Bassballs filter architecture

[lepra85]

Hi all,
I am trying to fully understand the bassballs schematic.
There are some posts related with the envelope follower section, but i didnt find too much info related with the sweeping filter section.
I looked up and seems that this filter architecture is not one of the most kwown ones (state variable, multiple feedback, sallen key...)
Does anybody knows the bassballs filter architecture?
Does anybody knows the ecuations for the Q and the center frecuency?

Thanks for reading.

[Kipper4]

A schematic when asking such a question is helpful my friend.

http://guitarpedalbuilders.blogspot.co.uk/2013/01/ehx-bassballs-schematic.html

My answer
Dual multiple feedback filters //
Google it and the equation.
Try GeoFx articles on wah too
In a rush, time for work

[Transmogrifox]

It's a bridged-T filter architecture, but the only difference between this and the multiple feedback is where you inject the signal, so Kipper is right as far as I'm concerned.  The basic equations for the MFB filter center frequency and Q will be identical since the resonator is the same. 

Because of the point of signal injection  BassBalls will have a peaking low-pass response instead of the typical band-pass response you would expect from this permutation of MFB.


The transistors act as variable resistors. 

[PRR]

Yup. Replace Q2 with a (variable) resistor, you see the twin-Tee. That's in a feedback leg so you get the inverse response (I guess a bump instead of a notch).

The Q of a plain naked twin-Tee is spoiled by interactions. Here both ends see "zero" impedances. I'm not sure that's commonly found in filter-books, though the math should be straightforward.

I am sure you can not get control of both F and Q with one variable part.

If you plan to use a BJT Q2 to shift frequency, this will only work for "very small" signals at Q2 collector. Maybe 20mV-30mV? The signal at this node may be only the Highs, so it may be naturally small. Tee-network value ratios also set the relative level of Q2 signal to external signal level. This may also limit the useful range of F and Q.

[anotherjim]

Here a variant..

[anotherjim]

And I wanted to add a Bassballs scheme but got called to dinner, here it is...


A question I have is - what are those 560p caps C7 & 11 doing? On the op-amp outputs? Why?

[Transmogrifox]

A question I have is - what are those 560p caps C7 & 11 doing? On the op-amp outputs? Why?

I don't think I have an answer but I have some guesses.

One guess is these are added to do something to the behavior of the filter when the op amps go to the rails...maybe softens the op amp clipping?

Another possibility may be related to op amp stability.  Typically capacitive loading pushes an op amp more toward an unstable condition, but then maybe it compensates something when there is a 2nd order RC network wrapped around it.

Maybe it's intentionally used to dominate op amp internal frequency response for proper (repeatable) phase alignment on each filter so that the two filters sum together properly.

Maybe the answer is (D):  None of these
or.
(E):  The drafter was on crack and put components on the schematic where he saw them in a vision.  The original engineer got promoted before getting to review the schematic and the replacement engineer assumed the original engineer knew what he was doing so he went with it, even though he didn't know why these were added.