[sdiy] SSM2164 Phaser - another way?

Mattias Rickardsson mr at analogue.org
Wed Feb 19 11:17:43 CET 2014


On 18 February 2014 17:05, Tom Wiltshire <tom at electricdruid.net> wrote:
>
> I'm using a simple LTspice model that Olivier Gillet posted in one of the Mutable Instruments forums:
>
> * SSM2164 VCA cell
> .SUBCKT SSM2164 IN OUT CV
> * The two transistors in the input's differential pair have the same base
> * current, hence the same Vbe. The second transistor is referenced to
> * ground so IN is virtually grounded.
> V1 IN 0 0
> * Current gain expression
> G1 OUT 0 VALUE={-I(V1)*PWR(10, -V(CV) * 1.5)}
> .ENDS
>
> My thanks to Olivier for helping me out with that. It doesn't model the 5K impedance of the CV input, but it works well enough for everything else - and I've tried quite a lot of things with it.

I use an extended version if that model, including the CV impedance etc:


* SSM2164 VCA cell
*
* taken from Olivier Gillet's msg to synth-diy discussion list 2011-01-29
*
.SUBCKT SSM2164 IN OUT CV
* The two transistors in the input's differential pair have the same base
* current, hence the same Vbe. The second transistor is referenced to
* ground so IN is virtually grounded.
V1 IN 0 0
* Models 5k impedance for the CV input
R2 CV 0 5k
* Stability network
R3 IN 1 560
C1 1 0 560p
* Current gain expression
G1 OUT 0 VALUE={-I(V1)*PWR(10, -V(CV) * 1.5)}
.ENDS


The signal input of the 2164 is essentially a virtual ground, so your
30k+2164  becomes essentially a constantly grounded 30k.

It would be awfully nice if the 2164 behaved like a voltage-controlled
resistor, but it doesn't.
When connecting a 2164 according to the datasheet, to the
virtual-ground summing node of an OP-amp, the 2164 *output* will look
like a voltage-controlled resistor *from the OP-amp's point of view*.
But the 2164 *input* looks like the virtual ground summing node, so
the resistor you put there is the only resistor drawing any current,
and it's constant.

Speaking of the SSM2164 datasheet, did you notice the error in Figure
6? It's supposed to show THD+N vs. amplitude - which would be a very
useful graph! - but it has a 20-20k frequency scale on the X axis, so
it's impossible to know what amplitudes the curves show. Funnily
enough, Coolaudio just copied it to their V2164 datasheet... :-)

/mr


More information about the Synth-diy mailing list