[sdiy] "Timer Controlled VCO" (TCVCO) ?

[Michael Zacherl]

On 24.Sep 2014, at 12:40 , Neil Johnson <neil.johnson71 at gmail.com> wrote:

> Michael Zacherl wrote:
> 
>> I'm going through an awfully gibberish thread and found this post (without
>> it stating the source of information):
>> https://www.gearslutz.com/board/6535950-post313.html
>> 
>> "It uses VCOs. The output frequency of the VCOs is measured digitally. The
>> frequency data is used to adjust the control voltage of the VCOs. This is
>> opposed to the design of DCOs: DCOs use digital timers to trigger the
>> oscillator cycle directly while Schmidt uses a pure VCO path which is
>> digitally observed. This is similar to the well known autotune procedure of
>> classic synths - but the autotune is able to adjust the VCOs all the time(!)
>> and does not need to be operated manually from time to time."
>> 
>> So you're guess is probably right.
>> Moreover all this modulation options hardly would be possible with DCO-ish
>> designs, I'd say.
> 
> For that to work the autotune system needs to know when the VCO pitch
> is being modulated.

yes.

>  Assuming those are also analogue I guess it must
> carefully measure the pitch CV before, during, and after the frequency
> measurement (even if doing a single-cycle reciprocal measurement) to
> make sure it is not compensating for the wrong thing.

given, that the VCOs won't be terribly drifting (which I'd almost take for granted with a modern design),
it could be enough to do the f-checks when the synth is silent.
Then again, this would require some detection logic and a mute-circuit for the fm inputs.
And this needs to be fast enough not to interfere with the next sounds played.

> The classic autotune that the above text mentions takes over control
> of the VCO so there are no artificial errors during the frequency
> measurement.


however one puts it, it seems to be tricky. ;)

m.





--
http://mz.klingt.org