Asignal generator is what the name implies: a generator of signals used as a stimulus for electronic tests. Most circuits require an input test signal with time-varying amplitude for characterization. It may be a true bipolar alternating-current (AC) signal with peaks oscillating above and below a ground reference point. Or it may vary over a range of direct-current (DC) offset voltages that are either positive or negative. The signal may be a sine wave or other analog function, digital pulse, binary pattern, or a purely arbitrary wave shape. For its part, the signal generator can provide "ideal" waveforms or it may add known, repeatable amounts and types of distortion (or errors) to the signal (Fig. 1).
Signal generators have hundreds of different applications. In the electronic measurement context, however, they fall into three basic categories: verification, characterization, and stress/margin testing. On the verification front, for example, wireless-equipment designers developing new transmitter and receiver hardware must simulate baseband inphase (I) and quadrature (Q) signals with and without impairmentsto verify conformance with both emerging and proprietary wireless standards. Some high-performance arbitrary waveform generators (AWGs) can provide the needed low-distortion, high-resolution signals with two independent channels: one for the "I" phase and one for the "Q" phase. Sometimes, the actual RF signal is needed to test a receiver. In this case, AWGs with sample rates to 24 GSamples/s can be used to directly synthesize the RF signal.
Another common application is the stress testing of communication receivers. Engineers working with serial-data stream architectures need to stress their devices with impairmentsparticularly jitter and timing violations. Advanced signal generators eliminate hours of calculation by providing efficient built-in jitter editing and generation tools. These instruments can shift critical signal edges as little as 200 fs (0.2 ps).
There are several ways to create waveforms with a signal generator. The choice of methods depends upon the information available about the DUT and its input requirements, whether there is a need to add distortion or error signals, and other variables. Modern high-performance signal generators offer at least three ways to develop waveforms:
Create brand new signals for circuit stimulus and testing;
Replicate: Synthesize an unavailable real-world signal (captured from a test instrument); and
Generate ideal or stressed reference signals for industry standards with specific tolerances.
The arbitrary/function generator (AFG), which serves a wide range of stimulus needs, is the prevailing signalgenerator architecture in the industry today. Typically, this instrument offers fewer waveform variations than its AWG equivalent but with excellent stability and fast response to frequency changes. If the DUT requires the classic sine and square waveforms (to name a few) and the ability to switch almost instantly between two frequencies, the AFG is the right tool. An additional virtue is the AFG's comparatively low cost, which makes it very attractive for applications that do not require an AWG's versatility.
Whether the engineer wants a data stream shaped by a precise Lorentzian pulse or a complex modulated RF signal to test a GSM- or CDMA-based handset, the AWG can produce virtually any waveform. To create the desired output, the engineer can use a variety of methods ranging from mathematical formulae to "drawing" the waveform. Fundamentally, an AWG is a sophisticated playback system that delivers waveforms based on stored digital data that describes the constantly changing voltage levels of an AC signal. Yet its block diagram is deceptively simple (Fig. 2).
To understand the AWG, it's necessary to grasp the broad concepts of digital sampling. Such sampling involves defining a signal using samples, or data points, that represent a series of voltage measurements along the slope of the waveform. These samples may be determined by measuring a waveform or using graphical or mathematical techniques. Using stored information, the AWG reconstructs the signal and feeds the data points through a digital-to-analog converter (DAC).
Together, stimulus and acquisition instruments make up a complete solution that can drive a device-under-test with complex real-world signals and acquire the resulting outputs. Signal generators provide the best way for the designer to control what goes into the device. This control is often needed in order to make sense of what comes out of the device.
Similarly, the signal generator makes margin testing and characterization possible. Working with a signal generator and test instrument, engineers can explore the limits of their design's performance. In doing so, they introduce deliberate stresses with the source and measure the results. While an AFG is ideal for certain applications, the AWG offers a degree of versatility that few instruments can match. With its ability to produce any waveform imaginable, the AWG supports the full gamut of microwave and RF test requirements.