[Re: VC ADSR (was Pulse VCO Help)]

[tomg]

>From: Harry Bissell

Hey Harry;

>Tom: I just saw your mods. Couple'a comments...
>1) Why the feedback on the 3080? Chris' circuit really isn't a sample/hold its
>a "charge pump" and an integrator. Are you relying on the feedback to stop the
>circuit from drifting? I have to think on this one more... there may be a
>"hidden weenie" in there. 

If you don't use a sample & hold configuration you can't drain the charge
off the cap for decay/sustain. Mine isn't a sample/hold ether. It's a voltage
controlled resistor.

>What about offset trim for the 3080? Needed?

Not really, it becomes very stable when you close it up.

>2) The fet is probably redundant. The TL074 has plenty high enough impedance
>for this circuit. You might need it with a .001uF cap, like in a filter, but
>not with a 10uF.

A 10uF was too big it had a whopping 25sec charge time but the note tended to
linger just a little (slow discharge time) so I replaced it with a 4.7uf. With the 4.7uf 
cap and fet charge time is 12secs. Replacing the fet with the 074 in the loop CT is 
7secs. I figure 4secs. is worth a fet ?? I read "you should always use all of an ic 
when you can" somewhere one time and it stuck with me, so I used the extra 
amp as a load buffer. It' up to you if you are going to build one for yourself.

>3) Chris needs the sustain buffer because of the low impedance of the charge
>pump. You don't need it. you have 220K input impedance which should have
>almost no effect on the sustain drive...

If you don't care about mixing voltage you can hook pin 15 directly to the
wiper of the sustain pot. The problem comes up when you add a 10K resistor
adder, all the sudden the pot has no range at all. 

I wanted to connect pin 3 of the 3080 directly to pin 13 of the 4052 and use
220K resistors for mixing and pull-up, but the 3080 didn't like being connected
like that, going high and staying there. 

So this is a solution that works. I'm sure something better could be done.

>4) If you can stand the offset of diodes in your voltage inputs, it would
>eliminate the backfeeding of the CV from the pots back into other circuits.

20K is usually enough. Adding diodes would be a good idea. As long as
it doesn't load anything down. Why not?

>I like your transistor input circuit. Was Chris' schematic wrong about the
>trigger in on the 555? I thought it needed to go negative...

Wellllll....yes. You are correct, Chris' circuit trigged on key up not key down.
That is a little unnatural feeling. I'm sure he would have gotten that fixed.

>On your reset to the 555, I like a switch so that there are two modes, 

1) with a short trigger, the unit gives AR response, ie if the attack peak
hasn't been reached by the time the gate is gone, the release starts right
away
2) same as above, but the attack time goes to full peak no matter what. This
is real useful for filter sweeps where you want to trigger and walk away. I
have manual triggers on all my AR or ADSR units.

That's easy enough to do. Get a dpdt, use one side for pin 4 of the 555 
switching between 3.9K/470ohm resistors and +V. Use the other side
of the switch to switch pin 9 of the 4052 between the 10K resistor going
to gate and pin 10 of the 4052.  

I got it to cycle, so you could use it in lfo mode if you wanted. But was
unhappy with the range, so I left it off. If you want to try it connect pins
9 and 10 on the 4052 and pins 6 and 2 on the 555. You could use another
dpdt switch if you wanted.

Here is a short list of things I fixed or changed.

I did not label the resistor connecting to pin 2 of the 555 to +V. It's 100K.

I mislabeled the pins on the opamp used for the inverter out. The output is 
correct the (-) input is pin 9 the (+) input is pin 10 of the TLO74.

I changed the charge cap to 4.7uF from 10uf.

I changed the 22K threshold resistor on pin 13 of the TLO74 to 18K. 

-tg