do you mean you're missing the first spectrum because you put the figure after the first instance of subplot? you should move figure(3) before the for loop. with this you create the subplot in whatever is the active figure/axes for the first iteration, jump to figure(3) to then plot the first iteration fft. however with that you overwrite that first iteration with the subsequent subplots.
if you're commenting that there is no frequency response maybe taking the log10(eeft(1:n)) would help you see the smaller frequency responses.