Zero loss front end design targets cost/power/performance compromises

Zero loss front end design targets cost/power/performance compromises
Wireless LANs are growing in the marketplace as more users take advantage of the freedom to wirelessly access their data where it is most useful. Significant advances in RF technology underpin this wireless networking revolution.

One recent example is a zero-loss radio front end design for WLAN applications. We believe this approach offers a solution to the difficult tradeoffs and compromises inherent in the design of wireless LAN and similar RF solutions. Basically, it eliminates RF switch insertion losses and the resulting compromises, while reducing component count and cost. As a result, solutions using this technology can offer reduced power consumption, complexity and cost along with increased range and performance.

Conventional designs of half duplex radios typically use the same antennas for both receive and transmit. These antennas are connected to the receive and transmit channels of the RF transceiver via a high frequency switch, typically labeled "T/R Switch". In addition, wireless applications such as WLAN, operating in severe signal fading conditions, demand antenna diversity to improve signal reception. Antenna diversity is achieved by selecting, via a second high frequency switch (or the "antenna diversity switch"), one of two separately spaced antennas. Thus, this architecture has independent control of transmit and receive signal paths and the antenna employed through these two in-line high frequency switches.

As these high frequency switches — RF switches — are located after the power amplifier and are designed to operate under high current conditions. This high current requirement and the need to operate on a very wide frequency band while minimizing signal losses and part costs, results in a typical insertion loss of approximately 1.5dB for commercially available RF switches.

Since conventional radio front-end designs for half duplex systems implementing antenna diversity have two such switches between the antennas and the RF transceiver, the overall signal insertion loss incurred for this topology is about 3dB in both the receive and transmit channels.

The insertion loss of these RF switches should be minimized as it directly reduces the transmitter output power and decreases the receiver sensitivity. To cope with the transmitted signal attenuation, the typical solution is to increase total transmitter power. This simple approach has the drawback that it results in a more power- consuming transmitter, a cardinal issue for the battery-powered mobile wireless devices proliferating in today's market. As for the receive channel, the loss of receiver sensitivity decreases the effective range of the RF system, potentially affecting the costs of providing wireless coverage over a given area.

Therefore, the use of high frequency RF switches for providing antenna diversity, as well as antenna selection for transmit and receive duplex RF connectivity, has a tremendous impact on the performance and system costs of wireless LAN solutions. An innovative design approach is required to eliminate the performance and cost penalties incurred by the use of these switches to obtain the necessary RF functionality.

Enhancements to this conventional radio front-end design call for the use of less than two RF switches for some of the possible receive or transmit paths. However, the zero-loss front-end approach eliminates the transmit/receive loss due to the switches, while achieving the desired antenna diversity and transmit/receive switching functionality.

Cutting switches
More specifically, this zero-loss front-end design adds a third antenna specifically for transmission, eliminating the T/R switch. The approach also eliminates the antenna diversity switch via the addition of a second low noise amplifier (LNA) that enables an independent signal path for each of the two receiving antennas. Antenna diversity for the receive path is enabled by simply powering on the LNA of the antenna with the best signal to noise ratio and powering off the other one.

This approach offers significant additional advantages. The 3dB saving on the transmit path requires a transmitter power amplifier rated at only half the output power, reducing its power consumption by a factor of 2x. For mobile, battery-operated wireless devices, as well as other similar applications, the minimization of power consumption is a major advantage. Alternatively, the same power amplifier results in about 3dB more output power radiated by the antenna, increasing the range of the WLAN system. Overall, a much more power efficient solution is reached when using this zero-loss front end design.

A zero-loss front-end design is simplified with the EN303 RF chip by integrating low-noise amplifiers (LNAs) on the chip itself. This approach trades the cost of two switches for that of an LNA and an inexpensive printed or simple antenna to obtain higher performance, reducing the total bill-of-materials cost. Source: Envara Inc.

On the receive path, removing the 3dB insertion loss of the RF switches results in an equivalent 3dB gain in receiver sensitivity. For indoor, wireless LAN and other similar applications, this gain translates to an approximately 25% increase in indoor coverage range. Such a large increase in coverage range can have significant ramifications for network implementers using this technology as it can reduce the number of required access points for a given structure or area. Alternatively, receive path loss margin is increased by 3dB, allowing room for design margin, strengthening the robustness of the WLAN system, and enabling greater flexibility for the OEM product designer.

High frequency switches tend to be more complex than LNAs, so the elimination of switches in place of LNAs reduces system complexity. In terms of BOM costs, this approach trades the cost of two switches for that of a LNA and an inexpensive printed or simple antenna to obtain higher performance. In addition, since high frequency switches are manufactured with GaAs technology, these external components can not be integrated onto a low-cost silicon-based RF chip.

For example, the EN303 RF chip provides LNA integration into the RF chip, therefore avoiding the use of high frequency switches. The overall result of this approach, when integrating the LNAs into the RF transceiver, is a reduced number of components, reduced manufacturing complexity and reduced BOM costs through the creation of a more highly integrated RF chip solution. Thus, it saves overall chipset cost and provides better RF performance.