Statistics - HP 54710A User's Reference Manual

Measurements

Statistics

Statistics
To confidently predict the reliability of a digital system, you must know the
statistics of its behavior. Singular measurements of setup time, hold time,
propagation delay and skew do not allow you to adequately predict the
probability of errors due to timing violations. However, a design based on
worst-case measurements can be too conservative. A worst-case
measurement does not show how frequently, or under what conditions, the
worst case will occur.
Measurement statistics in the HP 54720 give valuable insight. To view
statistics, select the Define Meas menu. The Statistics softkey allows you to
display either the minimum and maximum or the mean and standard
deviation of all selected measurements. The most recent value is always
displayed.
You need to know how much the oscilloscope's time-interval measurement
varies statistically because this affects measurements of the statistical
behavior of your system. Jitter introduced by the oscilloscope has two
components: a fixed part and a variable part. The fixed component of jitter is
the same at all points on the time base, including the trigger point. The
variable component increases as the delay from the trigger point increases.
To view the fixed part of the jitter, set the Position control in the Time Base
menu to zero. Set the Reference control in the Time Base menu to Center.
Set up the oscilloscope to view a fast-rising edge from a pulse generator, such
as an HP 8131A. The triggered edge should be at the center of the screen.
Press the Trigger menu key. You will see the trigger level marker displayed
on the screen. Ideally there would be no jitter (no horizontal spreading of the
trace) where the waveform intersects the trigger level. Any spreading is the
fixed part of the jitter.
13–26