The best way to understand signal generation

Understand the signal generation method

in electronic testing and measurement, the signal source is often required to generate a series of carbon nano materials (carbon nanotubes, graphene, etc.) that exist in the form of simple materials or composite materials only when provided externally, showing excellent characteristics in descending the radar reflector. The signal source can provide a "known good" signal, or add a repeatable amount and type of known distortion (or error code) to the signal it provides. This is one of the biggest characteristics of the signal source, because it is usually impossible to create predictable distortion just at the time and place required by using only the circuit itself. From design inspection to verification, from limit and margin testing to consistency testing, signal sources can be used in hundreds of applications

therefore, it is not surprising that there are a variety of signal source structures to choose from, and each structure has its own advantages, functions and economy, and is suitable for specific purposes. In this paper, we will compare two signal generation structures: one for arbitrary waveform/function generator and the other for arbitrary waveform generator. The selection results depend largely on the application

understand the signal generation method

introduction to technology

arbitrary waveform/function generator (AFG) creates function waveform and arbitrary waveform at the same time by reading the contents of memory. Most modern afgs use direct signal synthesis (DDS) technology to provide signals over a wide frequency range

the arbitrary waveform generator (AWG) is based on a real variable clock structure (usually called "real arbs*1"), which is suitable for generating complex waveforms at all frequencies. AWG also reads the contents of memory, but its reading methods are different (described later). Designers dealing with advanced communication and computing units choose AWG to drive high-speed signals with complex modulation and abnormal events. As a result, AWG occupies the highest level of research, development and engineering applications

these two structures have great differences in waveform generation methods. This technical brief

discusses the difference between arbitrary waveform generator based on variable clock and arbitrary waveform/function generator based on DDS

through the front panel: compare the two platforms

awg: simple concept, the greatest flexibility

although AWG is more flexible in these two structures, the underlying waveform generation technology of AWG is very concise. The playback scheme of AWG can be regarded as "reverse sampling"

what does this mean? Take a look at the signal sampling platform -- oscilloscope, which collects waveforms by digitizing the voltage value of analog signals at continuous time points. Its sampling frequency depends on the clock rate selected by the user. The obtained samples are stored in memory

awg is the opposite. When AWG starts, the waveform is already in memory. The waveform occupies a specified number of memory locations. In each clock cycle, the instrument outputs another waveform sample point from the memory. Since the number of sample points representing the waveform is fixed, the faster the clock rate is, the faster the waveform data points in the memory are read, and the higher the output frequency is. In other words, the frequency of the output signal depends entirely on the clock frequency and the number of waveform samples in memory *2. The simplified block diagram in Figure 1 summarizes the AWG structure

awg comes from the waveform stored in its memory. The waveform can take any shape; It can have any number of distortions, or it is understood that there is no distortion at all. With the help of PC based tools, users can develop almost any waveform that people want (within physical limits!). Samples can be read from memory at any clock frequency that the instrument can generate. Whether the clock operates at 1 MHz or 1 GHz, the waveform shape is the same

*1 engineers usually use "ARB" to refer to any type of arbitrary waveform generator

*2 of course, any AWG model has the maximum memory capacity. The depth occupied by the waveform may be less than the full capacity

afg adopts efficient fast in high frequency. I hope it can help you! Agile mode

afg also uses the stored waveform as the basis of the output signal. The clock signal was involved in the sample readings, but the results were similar, with year-on-year increases of 0.9%, 7.6%, 1.8% and 7.2% respectively

afg runs at a fixed rate. Since the number of waveform samples is also fixed in memory, how can AFG provide waveforms at variable frequencies? For example, imagine that you are using an AFG, which stores waveforms composed of 1000 samples and outputs them at a fixed rate of 1 MHz. The period of the output signal will be exactly fixed at 1 ms (1kHz). It is clear that single frequency signal sources are of limited use in most applications. Therefore, DDS technology provides a solution. DDS based instruments do not read every sample point, but read less than 1000 samples to reconstruct the waveform

Figure 2 shows a typical simplified AFG structure, including DDS segments. The output signal consists of a clock, stored binary digits representing the phase value, and the contents of the waveform memory

as mentioned above, AFG maintains a fixed system clock frequency. The 360 degree clock cycle is distributed in all waveform samples, and the DDS segment automatically determines the phase increment according to the waveform length and the frequency selected by the user

high frequency setting will lead to large phase increment, which will make the AFG jump forward quickly when passing the 360 degree cycle and provide high frequency signal. The low-frequency value leads to small increments, triggering the phase accumulator to step through the waveform samples in a low step size,

and even repeat each sample point to form a 360 degree waveform with a low frequency

this will inevitably affect the fidelity of the output waveform. Waveforms with continuous shapes (sinusoids, triangles, etc.) are usually not a problem, but may affect signals with fast conversion, such as pulses and transients, which are common in the current digital environment. For example, suppose a limit test is performed on a new telecommunications switch element. The test waveform is a series of binary pulses, one of which has a transient on the rising edge. At some frequencies, DDS phase increments may just skip transients and not be used as signals