i have made my quad vca and adsr and they both work fine just the sustain doesnt fall to 0 while the key is still pressed its barely audible at the end of the envelope…is this something i could fix with a trim pot or something besides a gate?
Just to make sure it’s clear: Normally “sustain” means the typically nonzero level you get after the initial attack/decay and before the key release. You can set the sustain level to zero, though, in which case you’d expect decay to zero while holding the key down. Is that what you mean?
Also when you say “7555 ADSR” what specifically is the design you’re working with? Presumably one based on a 7555 timer and not a digital envelope generator chip.
The sort of thing you describe is not uncommon with analog envelope generators. The problem is that the envelope voltage is created by charging/discharging a capacitor, and there are diodes on the charging and discharging paths. Once the voltage drops below about a volt the diode conductance goes way down and it can take tens of seconds or more for the voltage to decay from there to (almost) zero. Kassutronics has a design that avoids that using precision rectifiers but it’s probably intrinsic to the design you’re using.
I’m just educated guessing it’s the sustain leaking voltage based on the trouble others have had and that I did to get it working…it goes through the envelope perfectly but at the end it still faintly keeps playing the note…I’m not sure what words to actually use
Even if sustain is set to zero you’ll still hear a faint note when the keystep is played
This is the yusynth design Eddy bergmans strip board layout
Yes, that’s definitely the diode issue. In fact the Kassutronics Precision ADSR design started off as a corrective mod of the YuSynth. So it’s an intrinsic problem, and as I said a lot of analog envelope generators have it. Often it doesn’t matter because you’re not letting notes decay to nothing before starting the next note. Or often you can add a small negative bias to the envelope that will bring it down to or below zero. But if it’s important you get a fast decay right down to zero, you’ll need something like the Kassutronics design or a digital envelope generator.
2 Likes
alright thank you for clearing that up for me so i stop thinking its something i need to fix. i have another layaout using the 3310(?) but i did not pick one up when i last ordered from ED
I also had this annoying issue with the YuSynth ADSR which really made it unusable for me. Fortunately there’s a simple fix by adding a single diode (D7 in the schematic below) in the output stage:
Original circuit:
Modified:
It reduces the maximum output voltage slightly but returns to 0V so you don’t end up with bleed-through in the VCA. Hope this may be helpful!
4 Likes
@CTorp tried it without success, I don’t know why, but I don’t think either of us really followed up on it. There does need to be a path to ground so current flows through the diode, otherwise there’s no voltage drop across the diode. The downstream module might (ought to) provide such a path but a 100k to ground would guarantee it.
Also, note the same problem occurs in an ADSR with the decay — with a diode in the discharge path it can take seconds for the voltage to reach the sustain level. That’s probably not a big deal in most circumstances, but if it is the Kassutronics design handles it; a diode drop on the output wouldn’t.
1 Like
Thanks for that suggestion, it’s always good to know about alternative approaches.
I’m always up for kludging circuits - a.few of my modules have hacked PCBs or stripboard add-ons - but in this case the kludged diode fix works fine with my downstream modules. I sidestepped the problem entirely for my second ADSR and built the MFOS circuit instead, which works well.
I have an AS3310 waiting for experimentation to build a voltage controlled version when I need (i.e. want) another one.
Yeah, looks like the MFOS design uses analog switches instead of diodes to control which paths charge and which ones discharge, which avoids the problem.