Don, once a harmonic is several dB below the predominant harmonics, it basically disappears in the totaling to get THD. So usually, you only have to consider the one or two predominant ones, unless you really care about the difference between 0.00200 and 0.00201% (which will be well within your error bars in any case). For example, let's say you have a system with 0.01% second, 0.005% third, and 0.001% fourth at 1V out (to make the math easy). THD will be the RMS sum of the harmonics divided by the input voltage (that's why I chose 1V!). In this case, sqrt(0.010V^2 + 0.005V^2 + 0.001V^2) = 0.011%. So if we ONLY counted the second and neglected the third, we're pretty damn close; even a harmonic 6 dB down from the predominant one barely moves the needle. If we count second and third, then neglected 4th, we'd have sqrt(0.01V^2 + 0.005V^2) = 0.011% (in other words, the contribution of that 4th is entirely negligible).
Like I said, this is a trivial thing to do with instrumentation that was exotic in 1979 but totally commonplace now.
I am going to run some additional tests today and will be back with more.
