[sdiy] SSM2164 Phaser - another way?

[Donald Tillman]

> On 2014-02-18 15:29, Tom Wiltshire wrote:
>> I know that phasers with the SSM2164 aren't anything new, but I
>> wondered if anyone had tried this:
>> 	http://www.electricdruid.net/images/SSM2164Phaser.png
>> It looks ok in the sim. It breaks the usual rules by taking the output
>> of the SSM2164 to a genuine ground rather than a virtual ground, but I
>> can't see why that should be a problem - is it?

On Feb 18, 2014, at 7:48 AM, rburnett at richieburnett.co.uk wrote:

> I don't think this will work.  The output current from the VCA is just dumped to ground, so it won't have any influence on the allpass filter behaviour.  I think you are assuming that the input to the VCA will behave like a variable resistance?  I don't think it works like that.  The input to the VCA will just appear like a fixed resistance, and it is the output current of the VCA that will be scaled by the CV so that it *appears* like there's variable resistor inside.  But, this will only control the cutoff frequency of the allpass filter if the VCA is actually included inside the allpass filter circuit.



1. I think this circuit will work, it just won't work well, as the range will be limited and it will clip weirdly.

2. Where did this circuit come from?  Any background on it?  The Electric Druid image suggests it's been simulated.  And "It looks ok in the sim" suggests that it works.  What does "ok in the sim" mean exactly?

3. Even though it might look like the SSM2164 is acting like a variable resistance to ground in the diagram, that's not what's happening.  To understand how it works you need to look at the simplified schematic on the SSM2164 data sheet and note that the input impedance of the 2164 changes with the control voltage.  

4. The way the 2164 operates is that it's a fixed gain transconductance amp that steers its output current between two directions: to the output for high gains and feeding back to the input for low gains.  Advantages over a traditional OTA are that the low gain settings are linearized and the 0 V setting is balanced to unity gain.

5. Tom, that Spice "model" is way too abstracted and simplified to work in this application.

  -- Don

--
Don Tillman
Palo Alto, California
don at till.com
http://www.till.com
650 888-9632