A new kind of ADSR?

[synthmonger]

I was thinking of a new type of ADSR today. One with individual Attack, Decay, Sustain, Release outputs. Each output would have it's own gate and/or trig signal to repeat itself, so you can get an AD, ADS, or ADSR.

Does this sound possible?

I don't have any EG's ready yet so I can't test it until probably the end of the week. So if anyone has a standard ADSR out there, do you think you can see if it's possible to tap out the individual stages?

[Uncle Krunkus]

The biggest problem for your idea is that most ADSRs output the voltage fluctuations which are happening across one cap. So if you wanted an output which didn't have a release section, for example, you'd have to just disconnect from the voltage across that cap. Then where would that CV go? Straight to 0V? (just use another ADSR with no release dialled in) or would you want it to hang where it was? Then you'd need to S&H, and you'd still get droop eventually, which would be like a really slooooow release.
But I can see where you're coming from, if you just used the logic to run some other 4066s to only send out certain sections. Dropping to 0V elsewhere.

[Dave Kendall]

You can *sort of* get what you're describing from the EFM VC RADSR.
http://www.modular.fonik.de/files/EFMforum/orgEFMfiles/vcradsr.pdf
Pin 1 of the 4001 NOR gate produces a brief trigger pulse at the end of the attack phase, pin 14 of the TL074 gives a pulse at the end of the attack and decay phase, and I guess you could derive a pulse when the gate goes low at the end of the sustain phase using a bit of extra circuitry.

I guess any ADSR using logic as part of it's design should be able to be modded to produce trigger outputs. Perhaps this one on the Elby site.

I did hook up an end of decay pulse socket on my VC RADSR, but haven't actually used it yet for anything yet.....

cheers,
Dave
(edited to include VC RADSR link)

[synthmonger]

thanks for the link! I'll breadboard that asap and see what I can come up with.

[EdisonRex]

I have long thought that the envelope generator design would be best done using a PIC (or any other programmable microcontroller). This sort of functionality fairly screams out for that sort of approach to me...

[synthmonger]

Bah! You can do anything with PIC'ing around. Though I don't have the microcontroller hardware to program any PIC's so I'm stuck with analog means.


If this is feasible how difficult do you think it would be adding linear/exponential, and/or logarithmic voltage outputs to each stage?

[widdly]

I used to avoid all the pic projects out there. But a couple of weeks ago I built a jdm2 programmer like this... http://www.jdm.homepage.dk/newpic3.htm It works fine off my old linux machines serial port and it cost a few dollars to build.

I have since successfully built a midi to gate board with 6 outputs and a midi to CV board both using some 16f84's I got for cheap.

So bah to your no picing excuses


Thinking about your idea. You could have a universal envelope stage circuit. You could build many copies to create multiple stage envelopes.

It would have..

an input for the start value
an input for a trigger
a knob for the length (time)
a knob for the final value
a trigger output for when the final value is reached

It would sample the input value when the trigger was received and then slope to the final value over the time period.

The final release might be tricky. Perhaps you could use an inverted gate signal to trigger the release phase. This would allow multiple stage releases.

Then of course it would need CV for the the lengths too....and...and..and

[Rykhaard]

[ericcoleridge]

[synthmonger]

Ahhh! That should work perfectly for the gate/trig extraction. Surprised I haven't seen this one.[/quote]

[neandrewthal]

MFOS Retriggerable ADSR Envelope Generator With Dynamic State Indicator LEDs

If you can get a separate LED indicator per stage, why not a voltage?

[widdly]

On that MFOS, the separate cv outputs seem to all feed into IC-4b. I wonder if you could buffer these individually to get separate voltage for each stage.

[synthmonger]

[widdly]

The outputs of the 4066 in the bottom left corner are attack, decay/sustain and release. Instead of feeding them into the opamp (IC4-B) you could buffer each one separately. However the decay sustain are already mixed together. Maybe you could figure a way of using the extra switch the 4066 and the signal from the LED outputs to toggle the combined decay/sustain between two separate decay/sustain outputs.

Also, I think you will need comparators on the LED outputs if you want to use them as gates, since Ray mentions the LED's glow brighter depending on the envelope level.

edit..

actually I see the envelope output is fed back into the circuit so my idea wont work. Instead you could place the buffers just before the 4066's. Then feed the buffer into two 4066 switchs, one whose outputs are mixed together as in the diagram, and the other the feeds the individual outputs.

[synthmonger]

Well poop!

Doepfer A-143-2 does just what I wanted...well 99% of it, no gate output on the end of the sustain, but for everything else.

Ah well back to the drawing board...

[oldmanfury]

The ASM-1 ADSR can easily be set up to give triggers when the decay and release stages are activated:



The un-labeled output jacks to the side of D and R are trigger outs.

This is a great design for "chaining" ADSRs together and creating complex, multi-stage envelopes.

I have an eagle board layout for this dual ADSR... had a couple made up by futurelec and it works fine (though it could probably be improved upon). Can post it if you're interested.

-gerald

[synthmonger]

I'd surely be interested! The main reason why I want to make one with multiple trigger outputs is to chain ADSRs.

I breadboarded this http://yusynth.net/archives/Electronics/R-Kirschman-ADSR-1980.pdf yesterday and have been toying with it for different voltage output stages. I'm going to give a shot adding a polarity mixer for attack, decay + sustain and release with trigger outputs.

***A note for the schematic. The gate input needs a diode and a 1M resistor tied to ground or it'll trigger on it's own randomly.

I can get it to retrigger by feeding the cv output back into the gate, but it only makes it an AR and the release time is cut almost in half.

Ideally, if I can get it to retrigger at A, DS, R stages then I'd be a happy SDIYER

[synthmonger]

Well I figured out a great use for the extra switch. I forget which pins exactly (I'm at work now) but I found that routing a couple pins from the 4001 through the extra switch you can get it to freeze during it's cycle. Pretty fun lil' feature.