Tektronix TDS5052 User Manual

Creating and Using Math Waveforms
3- 172
TDS5000 Series User Manual
There are only two samples per cycle of a signal that have a frequency equal to
one half of the sample rate. This is the highest nonaliased signal that can be
output from the spectral analyzer. Thus, at this frequency, one sample of
acquisition jitter will show up in the spectrum as 180 degree phase variations.
Therefore, when you analyze phase, make sure to oversample the signal
sufficiently such that the one sample interval of jitter that is showing up in the
phase is acceptable to the accuracy you want to achieve.
Effects of Average and High Res acquisition modes: The result of averaging the
time domain acquisition using either Average mode or Hi Res acquisition mode
affects the frequency response of the oscilloscope. This is due to the one sample
of jitter in the acquisition system. Both Hi Res and Average acquisition modes
have the same affect on the frequency response. These modes cause the response
to roll off from a magnitude value of one at DC to a magnitude value of 0.63 at
Nyquist which is the frequency equal to one half of the sample rate. This is true
regardless of the real time sample rate setting.
Frequency Domain Averaging: Turn on averaging for a math waveform by
editing the math expression. Sometimes, it is more desirable to average in the
spectrum than in the time domain. For example, consider a signal that has time
domain components that are asynchronous to the trigger. If you turn averaging in
the time domain, these components may go to zero or produce strange nondeter-
ministic effects on the resultant waveform. Thus, these signal components may
end up not appearing in the spectrum. However, if averaging is done in the
frequency domain instead, then these components will be present.
Aliasing occurs when the input frequency of a signal is greater than one half of
the sampling frequency (the sample rate).
Set the sample rate high enough so that the signals in the spectrum appear at
their correct frequency as opposed to a lower aliased frequency value. Also,
complex signal shapes that have many harmonics in them, such as a triangle or
square wave, can appear to be OK in the time domain when in fact many of the
harmonics in that signal are aliased.
One way to check for aliasing is to increase the sample rate and observe whether
any of the harmonics unwrap to different frequency locations.
Another way to recognize aliasing is to realize that higher order harmonics
usually have decreasing magnitudes compared to lower order harmonics. Thus, if
you see a series of increasing harmonic magnitude values as frequency increases
then you can suspect that they may be aliased. In the spectral math waveform,
the actual higher frequency components are undersampled, and therefore they
appear as lower frequency aliases that “fold back” around the Nyquist point. (See
Figure 3--58.) You may test by increasing the sample rate and observing if aliases
unwrap to different frequency positions.
Recognizing Aliasing

Tektronix TDS5052 Specifications

General

Brand	Tektronix
Model	TDS5052
Category	Test Equipment
Language	English