Generate a pulse train, with the pulse defined either by a function or a sampled pulse.


pt = pulstran(t, d, 'func')

pt = pulstran(t, d, 'func', p1, p2, ...)

pt = pulstran(t, d, p)

pt = pulstran(t, d, p, fsp)

pt = pulstran(t, d, p, fsp, method)


The times at which to generate the pulse train.
Dimension: vector
When d is a vector it contains the time delay for each pulse in the train. See comments.
When d is a matrix, the first column contains the time delay for each pulse, and the second column contains the amplitude for each pulse.
Dimension: vector | matrix
The function to generate the individual pulses.
Dimension: string
The options are as follows.
  • 'rectpulse' (default)
  • 'tripuls'
  • 'gauspuls'
p1, p2, ...
The optional parameters to pass to 'func'.
Dimension: scalar
The sampled pulse to be replicated in the train.
Dimension: vector
The sampling frequency for p.
Dimension: vector
The interpolation method used with replicated p pulses.
Dimension: string
The options, as follows, are those for interp1.
  • 'linear' (default)
  • 'pchip'
  • 'spline'


The sampled pulse train.

Example 1

Plot a train of 2 ms rectangular pulses at 5 ms intervals.

fs = 20000;		% sampling rate of the output
ts = 1 / fs;
fp = 200;		% pulse rate
tp = 1 / fp;
t = [0:ts:4*tp];
d = [0:tp:4*tp];
w = .4 * tp;
p = pulstran (t, d, 'rectpuls', w);
plot(1000 * t, p);
title ('Train of 2 ms rectangular pulses at 5 ms intervals');
xlabel ('Time (ms)');
ylabel ('Amplitude');
Figure 1. pulstran figure 1

Example 2

Plot a train of user defined at 5 ms intervals, with increasing amplitude.

fs = 10000;		% sampling frequency of the output
ts = 1 / fs;
fp = 200;		% pulse frequency
tp = 1 / fp;
t = [0:ts:4*tp];
d = [0:tp:4*tp] + 0.5*tp;
np = length(d);
d = [d',[1:np]'];
p = [0.010 0.134 0.409 0.800 1.000 0.800 0.409 0.134 0.010];
fsp = 3200;		% sampling frequency of p
x = pulstran (1000 * t, d, p, fsp, 'pchip');
plot(t, x);
xlabel ('Time (ms)');
ylabel ('Amplitude');
Figure 2. pulstran figure 1


It is important to note that when using 'func' the delay times apply to the centers of the pulses, but when using p the delay times apply to the left endpoints of the pulses.

When using 'func' = 'gauspuls', gauspuls('cutoff', fc, bw, bwr, tpr) is called to set the pulse width, so 'tpr' can be passed to pulstran as 'p4'.