Flexible upconversion architectures satisfy global wireless standards

Achieving the goal of seamless voice and data connectivity affects the design requirements for wireless infrastructure. System capacity must increase. This requires larger bandwidth signals and multi-carrier capabilities. Higher power transmitters and more sensitive receivers extend the coverage range. To ensure ubiquitous coverage, the network needs smaller pico base stations to provide service to every nook of the city. Original equipment manufactures (OEMs) must be able to supply equipment that delivers this improved performance while maintaining efficient design practices and managing design costs.

Additionally, OEMs must employ systems that work in a variety of existing wireless standards including CDMA2K, WCDMA, GSM and EDGE, as well as emerging standards such as TD-SCDMA employed in China and WiMAX for broadband data services worldwide. A flexible architecture that works in most or all of these types of modulation cases serves well for minimizing design resources and improving reliability. Complexity is increased because frequency bands of operation vary globally. Frequencies from 800MHz to 2.1GHz are used for most voice communications, with the 3.5-GHz and 5.6-GHz ranges addressing data services.

These variables require a flexible up-converter with the RF performance to meet stringent base station specifications, all while maintaining good integration for a cost-effective and compact design. It must be capable of serving the multitude of frequency bands and a variety of the wireless modulation standards.

Transmitter Architecture
The two major architecture options for the transmitter are direct upconversion and superheterodyne. The traditional superheterodyne architecture consists of two mixing stages where the signal is first up-converted to a fixed intermediate frequency (IF) and then passed through a narrow band surface acoustic wave (SAW) filter. The direct conversion approach bypasses the IF stage and converts the signal directly from base band to the RF channel of choice. The block diagram of both architectures is shown in Figure 1.

1. Superheterodyne and Direct Up-conversion Architectures.

The direct upconversion approach utilizes a quadrature modulator and eliminates an additional mixing stage, an additional synthesizer, and a SAW filter. This greatly simplifies the design and reduces the bill of material (BOM) costs. This approach also offers the greatest flexibility, as it can be used for a variety of modulation schemes including CDMA, GSM, and OFDM.

Since there is no narrow band filter, the architecture can support a variety of signal bandwidths commensurate with the modulation scheme of choice. For example, the various bandwidths associated with CDMA2K and WCDMA can be supported as well as the variety of WiMAX signal bandwidths that typically range from 3.5 to 10MHz. Multi-carrier applications are equally supported since there is no bandwidth restriction.

Additionally, the direct upconversion architecture sustains digital predistortion linearization (DPD) signals which must be up to five times the bandwidth of the desired signals to incorporate the third and fifth order products that are tailored to cancel out the nonlinear effects of the power amplifier.

Direct Up-Convert Modulator
The direct upconversion modulator consists of differential in-phase (I) and quadrature-phase (Q) signals that are summed together at the output. It is imperative to use the quadrature modulator for the direct upconversion approach because the local oscillator (LO) component and the unwanted image signal, or sideband, are naturally suppressed without the need for a filter.

The amount of sideband suppression is determined by the amplitude and phase balance of the input quadrature components. The LO feedthrough is determined by the DC offset balance between the two input paths. It is best to have a device that inherently provides better than 35dB of suppression on each of these parameters as there is likely degradation over temperature. If further suppression is required, additional fine-tuning can be accomplished in the digital-to-analog converter (DAC). (It is best if the data converter supplies the I/Q interface with built-in adjustment capabilities for amplitude and phase balance as well as DC offset correction.)

The key RF parameters for the modulator include output power at 1-dB compression (P1dB), output third-order intercept point (OIP3), carrier feedthrough, sideband suppression, and output noise. The linearity and output noise parameters of the modulator are critical parameters that set the system performance because they dictate the operating output range of the device and limit the maximum output power of the entire radio. Good values for these specifications would be +9dBm P1dB and -163 dBm/Hz output noise. For high peak-to-average ratio (PAR) signals like CDMA and OFDM, the modulator must pass the peaks of the signal without significantly degrading the adjacent channel power ratio (ACPR) performance that is critical to maintaining compliance to the standard.