View Single Post
  #6   Report Post  
Posted to rec.audio.tech,sci.electronics.design,sci.electronics.repair
 
Posts: n/a
Default THD claims of audio signal generators


mc wrote:
My ancient Heathkit audio signal generator, which uses 3 tubes
and a light bulb in an oscillator circuit, claims to produce sine
waves with a total harmonic distortion less than 0.1%.
My semi-ancient Hewlett-Packard HP202B oscillator only claims
to produce THD less than 1%.
I would have expected the HP instrument to have considerably
better performance specifications because of its much more
elaborate circuit.
Are we simply looking at pessimistic vs. optimistic ratings?


if I were buying a low-distortion oscillator, neither would
be amongst my consideration.

What you're seeing is that a THD spec, all by itself, does
not the performance of an oscillator describe.

Compare the other specifications of the two oscillators,
e.g.:

* Response flatness: I would expect the HP to be
substantially better in spec and actuality

* Settling time after aburpt frequency change, I would
expect the HP to be FAR superior.

* Short-term amplitude stability vs time: I would, again,
expect the HP to be much better

* Long-term amplitude stability vs time: ditto for the HP.

* Phase noise/jitter: guess what? I'd expect the HP to be
much better.

If you need to build a simple, cheap, easy-for-someone-
of-unknown-skill-to-build oscillator, you make one set of
design choices. If you need to build a rugged, highly
stable, dependable unit which will has better short- and
long-term stability, can be swept rapidly in frequency while
maintaining that stability, and so on, you make a different
set of choices. Both choices have costs and consequences.

If I wanted cheap (at original prices only), I'd consider
the Heath, but I'd have to be willing to live with the fact
that I can't depend upon it as a source of stable,
dependable sine waves. If I needed a lab instrument that I
could depend upon to not have the amplitude bounce
substantially when I swept frequency, that I could
implicitly trust for amplitude flatness over the full range,
that would maintain its calibration for a long time, I
wouldn't buy the Heath.

But if I wanted dependably low distortion, I'd have
something else entirely.