[sdiy] Wanna help me make up a VCF homework? ;)

[René Schmitz]

Hi Aaron,

Aaron Lanterman wrote:
> I'm making up a problem now where I'm going to have them look at an 
> OTA-based 1-pole filter cell (in the usual four pole setup) and tell me 
> what the cutoff is as a function of the current to the pin on the VCA, 
> along with analyze the divide-down ladder, etc.
> 
> I've gotten three following so far:
> 
> 1) SSM2040 datasheet design
> 2) Ray Wilson's 4-pole - Ray's already showed up in HW #2 ;)
> 3) Polyfusion Lowpass VCF
> 
> Any others come to mind?
> 
> In particular, all the designs I've found drop a cap straight to ground 
> and then put in a buffer. I've yet to find one that actually follows the 
> one of the ARP patents, where coming out of the OTA the current goes 
> into an integrator formed with a cap in the feedback loop - I showed 
> that in class so it would be nice to find an actual example.

Two ideas:

1) Wasp filter. (Jürgens version is well drawn.) But you probably need 
to explain how CMOS gates can be viewed as sort of opamps. (look at the 
transfer function, linear range arround 1/2Vb, opamp rules)

The Norton based ARP filters (4072, 4075) are actually also of that 
type. The OTA is formed by a PNP pair, working onto a current mirror, 
which happens to be built into the LM3900. (Between the + and - inputs.)
If you draw that together using the equivalent schematic shown in the 
3900 datasheet (maybe redraw it a little so the current mirror shows 
more clearly) one can see how this can be analysed as a OTA, working 
into a high gain inverting amplifier (so some golden rules apply).

> P.S. I've only made vague references to the class about this list, since 
> I don't want to give away my secret weapon yet. When it comes time for 
> them to start their own modules designs, I'm going to require them to 
> join the list. ;)

I would guess they googled this up already. :-P

Cheers,
  René

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159