Stanford Research Systems SR785, Operating Manual And Programming Reference

Page: 108 / 546

2-28 Real Time Bandwidth and Overlap
SR785 Dynamic Signal Analyzer
Typically, time records with 50% overlap provide almost as much noise reduction as
non-overlapping time records when RMS averaging is used. When RMS averaging
narrow spans, this can reduce the measurement time by 2.
Time Record Increment
The Time Record Increment is how far the start of each time record is advanced between
measurements (as a percentage of the time record length).
If the increment is 100%, the start of the next time record is exactly one time record
advanced from the start of the previous time record (no overlap).
If the increment is 25%, then the next time record starts 1/4 of a time record advanced
from the start of the previous time record. This is referred to as 75% overlap since the
two time records share 75% of a record. The overlap is simply 100% minus the Time
Record Increment.
When the Time Record Increment is less than or equal to 100%, the measurement is real
time. All time points contribute to one (or more) measurements.
If the increment is 200%, the start of the next time record is advanced from the start of
the previous time record by 2 time records leaving a gap of 1 record. This means that the
data between the two time records is not measured.
When the Time Record Increment is greater than 100%, then the measurement is not real
time and some time points do not contribute to a measurement.
The actual time record increment for the measurements in progress is displayed in the
Real Time indicator in the status area (below the Input Ranges). If the indicator shows a
value greater than the requested Time Record Increment, it means that the measurement
cannot be made with the requested increment but is running with the smallest increment
possible. Factors which affect the processor’s ability to run real time include the
measurement type, averaging and source type.
Settling
When the frequency span or input signal path (gain, filtering, etc.) is changed, a settling
time is required before the FFT measurement is considered settled or valid. The
measurement is not settled until the discontinuity in the input data has propagated
through the digital filters and a complete new time record has been acquired.
If the time record increment is 100%, unsettled measurements are not displayed. After a
change is made which unsettles the measurement, new data is not displayed until the
filters are settled and a complete time record has been acquired.
If the measurement is running with a time record increment less than 100% and the
measurement is unsettled, unsettled measurements may be displayed. New data is
displayed after the filters are settled and a portion of the new time record has been
acquired. For example, if the time record increment is 25%, 3 measurements are made
before a complete new time record has been acquired. These first 3 measurements have
time records which contain data from before AND after the measurement was unsettled.
These unsettled measurements are displayed in half intensity indicating that the