another BFO idea (was:RE: Soft sync and medicine to cure it.)

[René Schmitz]

Hi Juergen, Martin and List!

I've toyed with two (albeit ordinary) vcos beating in the audio range,
before I built my base-band tzvco. 
(2 VCOs into RM -> LPF, plus another modulation VCO) 
The problem was drift and tracking. Re-hearing the recording I made back
then my impression is that it sounds somewhat better than the "base-band"
LTZVCOs, but I think that is due to the heavy filtering that was necessary
to remove the upper sideband. 

I suspect that with a BFO you trade one problem for another. Namely the
base-band circuits discontinuities due to offsets versus another
discontinuity, near the point where the two VCOs of the BFO lock into each
other. (They would have a dead-zone of a few Hz when they lock. That is
equivalent to a crossover distortion for the modulation signal.) The
question is only which of the two is less annoying.

Bye,
 René

At 13:46 18.09.00 +0200, Haible Juergen wrote:
>	>A talked to HAM people the other day. They spend a lot of thought
>	>for their local oscillator frequency planing, because avoiding
>	>locking of two neareby tuned VFOs is virtually impossible.
>
>Now that discarding the BFO concept in frequency shifters looks more
>reasonable than ever (IMO), there is another application that beckons
>just for the opposite: Ordinary audio VCOs with linear thru zero
>capability (LTZVCOs)
>
>Background: There are various LTZVCO circuits published (including my
>own at http://www.synthfool.com/diy/hj2vco.gif - *not* recommended for
>building)
>which to my knowledge all suffer from "commutating inaccuracies" when
>the modulation just approaches the zero point. The direction switch might be
>missed or delayed, and you get a unpleasant artefact. (It's there in my
>circuit
>for sure, and I remember reading similar findings about the Electronotes
>LTZVCOs.)
>Now, one should get rid of these problems easily with a BFO. And the locking
>around 0Hz would not be a problem, because the 0Hz point would normally
>not be used permanently.
>
>In short:
>
>FS should be able to approach 0Hz slowly / permanently
>-> locking is critical -> true TZVCO better than BFO
>
>Synth VCO will approach 0Hz dynamically
>-> delays and accuracy are critical -> BFO better than TZVCO
>
>
>Anybody built a BFO for use as audio range synth VCO yet ?
>The critical point would be fast switching in the VCO core,
>i.e. high frequency tracking I guess ...
>
>Anybody knows if the Bode FS's BFO is accurate enough to play
>tuned melodies ?
>
>JH.
>
>
-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159