Matrix

[Michael Zacherl]

On Aug 3, 2008, at 5:36 PM, Paul Maddox wrote:

> Michael,
>
>> The Xpander's 16 types are a bit of a compromise due to restrictions
>> of the CEM3372 used in the Xpander (I love this synth, BTW!).
>> But that was the starting point ...
>
> It's not really a restricton of the filter (which I thought was an  
> SSM2040).

It is - no direct input for the filter itself. Just the two VCAs for  
the signal (mostly used for the VCOs)
So for HP they have to turn the 1st stage into a simple buffer by  
tuning it very high (as a LP) and use its output as input for the next  
three poles.
That's the reason why there's no 4-pole HP and 5 rather strange modes  
called xxx + 1 Pole-LP (which is when the first filter cell is not all  
open.
Nicely solved by this guys back then, IMHO!

> The idea is a simple one, take each of the four poles and put the  
> into a
> VC-Mixer, with the option to invert them. For a fifth input include  
> the
> original signal, again with mix and invert options.
>
> All you need is a 5 input VC-Mixer with invert option for each  
> input. A
> simple micro to remember and recall the settings you choose.
>
>> This thing should be much more generic, not only used for filter(s),
>> but could be part of sort of an uber-filter, if someone wishes that.
>
> Huh?
> "not for a filter.... Sort of uber-filter"
> Either it's a filter or it's not, or am I missing something?

Nah ... I didn't re-read before sending obviously. :-(
There was some progression in this other discussion I didn't speak of.
 From VC-mixer to VC-matrix.
Once this matrix was designed it could be part of an über-filter (more  
poles, feedbacks, groups for two or more independent filters in one,  
whatever) or something stand-alone or part of another unit like acting  
as modulation matrix, etc. (just thought "preset aid")  ... whatever.
It's about the principle: Lots of inverting VCAs sitting on the cross- 
points while being  µC-controlled.
Just having visions ...

>> So that it could be used as a stand-alone module/unit or as part of
>> something else.
>> Basically a large µC-controlled network of VCAs.
>
> Simple enough.
> SSM2164 should suffice.

Thanks, looks resonable. Is that a current part?

At the moment I rather fancy some rotary switches for something  
really, really simple. No visions.  ;-)


    Michael.
--
feed your perception: http://blauwurf.at