Oscilloscope Basics: Stabilizing Waveform Display, Pt. 1

Figure 1: A free running oscilloscope starts each
acquisition at a different point on the waveform,
resulting in an unstable display.  A triggered oscilloscope
starts each acquisition at the same point on the
waveform, resulting in a stable display. 

An unsynchronized, unstable oscilloscope display is useless for making measurements, but proper triggering can synchronize the oscilloscope sample clock to specific waveform events so that the acquired waveforms appear stable on the display.  Let’s look at why signals can appear unstable and what to do about it.  

Oscilloscopes are sampling devices; they sample the incoming signal at a uniform rate.  The timing of a signal applied to the input of an oscilloscope is most probably asynchronous with the oscilloscope’s sampling clock.  If the oscilloscope timebase is allowed to run free—that is, not synchronized to the timing of the input signal—then each oscilloscope acquisition potentially begins at a different point on the input waveform, as shown in Figure 1.

Oscilloscopes are also a batch processing devices. They acquire a fixed number of samples, then process and display the data before beginning the next acquisition. If the oscilloscope timebase is running free because a trigger has not been defined, then those acquisitions that all began at different points on the waveform will appear unstable when displayed sequentially, like a flickering old-time movie whose frames are subtly misaligned.

Triggering means that the oscilloscope doesn’t just keep churning out acquisition after acquisition at whatever rate its clock allows, it waits for the appearance of a specific event before starting the next acquisition.  In the case of the Edge trigger shown in Figure 1, that event is the input signal crossing a specified trigger level on the positive slope.  

Note that the oscilloscope has been sampling data all along, storing it in the acquisition buffer. What the trigger actually does is tell the oscilloscope from what start sample point/time to what end sample point/time to “snapshot” and display on the screen. That range of samples is considered the acquisition, which, thanks to the trigger, has been synchronized to the input signal so that the samples occur at consistent points on the waveform from acquisition to acquisition.

After the acquisition is complete, the oscilloscope “processes” that set of samples before displaying it.  As a result of the trigger defining the start of the acquisition, the processing always begins at the same point on the waveform.  The displayed waveforms therefore appear stable, as more or less the “same” sample values occur at the same time on the display. The measurements made on those acquisitions are also showing the results of a consistent data set.

Stabilizing the display with an Edge trigger is simple enough if you are sampling repetitive waveforms that do not much change cycle to cycle. In fact, you’re never actually without a trigger, as the oscilloscope defaults to a positive-going Edge trigger at the 50% amplitude “zero” level. However, quite often waveforms can show many deviant anomalies that challenge what is the 50% amplitude, such as with noisy signals, or are not simple sine waves or pulse trains that are going to consistently hit that level following roughly the same lapse of time. It may well be that your signal requires that you trigger on a different property than a 50% edge to consistently define the start of each acquisition.

In that case, other types of triggers or various types of signal conditioning can help tell the oscilloscope what samples to ignore when looking for the place to start acquisition, which we’ll discuss more in our next post.