[Devices & ICs] Single Chip Holds WiMAX Transceiver The industry’s first single-chip WiMAX transceiver supports wireless broadband applications from 2.3 to 2.7 GHz with many cost-saving features and functions. Walter Lau | ED Online ID #15789 | June 2007 Broadband services, such as Internet access and streaming video, are coming to a wireless transceiver near you. Many wireless service providers are already offering forms of wireless broadband service through their existing cellular networks. Many are also considering WiMAX equipment as an effective way to augment their networks with broadband-wireless-access (BWA) capabilities. Of course, affordable WiMAX service requires low-cost WiMAX equipment, and the new model MAX2837 wireless broadband RF transceiver integrated circuit (IC) from Maxim Integrated Products is a single-chip solution in that direction. The transceiver operates from 2.3 to 2.7 GHz and supports fixed, portable, and mobile WiMAX operation. WiMAX, which stands for Worldwide Interoperability for Microwave Access, has been called a wireless replacement for wired digital-subscriber-line (DSL) technology. The wireless-metropolitan-area-network (WirelessMAN) technology provides high-data-rate communications for fixed-station subscribers at distances as great as 30 miles from a base station and for mobile station subscribers at distances as great as 10 miles from a base station. The operating “rules” for WiMAX equipment are defined by several IEEE standards, with IEEE 802.16-2004 detailing fixed WiMAX system requirements and IEEE 802.16e-2005 covering WiMAX requirements for subscribers on the move. For the IEEE 802.16-2004 fixed standard, licensed frequencies fall around 2.5 and 3.5 GHz while unlicensed frequencies are around 5.2 and 5.8 GHz. Mobile WiMAX, which is formulated to allow roaming of mobile units with handoffs from one base station to another, as defined in IEEE 802.16e, is designed to operate in the 2.3-, 2.5-, 3.3-, and 3.4-to-3.8-GHz frequency bands. Fixed WiMAX is based on orthogonal frequency-division-multiplex (OFDM) technology while mobile WiMAX uses both OFDM and orthogonal frequency-division-multiple-access (OFDMA) technology. WiMAX employs a number of different modulation formats, including BPSK, QPSK, 16QAM, and 64QAM, switching the modulation format according to the strength of the received signal. By this adaptive modulation approach, transmission rates as fast as 75 Mb/s are possible. WiMAX signals, which have a frame structure with preamble and data sections, are sent in bursts. The noise-like signals are converted from the frequency domain to the time domain by means of Fast Fourier Transforms (FFTs), with as large as 2048-point FFTs used for WiMAX signals. The MAX2837 single-chip transceiver (Fig. 1) incorporates both the RF-to-baseband receive signal path and the baseband-to-RF transmit signal path. It is based on an advanced SiGe BiCMOS process that features low power consumption but also the low noise figures needed for good receiver sensitivity. The transceiver features a direct-conversion, zero-intermediate-frequency (zero-IF) architecture that includes a fast-switching sigma-delta fractional-N phase-locked-loop (PLL) frequency synthesizer with 20-Hz step size and low phase noise. The zero-IF architecture exhibits extremely low carrier leakage into adjacent channels without the need for frequency correction. In contrast to traditional superheterodyne transceiver architectures that employ an external surface-acoustic-wave (SAW) filter for channel filter, the MAX2837 integrates monolithic baseband filters for its receiver and transmitter sections. The wireless transceiver IC offers programmable channel bandwidth filters that match the channel-bandwidth requirements of the WiMAX standards, at bandwidths from 1.75 to 28.0 MHz. The on-chip channel filter can be programmed for all of the WiMAX bandwidths: 1.75, 2.5, 3.5, 5.5, 6, 7, 7. 8.75, 10, 14, 15, 20, and 28 MHz. An SPI™-compatible interface enables electronic programming of channel filtering which allows one basic radio design to support virtually all geographic markets where channel bandwidths differ according to each country’s spectrum regulations. In support of the synthesizer, the well-equipped IC also includes a voltage-controlled oscillator (VCO) and a digitally controlled crystal oscillator (DCXO). The DCXO is designed for use with a low-cost crystal resonator, to save the cost of a temperature-compensated crystal oscillator (TCXO). With the synthesizer, inclusion of the DCXO allows the implementation of digital automatic frequency control (AFC) with the MAX2837 for improved system range while maintaining the minimum bit-error-rate (BER) performance required by the WiMAX standards (10–6 BER). The on-chip DCXO enables the use of low-cost crystals with ±25 PPM or higher frequency tolerance over temperature. The DCXO’s tuning step is less than 1 PPM and is monotonic in nature, allowing the development of an AFC algorithm that maintains total frequency error within 1 PPM over the full operating temperature range. The fractional-N PLL synthesizer even enables the use of an inexpensive non-tuned crystal. Instead of digitally tuning the load capacitance of the crystal oscillator, the RF frequency can be adjusted in 25-Hz steps to remove the frequency error. The sampling clock error (and symbol timing error) that is related to the reference clock (the crystal oscillator) can be corrected digitally in the WiMAX modem’s interpolation of the I/Q samples to account for drift in the case of the sampling clock. The MAX2837, which includes a baseband/control interface, features circuitry for on-chip DC-offset cancellation, in-phase/quadrature (I/Q) error detection, and carrier-leakage detection. Very little additional circuitry is needed for form a complete WiMAX RF transceiver front end, except for an RF bandpass filter, the crystal resonator for the reference oscillator, an RF switch, a power amplifier for the transmitter, and a small number of passive components. The receiver portion of the MAX2837 features a an integrated low-noise amplifier (LNA) with noise figure of 2.3 dB with input third-order intercept point (IIP3) that is adjustable from –8 to –11 dBm at the maximum gain setting. The peak-to-peak gain variation over the full RF input range is typically 0.8 dB (Fig. 2) while the RF input return loss is typically 13 dB at all setting of the integrated LNA. The total voltage gain is typically 99 dB, which is comprised of adjustable RF and baseband gain. The RF gain steps are 8, 16, and 32 dB while the baseband gain range is typically 62 dB (in 2-dB minimum steps).
| |||||||||||||||||||||||||||||||||||||||||||||||||
Resources

RSS














