Re: audio modulated SSTC

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: Jan Wagner <jwagner-at-cc.hut.fi> 

Hi,

On Tue, 3 Feb 2004, Tesla list wrote:
 > Original poster: herwig.roscher-at-gmx.de
 > Trying to improve quality and output power of my audio modulated
 > SSTC, I need to understand the theory of operation better. There are
 > two statements, I do not fully comprehend:
 >
 > 1. "For audio modulation, you must use push-pull mode *since audio
 > modulation requires CW operation*."
 >
 > Does "CW operation" mean, that both half waves need to have an
 > *equal* length? And, that pulse width modulation of a single ended
 > load will not lead to good-quality-audio-modulation?

I don't know where you found these statements, but they appear
to be quite wrong. Firstly, you don't need push-pull at all - a simple
half- or full-bridge is sufficient. Second, whether the load is
single end tied or not, simply does not matter, for example compare to
power amps driving speakers: single ended or cross-over.

Third, CW usually means "continuous wave" (no carrier
deviation/moedulation) e.g. radio amateur morse broadcasts, but with a
SSTC it is taken to mean "continuously powered", that is, you don't pulse
the 300VDC/whatever supply voltage (or blank out the SSTC driver) as for
example in 10ms ON then 10ms OFF, or similar (aka staccato, interrupter,
interruptor).

 > 2. " Self-resonant systems adjust the driver frequency using the
 > detected maximum output power of the resonator."

Actually they just keep the phase difference between drive signal
voltage and TC current draw at zero degrees. As a consequence, the
TC runs efficiently and at maximum power, though the control system
itself doesn't "detect" maximum power.

 > Since audio modulated
 > circuits are being power modulated, the output power of the resonator
 > and the control signal to the driver will vary according to the audio
 > being fed into the TC system. Therefore audio modulation will not
 > work with self-resonant systems."

Not really correct, either - up to date, the SSTC audio mod circuits up to
date use pulse width modulation and a fixed drive frequency, because this
is so trivial to build. The simplest self-resonant system doesn't use
PWM, but this does definitely NOT mean that audio modulation would not
work with self-resonant systems, too!

Vacuum tube driven tweeters are (second?)grid modulated (AFAIK), similar
to if you modulated the supply rails of a self-resonant SSTC. So, there
you already have one way to do the audio modulation: modulate the supply
rails. You could add a class A/B prestage, class D prestage (buck
step-down or sepic), etc.

Or, leave the rails like they are, and add a voltage dependend phase shift
(delay) to the feedback signal. Phase modulation accomplishes the same as
PWM modulation. Although with the TC load the result isn't linear enough
and will lead to audio compression, so you'll need to pre-compensate for
that or add an outer feedback loop (microphone? base current averaging?
etc)

Phase modulation is bound to be much better than PWM modulation,
efficiency and EMI/EMC wise. I've some ideas on this but haven't yet found
the time to do real experiments, got too many other projects right now,
but hopefully sooner than later...

cheers,

  - Jan

--
****************************************************
  Helsinki University of Technology
  Dept. of Electrical and Communications Engineering
  http://www.hut.fi/~jwagner/ - jwagner-at-cc.hut.fi