AW: Re: REQ: Phase Shifter Schematics

[Magnus Danielson]

> 
> > of the ElectroHarmonics SmallStone Phaser which is a quite nice guitar ph=
> > aser.
> > It should be a quite simple task to make this thing CV controlable.
> >
> > There is 4 1-pole allpass filters which are built around the CA3094 chip.=
> >
> > Note that the control input of the CA3094's is a current inputs. The data=
> 
> These are surely intended to be current inputs, but in that special 
> application
> they are connected to the buffer output of another 3094 over a 100 Ohm
> resistor. So I guess that they actually *voltage*-control the phase shifter
> stages, taking advantage of the exponential relation between V and I
> on a current mirror.
> Just a guess - not sure about this at all, but it would make sense.
> (And there's a question remaining: The inputs are all tied to the same
> control node in parallel, but each OTA is on a different chip, so
> the current distribution should be interesting!)

Well, to start of with is there no true current mirror, it is the input to an
NPN resistor base. The Emitter of the transistor is hooked to the negative
rail and the collector is connected to the emitters of the input transistor
pair, just as you migth expect. The 100 Ohm resistors will just act as
current dividers and also reduce the peak current somewhat. I think they just
add to tracking rather then being a major characteristic setter.

BTW. I _know_ that the current will change the "frequency" linearly. I have 
done
the math exercise on the full CA3094 model with surrounding components just as
they appear in the SmallStone schematics. The sweeps that the SmallStone does
over the frequency range is not linear but exponential (or close). This could
be achievied in the sweep generator but not sure rigth on hand... but that one
sure looks like triangle to me... (no scope around to check waveform on my
SmallStone).

Cheers,
Magnus