In article , Don Pearce
wrote:
On Tue, 22 May 2007 16:03:55 +0100, Jim Lesurf
wrote:


Well, I've used both 'traditional' THD+N kit that works by nulling or
filtering the test sinusoid, and a specan/generator that autofinds the
harmonics and works out THD. The second didn't seem at all 'tedious' to
use as it automated the process. It was also faster than the old
distortion kits I used to use that took some seconds to null down for
low distortion.


There is a distinct advantage to the second, in that it gives you
information about the nature of the harmonics. If you are making this
measurement as part of a development process, that can be invaluable.

When using the ST1000A I used to find it useful to view the result of the
nulling on a scope. i.e. the waveform with the test sinusoid removed. This
gave similar info provided you could 'interpret' the results. For example,
very useful when adjusting the bias of an AB output as you could see any
crossover error appear or vanish as you twiddled the bias.

This has an advantage over power/frequency plots of the harmonics as it
shows the effects of the relative phases as well. If I still had an ST1000A
I'd probably be using it, but I lost access to it 20+ years ago...

FWIW A month or two ago I wrote a simple example program that does this for
a computing magazine. Found it quite interesting to view the nulled
patterns from an old Shure cart playing a test LP.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html