The only reason I got an array of clusters was because I fed the PSD function with an array of waveform data. The reason I did this was because I believe you stated before that you had the same thing (I may be wrong). You could just get rid of the FOR loop and it should work, writing to file.
Laurence C. Applications Engineer National Instruments
I just realised when I compared the saved graph in the TDMS file to the displayed graph from the front panel. The frequency is scaled some way. The saved frequency spectrum is showing the same shape but at a higher value. The one on the right is the data saved in the TDMS file, the other one is plotted directly from the PSD vi.