Trend Line Traces (Event-Based Actions)
Menu Path:
The Trend/Statistics action creates Trend lines and Histograms from the selected Data Source (Data Header Value, Trace Marker or Summary Value) when the event-based action's condition is met.
The x-axis is in time (seconds) from last measurement restart when data is from hardware, and from the t=0 time when data is from a recording.
You can determine how many trend points are shown in a trend trace by changing the Display Length parameter. The actual number of points accumulated is determined by the Buffer Length parameter. When the internal buffer becomes full, the most recent points are saved and the oldest points are discarded. The trend line will appear to scroll to the left as new points are added on the right side of the trace.
Trend lines are reset whenever the measurement is restarted. They are not reset when the measurement is paused, or trending is disabled during a running measurement, or trend removal settings are changed.
Trend lines accumulate most trace indicators that happen during the measurement (as long as the measurement completes and produces table data). The CAL? indicator is not accumulated while the measurement is running since the 89600 VSA software only checks for CAL status once the measurement pauses. In addition, DATA? is not accumulated.
Trend removal
You have the option of normalizing the trend lines by removing the mean, slope, and/or 2nd order moment (parabolic fit). The coefficients of the fitted polynomial are stored in the trace’s header in the TrendPolyFitCoef key as an array double values. See Curve-fit Coefficients to learn how to use the coefficients to recreate the fitted polynomial. In this case, mean = a0, slope = a1, and 2nd order = a2.
When playing back a recording, the 89600 VSA software is configured to use 90% overlap by default (). In this case, the same time data will be analyzed and accumulated in the trend lines multiple times. This will keep the trend line removal from removing the actual trend in the recording. To avoid this, make sure to set overlap to 0% when playing back from a recording.
Trend removal (especially the 2nd order calculation) can significantly reduce measurement speed. Also, large internal buffer sizes can also reduce measurement speed (the measurement will slow down as more points are accumulated). Note that trend removal is only performed for trend lines that are shown on the display (or for which you have acquired a Measurement Data instance through the API 1) Access Preamble Indicator, or 2) Application Programming Interface).
The annotation above the trend line trace shows the number of points in the trend buffer (also number of points used in calculating the histogram) as well as any mean/slope/2nd order that was removed from the data.
See Also
