Hello John,
You are getting these results because of your very short
rise and fall times and short 'crossover' pulses.
This was indeed the trick to get a "useful" result for THD
in the .FOUR report. I have had used 10us rise and fall time
in my example which I uploaded yesterday.
Best regards,
Helmut
--- In LTspice@..., John Woodgate <jmw@...> wrote:
In message <kkfbb8+osfi@...>, dated Sun, 14 Apr 2013, Echidna
<mchambin@...> writes:
Indeed, I was tricked by the .FOUR function default calculation on 9
harmonics only. I tried with 99 harmonics and got a much different
result.
You are getting these results because of your very short rise and fall
times and short 'crossover' pulses. If you do an FFT from 'View' you can
see that the harmonic spectrum goes to well over 30 MHz. This isn't
realistic for an audio amplifier, if that is what you are working on.
But calculating THD by 'adding' (r.s.s.-wise) the harmonic amplitudes is
seriously error-prone in my experience. It's much better to notch out
the fundamental with a filter (simulated notch filters always work well,
unlike real ones) and measure the r.m.s. value of what is left.
--
OOO - Own Opinions Only. See www.jmwa.demon.co.uk
They took me to a specialist burns unit - and made me learn 'To a haggis'.
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
