[Re: [Re: multistage rate level env generators]]

[M.Ruberto]

Grant,

last night i tested my idea for the window comparator. the test circuit was
very simple. i used a 1458 dual op-amp with both outputs ORed together with
diodes. amp1's - input was connected to the wiper of a pot connected between
the 12v and ground rails. amp2's + input was connected through a diode to
amp1's - input. amp1's + input was connected to amp2's - input which was
connected to the ouput of an env generator module. in this setup the
comparator's output is active low. the window was .25 volts. as i triggered
the envelope the comparator changed states on both upward and downward
transients with nice short pulses. for my design this is perfectly acceptable
behavior.

tonight i will add a differential as per your idea and see how well this works
and compare it to the results from last night to see if there is a real
advantage.

as for adding sustain to a stage i think it will require some logic and a
controlled switch between the output of the multiplexer and the charging cap.
not so bad parts wise. i believe the use of a 4 bit magnatude comparator will
be necessary to provide a trigger used to stop the process from looping over
and over. of course a switch can be thrown in to put it in looping mode
thereby giving a very complex LFO. i still need to devise a scheme to start
the envelope upon recieving a gate. i am going to order the necassary parts to
create a prototype tonight since i am getting quite optimistic about this
thing actually working. 

i'll let you know how the tests go.

-<mike>-

"Grant Richter" <grichter at execpc.com> wrote:
> My instinctive reaction is that the window comparator without the 
> differential amp would only work for ramps moving in one direction.
> If the thresholds of the window comparator were set to say +/- 20 mv then
> the window comparator would work for both positive or negative going ramps.
> I could be wrong.
> 
> Another advantage is the ability to easily adjust the "closeness" of
> approach before advancing. There will be an optimum level which would be
set
> by the intrinsic noise in the system. Too high and the ramps will be cut
> short, too low and the comparators will "stutter" because of noise, giving
> you more than a one stage advance. This might be a front panel control,
> since cutting short the slopes will speed up the whole envelope and might
be
> musically interesting as a control axis?
> 
> Please let me know of your progress.
> 
> ----------
> >From: "M.Ruberto" <indridcold at usa.net>
> >To: grichter at execpc.com
> >Subject: Re: [Re: multistage rate level env generators]
> >Date: Mon, Oct 23, 2000, 2:05 PM
> >
> 
> > Grant,
> >
> > this is exactly what i was thinking just better explained (see my
followup
> > posting). a window comparator has 2 reference inputs which i think can be
used
> > like this. reference hi takes its input from the MUX connected to the
wipers
> > of the level pots and then put a diode from this input to the lo
reference
> > input to create .6 volt window. then the comparators input comes from the
> > buffer. the output from the comparator is used to clock the binary
counter
> > that switches the MUX. your idea of a differential amp is interesting. let
me
> > think about it - it sounds better than my idea above. i love the idea of
> > bipolar envelopes!
> >
> > i think i better order some parts, it's time to start breadboarding!
> >
> > -<mike>-
> >
> > "Grant Richter" <grichter at execpc.com> wrote:
> >> You don't really need a microprocessor for this. Some CD4000 logic and
> >> op-amps would do the job cheaply.
> >>
> >> Lets say you make a regular 8 stage sequencer out of a CD4051, a counter
> >> chip and output buffer. You could have 8 10K pots to set 8 voltage
levels.
> >>
> >> If you put a 1 meg pot on the wiper of each 10K pot and a cap on the
4051
> >> output before the buffer you could have adjustable portamento (slope or
> > slew
> >> time) for each stage. (Gabe Catanzaro showed me this one) If you want
> > linear
> >> segments, use a voltage to current converter before the cap.
> >>
> >> This could be the start of a multi-ramp envelope, but you need to make
the
> >> circuit self clocking. You need a circuit that decides when the slew of
the
> >> previous stage is close enough to the target voltage to step the counter
to
> >> the next "segment".
> >>
> >> Use a second CD4051 connected directly to the pot wipers before the
> >> portamento pots, and address it from the same counter as the first. Use
> > this
> >> output as one input to a standard differential amp, and the slewed
output
> > as
> >> the second input. This will give you a voltage proportional to how
"close'
> >> you are to the target voltage, or how much of the ramp time is
completed.
> >> (It will get closer to zero as the ramp approaches the target).
> >>
> >> That voltage could be positive or negative depending on if you are
slewing
> >> up or down. So a window comparator could be used to derive a "close
enough"
> >> signal. That signal could be used to advance the counter to the next
> >> segment.
> >>
> >> A Set-Reset flip-flop starts the process and the last stage resets the
> >> flip-flop and puts everything back to start up again. Additional logic
> > could
> >> be used to make one stage a "sustain" stage and the pots could be
powered
> >> from +/- to make bipolar envelopes. The circuit that decides when the
> >> ramp is close enough and generates an advance pulse is the key.
> >>
> >> ----------
> >> >From: patchell <patchell at silcom.com>
> >> >To: "M.Ruberto" <indridcold at usa.net>
> >> >Cc: synth-diy at node12b53.a2000.nl
> >> >Subject: Re: multistage rate level env generators
> >> >Date: Sun, Oct 22, 2000, 10:08 AM
> >> >
> >>
> >> >     Although I never actually completed it...I did something like
that,
> > but it
> >> > was hybrid.  The core was analog, used an OTA to make the standard
> > voltage
> >> > controlled RC network, it had two inputs, one to control the rate, one
to
> >> > control the level it would ramp to.  And there was a window comparator
to
> > tell
> >> > when the output was settled to the level voltage.  The other half was
a
> > micro
> >> > controller.  The parameters were stored in memory.  In theory, I could
> > have as
> >> > many stages as I wanted....but, I will never know, I abandoned the
> > project
> >> long
> >> > ago.  I keep thinking I will get back to it sometime....maybe.
> >> >
> >> > "M.Ruberto" wrote:
> >> >
> >> >> hi,
> >> >>
> >> >> lately i've been thinking about designing a multi-stage rate/level
> > envelope
> >> >> generator. i think i had this idea once before but never moved on it.
i
> > have
> >> >> some interesting thoughts about how to make it work but before i
start
> >> >> protoboarding i thought i'd ask if any of you folks have seen this
done
> > in
> >> >> analog before. if there are designs out there i would like to see
them.
> >> >>
> >> >> thanx
> >> >>
> >> >> -<mike>-
> >> >>
> >> >> ____________________________________________________________________
> >> >> Get free email and a permanent address at
http://www.netaddress.com/?N=1
> >> >
> >> > --
> >> >  -Jim
> >> > ------------------------------------------------
> >> > * Visit:http://www.silcom.com/~patchell/
> >> > *-----------------------------------------------
> >> > *If you didn't buy a home in Santa Barbara,
> >> > * You didn't pay enough!
> >> > ------------------------------------------------
> >> >
> >> >
> >
> >
> > ____________________________________________________________________
> > Get free email and a permanent address at http://www.netaddress.com/?N=1
> > 


____________________________________________________________________
Get free email and a permanent address at http://www.netaddress.com/?N=1