For an FFT performed with the 'Amplitude' setting selected, the y scale of the FFT plot gives the relative strength of each individual frequency present in the original signal. Since the strengths are relative to the original signal, the scale is essentially unit-less. For example, if you perform an FFT on a dataset such as 1*sin(x)+2*sin(2*x), the FFT will have two spikes, one with an amplitude of 1 and the other with an amplitude of 2. These two amplitudes represent the strength of the two harmonic components in the original signal (as can be seen by looking at the scalar multipliers, 1 and 2, applied to each sine wave in the original signal equation).
When performing an FFT with the 'Power' setting selected, the relative strengths of each individual frequency present in the original signal are squared and normalized before being represented in the plot. In our example above, the ratio between the two individual frequencies would therefore be 4 (as opposed to 2 for the Amplitude setting).
As for the difference between our results and your results (which I can only assume you calculated manually), it is too difficult to determine that without having more information.
Sincerely
Ryan Toomey
[This message has been edited by rtoomey (edited 07-28-2000).]