implementing direct form 2 without using filter command

x = randn(200,1);
for n = 9:length(x)
    y(n) = 9*w(n)+w(n-1)+w(n-2)+33*w(n-3)+12*w(n-4)
    w(n) = 2*w(n-1)-1.5*w(n-2)+0.9*w(n-3)-0.96*w(n-4)+x(n)
end

The above are the equations for direct form2 realisation of a transfer function.But it is throwing an error for variable w(n).what is the value of w(n) so that the error is removed.

[Jan]

I've formatted your code to make it readable. Please use the "{} Code" button.

Please post the error message, when you mention, that there is one. Letting us guess the error at first wastes time.

Answer 1

[Jan]

Have you seen a working version of a direct form II implementation: http://www.mathworks.com/matlabcentral/answers/9900#answer_13623

[s]

I have seen the direct form 2 implementation you have provided which uses "function [Y, z] = myFilter(b, a, X, z)" i am not allowed to use the above filter command.Is it possible to get the output just by using the difference equations i have provided. Thanks

[Walter Roberson]

"myFilter" is an implementation of direct form 2 that Jan has written and does not use the MATLAB "filter" command. You could name "myFilter" anything that is a valid MATLAB identifier, including

function [Y,z] = Supercalifragilisticexpialidocious(b, a, X, z)

[s]

Thank you very much for the clarification...