To find Group delay : Error using diff Difference order N must be a positive integer scalar.

To find Group delay : Error using diff Difference order N must be a positive integer scalar.

I am trying to find the group delay of a modulated signal , which is derivative of phase delay , but while using diff iam getting this error "Error using diff Difference order N must be a positive integer scalar" , can anyone sugget method to do this

this is the code i tried to do

close all;
dt=.0001;
fs=1/dt; %sampling frequency
fn=fs/2;
n=100;
t=dt*(-n/2:n/2); %time base

f0 = 200; %to modulate GP multifly by cos(2ft*pi)
xt = cos(2*pi*t*f0) .* cos(2*pi*fn*t); 
y = hilbert(xt);
m = abs(y);
g = angle(y);
ph = -g/(2*pi*f0);

gd = diff (ph,(2*pi*f0));
figure
subplot(3,1,1); 
plot(t,xt,'b'); 
title('signal');
xlabel('Time(s)'); ylabel('Amplitude');

subplot(3,1,2); 
plot(t,ph,'b'); 
title('phase delay');
xlabel('Time(s)'); ylabel('phase');

subplot(3,1,3); 
plot(t,gd,'b'); 
title('group delay');
xlabel('Time(s)'); ylabel('phase');

[Walter Roberson]

ph = -g/(2*pi*f0);
gd = diff (ph,(2*pi*f0));

What was your hope that this would accomplish?

[Athullia Emmanuel]

group delay is taken as the derivative of phase delay

so i thought taking differential would give me the group delay of the signal envelope.

Answer 1

[KSSV]

Replace the line:

gd = diff (ph,(2*pi*f0));

with

gd = diff([ph (2*pi*f0)]);

[Athullia Emmanuel]

thank you , the code was completed well and stoped getting the error message.

[Walter Roberson]

I don't think this is right for the circumstances

Answer 2

[Walter Roberson]

MATLAB has two diff() functions. The one that your equation is for, the calculus derivative, is only used for symbolic expressions and symbolic functions.

The other one applies for all numeric types, and is the successive difference operator, x(2:end) - x(1:end-1) . In the case of equidistant points each 1 unit apart it acts as an approximation of numeric derivative.

As you have numeric data, I recommend that you use gradient(), which does more accurate approximation of numeric derivative.