29 March 2021

How to "Layer" Measurement Tools

Figure 1: Multi-grid display "layers" multiple measurement tools to find hidden glitch.
Figure 1: Multi-grid display "layers" multiple
measurement tools to find hidden glitch.
Teledyne LeCroy oscilloscopes have four, distinct sets of measurement tools, including measurement graticules, cursors, parameters and graphs. These tools developed historically and are designed to be “layered” on multi-grid MAUI® oscilloscopes so that each addition brings a new level of understanding and insight. Even on oscilloscopes that do not have multi-grid displays, as shown here, several measurement tools can be applied at once for added insight. Read on to see how, properly combined, they can help you find waveform anomalies and assess their frequency of occurrence in a few, simple steps.

The oldest of these tools, developed in the original analog oscilloscopes, is the use of an overlaid calibration graticule. Most oscilloscope graticules have ten horizontal division and eight vertical divisions.  By counting the number of boxes between either vertical or horizontal events on the trace and multiplying by the appropriate scale factor, you can estimate the amplitude or time duration between events.

Figure 2: The relative vertical cursor marks the amplitude levels of two horizontal cursor lines and displays them in the trace descriptor box along with the difference in amplitude.
Figure 2: The relative vertical cursor marks
the amplitude levels of two horizontal cursor
lines and displays them in the trace descriptor
box along with the difference in amplitude.
Cursors are moveable vertical or horizontal lines projected onto the graticule that can be aligned with significant trace features.  Single, absolute cursors show the value of the vertical or horizontal point where they are positioned. Dual, relative cursors show the value of and the difference between the two points where they are positioned. Figure 2 shows an example of a relative vertical cursor measurement of the peak-to-peak amplitude of a clock burst. The absolute amplitude of each of the cursors relative to the trace offset, along with the difference between the cursors, is posted on the trace descriptor box, as shown in the detail.

Timing measurements can be made using the relative horizontal cursors by placing them on two points of a single trace or multiple traces, as shown in Figure 3. The cursors read the amplitudes at the cursor crossings, as well as the absolute time relative to the trigger time. Amplitude readouts are in the trace descriptor box, while the time readouts measuring delay between the two traces appear beneath the timebase and trigger descriptor boxes.

Figure 3: Relative horizontal cursors measure timing between two signals.
Figure 3: Relative horizontal cursors measure
timing between two signals.
Cursors are a quick method of obtaining measurements within one acquisition, but they are not very exact. Their accuracy depends on the placement of the cursor lines on the trace. 

Automatic measurement parameters are based on industry standard definitions of how to calculate amplitude, width and other typical waveform measurements.  They are computed from each occurrence of the parameter over multiple acquisitions, up to millions of measurements. Figure 4 shows six, typical parameters being used to characterize a pulse width modulated waveform.

Figure 4: Six measurement parameters with statistics used to characterize a pulse width modulated signal.
Figure 4: Six measurement parameters with statistics
used to characterize a pulse width modulated signal.
On all Teledyne LeCroy MAUI oscilloscopes, key statistical values are automatically computed for each parameter, including the mean, minimum, maximum, and standard deviation. You can choose to display these statistics on the Measure table in order to gain an immediate and highly accurate view of how this parameter has ranged through many acquisitions, as opposed to the cursor readout that will only show you the measurement taken on the last acquisition. In fact, with a repetitive trigger, cursor readouts are nearly impossible to read, they change so rapidly, whereas the statistics on the measure table are usually stable enough to actually read.

Now this is where it starts to get interesting—look at the statistics for parameter P3 reading the width of trace C1 in Figure 4 (you can click on the image to expand it).  It has a mean value of 190 µs and a range of values from167 ns to 240 µs. Below the parameter statistics is an iconic histogram, called a histicon, which shows the distribution of the parameter values.  The histicon for parameter P3 shows the bulk of the measured values are towards the upper range, but there is a tiny outlier near the minimum value at the far left.  

Figure 5: Track of the width parameter helps locate the outlier shown on the P3 histicon.
Figure 5: Track of the width parameter helps
locate the outlier shown on the P3 histicon.
Most oscilloscopes have the ability to graph the measured values either versus measurement instance, which is called a trend plot, or versus time, which is called a track plot. In Figure 5, a track plot of P3 has been laid over the acquired pulse signal. Note the sharp dip in the track near zero seconds. This is the location of the outlying minima that showed up in the histicon, which first suggested to us there might be something of interest to be revealed if we looked at the data another way. Simply zooming on the dip in the track shows where the event occurred in the signal.

On some oscilloscopes, we could generate a full histogram of the parameter simply by touching the histicon. Histograms have their own measurement parameters and cursors, which provide additional insight into the statistical distribution, such as what is the min/max value occurring within a particular bin of the histogram, or how many measurements total fall within a given bin. 

Figure 1 at the start of the post shows all these measurement tools “layered” into a multi-grid display combining the acquired waveform, measurement parameters with statistics and histicons, the track and histogram of the width parameter, as well as a zoom centered on the minima event marked in the track. 

The zoom of the signal (upper right Figure 1) shows the narrow glitch outlier, which is only 166 ns wide and otherwise quite easy to miss. The histogram indicates that 30 of them have occurred in the 770 measured width values, so we know it occurs about 4% of the time.  By using the measurement tools in combination, the presence of the narrow glitch, its location, and its frequency of occurrence have all been learned in a matter of minutes.

Watch Steve Murphy demonstrate more uses of oscilloscope display and measurement tools in the on-demand webinar, “Oscilloscope Coffee Break Series: Part 2, Optimizing Your Display & Using Cursors and Measurements”.

See Also:

Getting the Most Out of Your Oscilloscope, Cursors and Parameters

Getting the Most Out of Your Oscilloscope, Tracks and Trends

Using Histograms, Part I

Using Histograms, Part II

Using Histograms, Part III

Using Histograms, Part IV

No comments:

Post a Comment