Real Time Bandwidth and Overlap 2-27
SR785 Dynamic Signal Analyzer
Real Time Bandwidth and Overlap
What is Real Time Bandwidth?
The Real Time Bandwidth is the largest frequency span whose corresponding time
record exceeds the time it takes to compute the FFT measurement. At this span and
narrower spans, it is possible to compute the measurement for every time record without
skipping any input data. The spectra are computed in "real time".
At spans larger than the real time bandwidth, the measurement computations take longer
than a time record. The analyzer can not keep up and some input data must be ignored.
Data is lost between time records while the computations are in progress.
For FFT Resolutions greater than 100 lines, the SR785 can usually compute the
measurements in both displays in less than a time record for all spans. This includes the
real time digital filtering and heterodyning, the FFT processing, and averaging
calculations. The real time bandwidth in this case is 102.4 kHz. Every input sample
contributes to a time record and a measurement.
Averaging Speed
Real time bandwidth has a direct affect on measurement times when Averaging is On.
Consider the averaging of full span (102.4 kHz) 400 line FFT’s. If the real time
bandwidth is only 10 kHz (as in many analyzers), it takes 40 ms to compute the
measurement for a 4 ms time record. This means 9 out of every 10 time records are
ignored and only 25 full span measurements can be computed in a second. When
averaging is on, this usually drops to 10 measurements per second. At this rate it's going
to take a couple of minutes to do 500 averages.
The SR785, on the other hand, can make real time measurements at full span (102.4
kHz). This results in 256 measurements per second (on each display!). In fact, this is so
fast, that the display can not be updated for each new measurement. The display only
updates about 8 times a second. However, when Averaging is On, all of the
measurements contribute to the average. The time to complete 500 averages is only a few
seconds. (Instead of a few minutes!)
Overlap Processing
What about narrow spans where the time record is long compared to the processing time?
The analyzer computes one measurement per time record and waits and does nothing
until the next time record is complete. The update rate is one measurement per time
record. With narrow spans, this is quite slow.
With overlap processing, the analyzer does not wait for the next complete time record
before computing the next measurement. Instead, it uses data from the previous time
record as well as data from the current time record to compute the next measurement.
This speeds up the update rate. Remember, most window functions are zero at the start
and end of the time record. Thus, the points at the ends of the time record do not
contribute much to the FFT. With overlap, these points are "re-used" and appear as
middle points in other time records. This is why overlap effectively speeds up averaging
and smoothes out window variations.
Summary of Contents for SR785
Page 4: ...ii ...
Page 10: ...viii ...
Page 80: ...1 64 Exceedance Statistics ...
Page 158: ...2 78 Curve Fitting and Synthesis SR785 Dynamic Signal Analyzer ...
Page 536: ...5 136 Example Program SR785 Dynamic Signal Analyzer ...