[jitter-timing] Rise & Fall Times - Part 3
- From: "JitLab" <jitlab@xxxxxxxxxxx>
- To: <jitter-timing@xxxxxxxxxxxxx>
- Date: Tue, 23 Sep 2003 19:49:52 -0600
Rise & Fall Times - Part 3
Measurement of Rise Time & Fall Time:
The measurement system should have enough analog bandwidth to measure the
edge times faithfully. This means that the oscilloscope and the probe
together should have at least twice the bandwidth of the fastest edge rates
to be measured. To estimate the highest frequencies of significant level in
an edge, divide 0.35 with an estimate of the edge time (i.e. rise time or
fall time).
If a digital oscilloscope is being used, an edge should get at least 5
samples for the measurements to be reasonably accurate. Normally, the lesser
bandwidth the measuring system has, the larger the measured rise or fall
time will be from the actual value. This means that the measured value for
rise & fall rates will always be larger than the actual value. However, this
is not always true. If the probe used for measurement is highly
under-damped, it will result in ringing causing the measured signal to have
faster rise & fall times than the actual signal.
You have to use cursors to measure the rise & fall times in analog
oscilloscopes. In modern digital oscilloscopes, the measurements are
automatically done by the oscilloscopes. However, all digital oscilloscopes
measure only one edge in the whole acquisition (normally first). To have a
higher confidence in the measurement, multiple measurements must be made.
One way is to let the oscilloscope run in continuous run mode and let the
statistics accumulate over time. This approach may be time consuming and
will not measure consecutive edges.
The better way is to use a post processing software. These software are very
fast and can acquire data from the oscilloscope automatically, process
millions of data edges in just a few seconds, display the results
graphically in time domain as well as in frequency domain, with statistics
along with the time stamp of max and min values. You can even save waveforms
from the oscilloscope for post processing by the software.
Links:
Measurement & Analysis Software for Timing & Jitter -
http://www.aubraux.com/free-jitter-software.php
Measurement Definitions - http://www.aubraux.com/jitter/measurements.php
Other related posts:
- » [jitter-timing] Rise & Fall Times - Part 3
