Thread Subject: Question on inverse filter

Dear all, I want to design an inverse filter, I inverse the spectrum of a chirp and do ifft to get the coefficients.

However, the convolution of the coefficients with the chirp is not close to delta function, anyone can help me? Thank you very much!


f0=10e6;
B =4e6;
Tr =6.4e-6;
a = 2*pi*B/Tr;
fs =200e6;
t = [0:1/fs:Tr];
chirps = cos(2*pi*(f0-B/2)*t+a/2*t.^2);

fftc = fft(chirps);
inversef = ifft(1./fftc)
figure;
plot(conv(chirps,inversef));