Some asm-1 notes and bug fixes

[tom at synth-diy.you-bastards.com]

I sent something like this to the list a few days ago but haven't seen
it show up yet.  Perhaps I sent it to a closed list from the wrong
email account, so here's take two.  If you see this twice, hopefully
one version or the other will make sense.

1. Filter distortion problem at high res:

I've been catching up on the past 5 months of synth-diy.  In December,
somebody working on an asm-1 was having a problem with their filter
distorting and sounding crappy at high res setting for certain cutoff
values.  I had the same problem, and have never seen it addressed
here.  In my case, the problem was caused by using TL082s as the
buffer opamps (positions U7 and U8 on the schematic available at

http://home.swipnet.se/cfmd/synths/friends/stopp/asm1vcf.new.pdf

).  When trying to drive these with a voltage below -12V or so (which
can happen at high res settings), they will go into phase reversal and
the output will go to +15V which sounds really nasty.  I swapped the
TL082s for LM358s which work fine.  Perhaps the schematic could be
updated to indicate LM358s or some other bipolar input opamp.

2. Using LFO to drive the ADSR gate:

The same person asked about using the LFO to drive the ADSR gate input.
It's not apparent from the schematic at

http://home.swipnet.se/cfmd/synths/friends/stopp/asm1adsr.pdf

but the LM358 at U2 is run between ground and +15V.  The asm-1 LFO
output goes to -15V which exceeds the common mode input range of the
LM358.  For safety, I put a diode in series with the gate panel jack
to clamp negative voltages.

3. Attack pot needs series resistor:

I found that when I turned the attack pot to zero ohms, the attack
peak would exceed the trimmed peak setting because the rise time was
so fast that there would be audible overshoot while the rest of the
circuit was switching to decay mode.  I put a 10 ohm resistor in
series with the attack pot to stop this and still get snappy attacks.

4. Inverting filter CV input fix:

Concerning the filter, the notes on Magnus Danielson's page

http://home.swipnet.se/cfmd/synths/friends/stopp/

(which I believe are transcribed directly from Gene's original notes)
say:

  An inverting amplifier is provided to allow positive modulation
  voltages to drop the VCF frequency, and because I had an extra
  op-amp. A 47k resistor is shown on the input to this inverter, which
  means that this input will have twice the "normal" sensitivity, or
  0.5 volts per octave

Ok, but the aforementioned schematic labels the input as -1V/oct when
it is in fact -0.5V/oct.  Furthermore, because U1-B is configured with
a gain of -2, it will clip if the input exceeds +/-7.5V.  This can be
corrected and control sensitivity maintained by making R5 100K and R7
47K instead of vice versa.

fTom.