[sdiy] Shruthi 4PM (was re something else...)

[rburnett at richieburnett.co.uk]

I don't have the filter in question so I can't test the actual hardware. 
  However, a quick circuit simulation will show you how much the 
stop-band attenuation of the band-pass and high-pass outputs is 
compromised by innaccurate summing resistances.  You could do a Monte 
Carlo analysis if you really want to know what the worst case mix of 
tolerances would result in.

-Richie,

On 2013-11-28 07:48, Dave Manley wrote:
> rburnett at richieburnett.co.uk wrote:
>> If you take a look at Olivier's technical description here...
>> 
>> http://mutable-instruments.net/static/documentation/pole_mixing.pdf
>> 
>> ...you can see that the cascaded 4-pole filter intrinsically produces
>> the four low-pass responses.  So these responses will be of good
>> quality
>> and not subject to resistor matching tolerances.
>> 
>> It is the other responses (band-pass, high-pass and notch) that are
>> constructed from carefully chosen weighted sums of the input and the
>> four lowpass outputs.  So the purity of these "constructed responses"
>> depends on how accurate the weightings of the various signals are in
>> the
>> final mix.  This is where tight tolerance on the mixing resistors will
>> reduce "non ideal" artifacts in the constructed responses, like LF
>> bleed-through in the stop-band of a high-pass filter.
>> 
>> -Richie,
>> 
> 
> This is exactly my point.  The non-LP modes are very sensitive to
> tolerance of the summing resistors.  I'd really like to see in bench
> testing how well the other modes match their calculated response.
> 
> -Dave
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy