analogspiceman, thanks for the proper spelling. The book title is in all caps and I've always seen "spice" always in caps (aside form the "LT" version). Thanks, also, for further explaining the proper etiquette in using the yahoo group.
Sanjib, the book has several figures illustrating the concept, but otherwise, doesn't go into much more detail in that regard. However, there's a lot more general info and tips on using FFT in the other sections you noted.
Yes, using smaller delta T's will give more accurate results, but at the expense of longer simulation times.
Ken, Thanks a lot for the information regarding the importance of selecting the proper sampling window in the FFT view menu. Are those algorithms elaborated in the book from WE?
Also, selecting a delta T (simulation step size) as minimum as possible would give results closer to reality, right? (though it would take more simulation time with reducing delta T keeing stop time the same).
A word of caution on the yahoo group, while most are helpful, one of the members is known to be very aggressively disparaging towards newbies.
Hey, I resemble that remark! (Which I will address shortly.)
But first, to Ken Wyatt and anyone else reading this, the correct spelling of LTspice is as a single word with the "LT" part in upper case and the "spice" part in lower case.
As far as the Yahoo LTspice group goes, the rule of thumb is no question, no matter how simple or basic, will be mocked as long as it has been prepared with at least as much effort in the asking as is expected in the answering. If there is an exception, it would be students trying to trick experienced group members into directly answering homework problems. Such behavior is cheating (whereas just getting help with basic understanding or where specifically "stuck" is perfectly fine).
Asking questions that are clearly and directly answered in LTspice Help or in yesterday's messages is sure way to bring on a public shaming for being too lazy to do any research and for inconsiderately cluttering the group with needless drivel. Also, you wouldn't believe how some people can ask vague questions with little or no background information and either don't upload a simulation file at all or fail to include all the necessary parts to make it run: "LTspice couldn't simulate my 5 volt buck converter. It just stops with an error message. How can I get LTspice to work right?"
Stop Time (T) is the duration of the transient simulation. For a fixed Maximum Time Step of, say, 0.01 ms, the value of T has a significant influence on the appearance of the FFT spectrum as well as on the precision of the harmonics minimum amplitude.
The Time Step parameter (delta T) is the interval between two transient measurements. If the value Stop Time remains fixed and equal to T =10 sec and if the number of samples is decreased by not being as demanding with the maximum value of Time Step, the frequency resolution is reduced and provides fewer details in the low frequencies.
Therefore, you have to consider both parameters for the FFT analysis to be close to reality. Finally, there are 12 models of sampling window within LTspice. For example, the Flat Top window will provide very precise harmonic amplitudes, but very bad frequency resolution, while the Hamming window will provide acceptable amplitude accuracy and very good frequency resolution. The Kaiser-Bessel window will provide more precise amplitudes and resolution, but is still limited for both. The others have similar kinds of tradeoffs. No window is perfect.
To my knowledge, there is no electronic copy available for this book.