diversity (4× or 2× Antenna Diversity)
As stated previously, the transceiver supports 600 MHz IBW, and working with two internal LOs, the whole 1200 MHz band can be covered by a single chip. As shown in Figure 6, the transceiver is configured to support four antennas (four channels) for the entire 1200 MHz band. LO0 is used for channels 0 to 3 to cover both U-NII-5 and U-NII-6 on all four channels. Likewise, LO1 will be configured at 6825 MHz for channels 4 to 7 for U-NII-7 and U-NII-8. Both 600 MHz bands can be sent to the baseband through the high speed JESD204C interface simultaneously. See Table 3 for the configuration details.
Low Power Solution to Support 1200 MHz with LO Frequency Sweeping
The 6 GHz band Wi-Fi spectrum can be split into 59 channels with 20 MHz bandwidth on each channel, or support seven channels with 160 MHz bandwidth on each. Instead of the above configurations of wide bandwidth, the transceiver can be configured as narrow bandwidth for low energy consumption. For example, with a data rate of 245.76 MSPS, the signal bandwidth can be 160 MHz, and the JESD lane rate can operate as low as 9.8 Gbps. The RF LO frequency can be configured flexibly within the 1200 MHz band to cover the whole 6 GHz band. The power consumption of the transceiver for this low power configuration can save 20% energy vs. the wide bandwidth configuration. The configuration example is shown in Figure 7 and Table 4.
Table 3. Wide Bandwidth Configuration of the Transceiver
| No. of Antenna | LO | IBW | Data Rate | JESD | Lane Rate | No. of Lane | |
| U-NII-5/6 | 4 | LO0 = 6225 MHz | 600 MHz | 983.04 MSPS | JESD204C | 32.44 Gbps | 4 |
| U-NII-7/8 | 4 | LO1 = 6825 MHz | 600 MHz | 983.04 MSPS | JESD204C | 32.44 Gbps | 4 |
Table 4. Narrow Bandwidth Configuration of the Transceiver
| No. of Antenna | LO | IBW | Data Rate | JESD | Lane Rate | No. of Lane | |
| U-NII-5/6 | 4 | LO0 | 160 MHz | 245.76 MSPS | JESD204B | 9.8 Gbps | 4 |
| U-NII-7/8 | 4 | LO1 | 160 MHz | 245.76 MSPS | JESD204B | 9.8 Gbps | 4 |
Wideband Observation Receiver for Spectrum Sweeping
For this application, the observation receiver can be configured to be 7.8 GHz, which covers the Wi-Fi 6 GHz band seamlessly. Figure 8 shows how the 6 GHz band is located at the upper band of the second Nyquist zone, and accordingly, at the first Nyquist zone, the reversed 6 GHz spectrum can be converted to the baseband by utilizing the NCO on the receiver data path.
In the wireless communication market, with the new technologies and the new spectrum continuously being adopted, a cost-effective solution is required as well by the operators. Therefore, a high integration and low power solution become more important. ADI’s 16 nm transceiver family integrates the eight channels with the high performance analog front end as well as the digital front-end functions (DPD, CFR, and CDDC/CDUC) on a single chip. Moreover, the zero-IF architecture provides a low power transceiver solution, and the power saving feature (DTX) is integrated into the device to further reduce the system power consumption by controlling the PA. Furthermore, its flexible configuration makes it an agile solution for various applications such as the wireless BTS and Wi-Fi systems, etc.
About the Author
Howie Jing is the product applications manager in the Wireless Platform Group (WPG) at Analog Devices and he works in the RTP office. He joined ADI China in 2011, starting as an applications engineer to support the transceiver products in China market. In 2019, he moved to the ADI U.S. office as an application manager and continued to support transceivers for worldwide customers. Prior to ADI, he worked in Maxim Integration (now ADI) as an applications engineer for the digital audio/video network and 3G cellular applications.
