AW(2): AW: "Stacked" Power Supplies - any thoughts?

[Stopp,Gene]

Thanks for the comments, Juergen. I think you have a good point. One
idea that I have is to make another matching cabinet with four more
panels in it, with plenty of room for additional "accessory" modules
(like more envelope generators). Here are some thoughts that I have so
far:

The PC motherboard has a sound card in it, to provide the MIDI I/O for
the software which drives the 8-channel DAC. I plan to use the existing
program that I wrote for my standalone PC-controlled 8-channel DAC,
which currently drives the old ENS-74 modular in my setup. That program
right now provides:

By DAC Channel:

1 - Pitch CV
2 - Note On Velocity voltage
3 - Pitch Bend voltage
4 - Modulation voltage
5 - Aftertouch voltage
6 - Volume voltage
7 - configurable to any CC#
8 - configurable to any CC#

The gate on DAC channel 1 is the Note On/Off status, and the gates on
DAC channels 2-8 provide a threshold on each of their voltages (with
settable threshold value). For example, if I set the threshold on
Velocity to 64, then the gate for DAC channel 2 goes high when the
velocity voltage exceeds 64.

As you can see, this provides pretty good coverage for monophonic
expression on 1 MIDI channel. Since it's all in the software, I can
easily expand the code for two-channel operation, or really whatever I
want. This includes slowly changing voltages for software envelopes and
ramps, and perhaps sync'ed LFO's and other arbitrary functions. So with
this in mind I hope to overcome my lack of hardware envelope generators.

Besides, for overall amplitude, I almost always use just an AR envelope,
and ADSR's for everything else.

On the same panel as the 8-channel DAC, I have the four VCA's plus the
four attenuators, since the DAC outputs are full-scale 7-bit voltages,
from 0 to about 10 volts (semitone steps at 1v/octave). The VCA's are
for controlling the depth of things using DAC voltages, and the
attenuators are for scaling the raw DAC voltages themselves.

Because my MIDI I/O is a sound card, I also brought out the sound card
audio lines (mic in, line in, line out) in case want to trigger .WAV
file sounds as well. This leads me up to a question - does anybody know
of a program that can play .WAV files from the DOS command line? Since
the code is currently running in QuickBasic, I'm thinking that all I
need would be a DOS command to play a wave file, and from within the
code I can call a DOS SHELL command whenever I want to trigger a sound
to the sound card output jacks. I'd rather not run Windows 3.11, and I'd
really like to stay away from Win95 since this is just a 486/66. DOS
6.22 should be good enough. When the QuickBasic code is debugged, I
usually compile to .EXE anyway, but I am open to using C if I have to.

Also, I've been thinking about that ADSR that's on the DAC panel. My
panel graphics only have the standard ADSR controls (4 pots, gate
inputs, manual trigger switch, output) but no CV inputs for time
constants, so it would be a little bit of a waste to use an ADSR chip
like the 3310 or 2056. Although these are fine ADSR's, it would be nice
to use them where the time constants are externally CV'd in addition to
the panel controls. So, I dreamed up another circuit. Remember my DIP
relay envelope generator experiments a couple of years ago? I'm thinking
of making one for this section, and for the resistive elements using
opto-couplers with the LED currents controlled by the panel pots. Based
on experiments with opto's, I know that you can get *extremely* long
time constants by charging and discharging a buffered cap with opto's.
You can adjust the LED current in the opto to get ramp times of hours,
or even days. The slopes are linear, but that would be OK, in fact
probably useful. This might be a nice way to get very long evolving
changes in a patch.

Current project status - all panels are built, all panel components are
installed, all knobs in place. It looks really nice, but of course makes
no sound. Next step - build internal cables from the PC to the panel
(keyboard, mouse, video, sound card), hook up the disk drives, and
verify the operation of the DAC. These parts have all been bench tested,
so I expect everything to go smoothly. However I am bracing myself for
the wiring of all the circuits to the panels - those experienced
builders out there know that this can take as long as all the previous
steps put together!

 - Gene

 ----------
From: Haible Juergen
To: Stopp,Gene
Subject: AW(2): AW: "Stacked" Power Supplies - any thoughts?
Date: Wednesday, June 03, 1998 8:35AM

Gene, after looking closer to your list, I wonder if you have enough
envelopes in your system.
In retrospect, I'd rather have more envelopes (and maybe not as many
filters) in my JH-3 system. I love to have several different filters to
choose
from, but every time I want to create one large patch all across my
system,
I get short of envelopes (I have 5), and still have at least two filters
unused.

Just thought I should mention this,

JH.

*==================================================================

> Upper left panel:
>
> Completely patchable ASM-1
> 4-input mixer
> two reversible attenuators
> AR envelope generator
>
> Lower left panel:
>
> 4 linear VCA's
> 1 Quad Passive Attenuator
> 1 VCADSR (probably a CEM3310 or SSM 2056)
> 1 IBM PC 486 motherboard w/ floppy and HDD
> Monitor, keyboard, mouse connectors
> Sound card audio inputs and outputs
> 8-channel DAC controlled by PC motherboard
> 8 CV outputs
> 8 Gate outputs
> MIDI in and out
>
> Upper right panel:
>
> 8 VCO/VCA pairs, saw/pulse
> 1 Master VCO CV controller
> 1 Output mixer for VCO/VCA pairs
> 1 Quad tri-square LFO
> 1 Shepard Function Generator (8 ramp and 8 tri outputs)
>
> Lower right panel:
>
> Quad VC Bandpass Filter Bank (4 in parallel)
> Switched Capacitor Filter
> 4-pole Lowpass VCF (cascaded 3080's)
> Stereo Digital Audio Delay
> Final Output Mixer
> Power Switch and voltage rail LED's
>
> The 8 VCO/VCA pairs, the quad LFO, and the filter bank are all
circuits
> on 4 partially-stuffed ASM-1 circuit boards.