Moog 24dB Highpass

[Martin Czech]

I found the moog 24dB highpass patent on Mikko Helin's
page. So far, so good.

[Synth diy page (moog pat.)] >> http://www.uta.fi/~helin/index-diy.html
[Patents (Moog etc)] >> http://www.uta.fi/~helin/index-pat.html

But this is only a "priciple of operation" drawing.
Has anybody a real schematic ?
Has anybody tryed to rebuild one, yet ?

1st order analysis:

The circuit is not differential and has not much to do with the famous
ladder, think of it as a normal 1st order highpass cascade, each
element decoupled ((darlington) emitter follower ? or unity gain 741 ?)
and substitute the resistor (to ground) of each element with two
transistors : one npn trany to some fixed positive potential and the
other (I guess pnp, the arrow is missing) to some fixed negative
potential. That's 8 transitors, they all have to be matched with the
usual method, but better would be to match at two points with some
small current ~100nA and a bigger one ~100uA, because of the circuit
(see below) . The control voltages are summed with  a noninverting
buffer and go from there to the 4 npn bases.  The sum is also inverted
and controlls the 4 pnp bases.  Since base current is exponentionally
depended on base emitter voltage this already makes the exponential
cutoff controll. No expo converter circuitry is needed, but this will
also mean a quite nonideal behaviour (temperature drift etc, and
saturation current will have much influence. Thus better matching may
be required.).  Base current controlls emitter-collector current
(Ic~b*Ib, but this is only valid if Ic<<1mA).  As previously discussed
in some "ladder" thread, the transconductance of the trannys is
proportional to Ic, voila exponential controlled "resistance" and some
capacitors make the voltage controlled filter.  The patent does not
show any feedback, I guess this can be a simple voltage divider pot at
the output that is fed back to the input.  I heard people say that the
sound is quite "dirty" and oscillation is possible, but non sine shaped,
much more partials. Ok, a highpass filter will always oscillate with
added distortion, since the higher partials of amplitude saturation are
not cut off (by definition). The particular "non-ladderish" design will
saturate faster than a differential approach, that is more distortion
at the same signal level.


comments invited.

By the way : merry christmas and a happy new year to all synth-diyers,
may your soldering iron never get cold in 1998!

m.c.

------------------------------------------------------------------------
m.c. has made it finally:  3 CDs out now; 72 min. minimum; "1"
(1994-1995),"2" (95-96),"three" (96-97); experimental stuff; mostly
Eimert/Stockhausen style; but also modern popular style