Tools tackle complex modulation

Between the dancing frequencies of Bluetooth and WiFi and the bursty transmissions of RFID and ultra-wideband, radio communications systems have become hard to test. Swept-spectrum analyzers are too slow to catch transient signals and cannot show the timing of frequency-hopping transmissions. Vector signal analyzers, intended for studying modulation patterns, have limited RF performance.

The bandwidth of high-speed communications schemes also adds to the difficulty of measuring wireless signals. Rates up to 800 Mbps will soon be broadcast from space to business rooftops. Satellite television is making use of complex digital error correction coding and phase-shift keyed modulation to transmit scores of digital television channels at high resolution with low signal levels.

To fill this growing gap between tool capabilities and RF communications complexity, companies are creating a new type of spectrum analyzer that provides a variety of modes for data display. They are also introducing frequency-based triggering to RF analysis. Display modes include parameter strip charts and waterfall displays of frequency and power (Figure 1) that show how the spectral characteristics of signals evolve over time.

Figure 1. The spectrogram view of the real-time spectrum analyzer shows signal intensity (color-coded) as a function of time and frequency, simplifying the analysis of changing signals. Courtesy of Tektronix.

These tools employ an old concept with several new twists. Fundamentally, the instruments use a tunable band-pass filter front-end to select a frequency band of interest, then digitize the signal for analysis. In this respect the instruments are similar to existing digital spectrum analyzers. One of the new twists, however, is that the instruments tag the data they capture with time markers before storage.

Tagging the data allows for multiple correlated views of the information. Users can, for instance, look simultaneously at a transient signal's waveform, its instantaneous spectrum, and a time history of spectral power distribution. This combination is useful for tracking a sequence of RF signals, such as frequency-hopping communications or an RFID query and response. The timing, strength, and spectrum of signal bursts are instantly apparent yet can be correlated to the more traditional spectral power and digital oscilloscope type displays for making detailed measurements.

This ability to generate a time history is key to another of the instrument's twists: frequency-based triggering. Designers can set the instrument to gather data only after a specific pattern of signals occurs, such as a Bluetooth link setup sequence. The trigger pattern can also be used to filter data collection, so only signals of interest get recorded and not the intervals between them. Because the trigger uses amplitude as well as frequency, the instrument can monitor an ongoing signal and capture data only when it changes.

At least two such instrument series have entered the market: the Real-time Spectrum Analyzer (RSA series) from Tektronix (Beaverton, OR; www.tektronix.com) and the Celerity Broadband Signal Analyzer (CS35000 series BSA) from Aeroflex (Plainview, NY; www.aeroflex.com). These instruments have similar display and analysis capabilities, but they differ in configuration and performance.

The Tektronix RSA series devices are self-contained, bench-level instruments with built-in display and capture memory. They have a signal bandwidth as great as 8 GHz, with a capture bandwidth as much as 15 MHz wide. The standard memory depth is 64 Mbytes, but a 256-Mbyte option is available. Users manipulate the instrument with front-panel controls or through a remote computer.

The Aeroflex Celerity BSA series instruments have a modular structure that offers several configurations. The core instrument runs Windows 2000 and can be remotely operated over a GPIB or Ethernet link, or manipulated locally using an optional keyboard and display. Onboard data storage can hold as much as 10 s of data capture. The unit can also stream its data to an onboard disk drive to store longer capture intervals.

The different configurations available for the BSA series start with a 40-MHz capture bandwidth but include the ability to run data-acquisition channels in parallel with an offset in sample time between channels. That offset makes the parallel channels act like a single high-speed analog-to-digital converter (ADC) for data capture. As a result, the BSA series can offer a capture bandwidth as great as 600 MHz, which means it can record and analyze entire communications frequency bands, such as LINK satellite channels.

Both companies' instruments offer insights into complex modulation and frequency-agile signals that traditional instruments cannot. They will not replace traditional analysis tools, however. Swept-spectrum analyzers provide superior resolution and dynamic range when working with stable signals, and vector signal analyzers are better at digital modulation analysis. But the ability to see the interrelationships of signal frequency, amplitude, and timing makes frequency-hopping and complex modulation schemes easier to understand.