[sdiy] "Ghost" Electronics

[Grant Richter]

For practical purposes a 2 millisecond step change is all that is required.
This would mean a band width from 1K on down. A fast envelope attack would
need to be encoded. A carrier frequency of 5K should do the job, and the
sidebands are the data of course.

> From: Martin Czech <czech at Micronas.Com>
> Reply-To: Martin Czech <czech at Micronas.Com>
> Date: Thu, 22 Mar 2001 09:36:08 +0100 (MET)
> To: harrybissell at prodigy.net
> Cc: synth-diy at node12b53.a2000.nl
> Subject: Re: [sdiy] "Ghost" Electronics
> 
> 
> :::I would not use PLL techniques. If you could square up the recovered
> :::audio, use the PV-1 pitch to voltage converter... Penalty of divide by 2 in
> :::frequency... but it will track very accurately and with the minimum of
> :::ripple. It "could" staircase at low frequencies, but you could filter that
> :::with only 10% of what a loop filter would need to be. TomG still has some
> :::circuit boards for the unit. I never thought
> :::of using it like this. Great Idea.
> 
> Great idea. If the voltage range of interest is coded as a 5000-15000 Hz
> sine wave (linear tri or sine wave osc. needed), why
> shouldn't that work with nice ripple rejection?
> I think the vco frequency must be way higher then that of the signal to
> be recorded, less sidebands, less ripple, like in fm radio.
> 
> Am I missing something?
> m.c.
> 
> 
>