[sdiy] DC-1 MHz, through-zero FM, controlled, oscillator without core capacitor

[Richie Burnett]

> Interesting! I had no idea that was how a Theremin worked. Those can
keep the pitch really well, actually.

Perhaps if you are a player skilled in the art! ;-))

> A PLL would mean completely digital control, wouldn't it?

For the coarse pitch control in steps, yes.

> Do PLLs like that respond well to audio rate modulation?

There's two options...

1. You apply frequency modulation to the PLL's reference, and you make the 
PLL's response really fast so that its control loop acts to make the VCO 
track the instantaneous modulation on the reference.

_OR_

2. You mix the frequency modulation signal directly into the VCO's CV input 
and you make the PLL's response *really* slow, so that the feedback loop of 
the PLL doesn't try to remove the modulation by correcting for what it 
perceives as an unexpected error!

Both techniques have their own set of drawbacks though.  I've seen the 
second technique used in FM broadcast transmitters that claim digital PLL 
stability in the long-term for the carrier frequency, but still allow 
instantaneous analogue modulation of the frequency to impart the audio.  The 
downside is that the PLL usually takes *ages* to lock after a step change in 
frequency because the control loop has to be slow enough not to remove all 
of the bass from the audio!

> The idea with a second core was to let the oscillator drift a little
bit owing to its analog nature, while also correcting for the gross
drift. Do you think that would keep things close enough in tune to be
musically useful?

I don't know.

I'd be inclined to have one of the RF oscillators be a crystal oscillator. 
Then you only have the stability of one RF oscillator to worry about and/or 
compensate for in your system.

But then my preference would be for DDS rather than PLL anyway, because you 
get much more agile frequency control, in very fine steps (milli-hertz,) 
with none of the control-loop settling issues of a PLL, yet still keeping 
the crystal oscillator accuracy and long-term frequency stability.

-Richie, 


-- 
This email has been checked for viruses by AVG antivirus software.
www.avg.com