PLAN B analog blog

It's interesting, this is the second blog which I've seen this
comparison being drawn.

The similarities to the Buchla 291 go as far as it's vactrol based
bandpass filter, with one of them used as an AGC (Automatic Gain
Control) on the front end.  The circuit I used for that however is
from the Perkin-Elmer Vactrol Application Note #1, not the 291.  

The Model 12 also has an internal AGC, but it's manual, not voltage
controlled.  In that bandwidth was not voltage controlled on the 12
there was no need.  On the Model 12 it's done with a dual pot, one
half controlling the resonance level, the other half controlling the
input gain slightly - much like a fine tune on a VCO. It augments the
gain set by the faceplate's input level pot.  I use a dual pot on
almost every module I make, but for a very different reason: to keep
the main PCBA rigid!  The Model 12 is the only one in which both
halves of the dual pot are active in the circuit.

Back to the AGC for a second ... asking the question what does
variable bandwidth actually do:  it restricts the width of the band
allowed to pass to the outputs obviously and in doing that lets less
frequencies through.  So as the band narrows so does the output level
of the filter.  A circuit is required to boost the output amplitude of
the filter slightly as the band closes down in width - they work
together. This is called an Automatic Gain Control and it's not just a
Plan B thing - these are commonly found in filters. On the ET Bandpass
that gain control is done a bit more efficiently than on the 12.  You
won't find yourself having to make adjustments to the level pot when
you change the resonance or bandwidth las you do on the 12.  On the 11
it's much more set and forget.  I didn't feel it was nessesary as
neither of these functions are voltage controlled.  On the 11 it's
another story - to assure the output gain would stay level as the
Bandwidth was VC'd, I had to come up with a way to adjust the level
accordingly.  The ACG on the 11 is basically a narrow-range vactrol
VCA - a voltage controlled pot - that's governed not to shut down or
open up all the way.  I use the very same circuit for the VC slewing
in the Heisenberg's SMOOTH output.

I went with vac's on the 11 for several reasons, yet did make a
prototype using transconductance amplifiers in place of vac's - four
of them.  I elected to go back to my first choice for a couple of
reasons.  One is cost - although vactrols are bit pricey, there are
much less components required in the drive circuit as on the 13700. 
There is also the cost of PCB real estate one must take into
consideration.  An added benefit of using vacs above the cost
reduction are their sonic fingerprint.  Although these aren't the same
type oused in the Model 13 (not quite as slow), they're still much
slower than trans amps, and I tend to like the whip effect that comes
from that.  It' doesn't 'ring' like the Model 13... whippie is a
better term. Like the tail end of snake, it sort of follows what's
going on in front. Also, the effect gleaned from the TWIN switch is
pretty much nullified using transconductance amps. Each of the two
feedback paths are gated by their own vactrol which are intentionally
running a bit behind one another.  This is what causes the thickness
produced when the switch is tripped.

A recap on the facpelate controls - I actually find the Model 11's
compliment much more similar to the Model 12's than the 291, outside
of the Feedback switch and the voltage controlled Bandwidth.  As in
almost all of the our modules, we used bipolar attenuators.  As with
the Model 12, the 11 as one bipolar control for Freq, and one that's
not.  Both have outputs for the 6dB and 12 dB taps, although handled
differently on each.  Level pots not found on the 291, the feedback
switch not found anywhere, the two tap outputs ot found on the Buchla
filter, etc.

I may have gone into too much detail here, hope you guys find a least
some of it interesting!

- P

--- In PLAN_B_analog_blog@yahoogroups.com, "Romeo Fahl"
<romeo.fahl@...> wrote:
>
> That's thoroughly wicked!
> 
> Say Peter, does this filter have any relation to the Buchla 291? I
see some
> similarities, at least based on the front panel content.
> 
> Romeo
> 
> On 8/29/07, (i think you can figure that out) <peter@...>
> wrote:
> >
> >   The Model 11 Bandpass Filter has been officially released. Shipments
> > to begin in two weeks. Go here for full module stats, specs and sound
> > samples:
> >
> > http://www.ear-group.net/model_11.html
> >
> >  
> >
>