Utility Envelope Generator CGS78

pragtig! · Joined: Jan 04, 2011 Posts: 3 Location: circuit

Hi there, have any people here build this module?
It all seems to work although the longest Attack setting I can get is under a second. is this normal? This is my first envelope.
I use 12V/-12V and 1M lin-pots (100k Sustain) and a timing-cap C of 2,2 Uf.
I have experimented with the designator resistors but they seem to affect the minimal timing, (pot ccw) and not maximal timing.

Dave Kendall · Posted: Wed Jan 05, 2011 10:56 am Post subject:

Hi.

This is a guess, but the ATTACK pot is specced on the schemo as 2M rather than 1M. Try a 2M2 pot, or if you haven't got one, experiment with a 1M resistor in series between the RATT 4K7 resistor and the lug of the attack pot nearest to it. That should show whether the larger 2M pot gives you a slow enough attack...

cheers,
Dave

pragtig! · Joined: Jan 04, 2011 Posts: 3 Location: circuit

Hi Dave,

Thanks for your reply:). I did what you suggested and got the longest Attack time yet; about one second. Problem is the initial Attack time (minimal Attack time, pot ccw) is about half a second now. Changing timing-cap C to 1 uF resulted in overal shorter times. Maybe a huge timing-cap?

Dave Kendall · Posted: Wed Jan 05, 2011 7:54 pm Post subject:

Hi.

Yeah - it figures that the shortest time is now *not* zero, as there is now a resistor between *off* and the pot before it starts slowing down the attack rate (the minimum is now probably about the same length as the 1M pot on its own was at the longest setting, but as a test, it shows that the size of the pot *does* have an effect). As you suggest, a huge timing cap would be worth a try, although it would have effects on the other parameters too most likely, increasing the minimum times perhaps. The various resistors in Ken's mods table will probably interact with this too...

Caps in parallel are easy to switch in and out to experiment, or as an extra mod or two. Fonik has a nice diagram here; http://modular.fonik.de/files/CGS_LFO_mod.jpg showing how to use a SPST centre-off toggle switch to get 3 different ranges from a CGS LFO.If you don't have an SPST centre-off, somewhere on E-M I posted a schemo for using a DPDT changeover as a range control for a CGS31.... search should find it These mods work for many designs where the cap is part of an oscillator or as part of a timing circuit - I've also modded an old EFM VC LFO and a VC ADSR this way.

FWIW, I've seen quite small poly 3u3 caps available , and recently got some very cheap metal polyester 4U7 caps from rapid electronics in the UK - works out 25 cents US each !!) , although they're quite large. Being non-polarized, it's easier to sub them in and out with less risk of toasting something.... Wink

I'd build mine up and have a look, but I'm completely re-doing the kitchen on a near-zero budget, so time is not available.... I'm playing hookey by typing this..... Cool

Let us know how you get on..... it might be a mod switch or two will help out. Definitely a bigger pot than 1Meg would help with the range issue....
Hope this helps some.

right....back to scrubbing the walls Sad

cheers,
Dave

pragtig! · Joined: Jan 04, 2011 Posts: 3 Location: circuit

Hi Dave,

Thanks again for your reply.
That all made sense! so I ordered some 2M pots.
I already made the front for this module in Serge-format and have no space left for a switch. everything is so tight in there. But I appreciate the link, will be helpful in future projects. Good luck with the kitchen and hope you will post when you have build yours!

Cheers, pragtig!

modulus in moduli · Joined: May 14, 2010 Posts: 17 Location: USA

To have the minimum attack time, you set the potentiometer near its minimum value, which should be near 0, so the value of Ratt makes all the difference.

To have the max attack time, you set the pot near its max value, in your case 1 Meg, so if Ratt is a few k Ohms, that does not make any difference.

In this envelope gen., you load a cap through a resistor. Then if you multiply the C and R values you'll get the time to load the cap to 63% of its final value, so that will give you an idea of the time range. Multiplying 1uF with 1MOhm gives exactly 1 sec, using 2.2uF with 2MOhm will give you 4.4 sec.

defex · Joined: Jun 27, 2008 Posts: 36 Location: Toronto, canada

I have just finished one side of one and am testing it. and i am having a bit of trouble.

I am not using the RG resistor so i am using 5V output

the attack will get about half way to the maximum level, then the decay takes over.

if the sustain is at zero, the decay will go down to a lower voltage than the zeroed "no signal" voltage (almost -1v)

if the sustain is above the level where the attack prematurely ends, then the decay will act a s a second attack, after a little glitch.

the release will start from a higher value than the sustain is at.(unless it is at max)

The only thing i changed was to use 470R on the main output instead of 330R but that shouldnt matter because it is only driving a scope, and the resistor is inside the feedback loop. for the output buffer.

checked all my soldering and correct transistors, and diodes the right way about, the usual stuff. RSUS is 270k

Its harder to fix something that almost works right than something that doesn't work at all Smile

andrewF · Posted: Sun Jul 24, 2011 5:18 pm Post subject:

defex · Joined: Jun 27, 2008 Posts: 36 Location: Toronto, canada

well i got back to working on this again.

what i have found is that yes, the LED signal is there, no trouble.

the attack always rises to half way, no matter what the sustain is set to. then it jumps to the decay period, which will go up or down to the sustain level. depending where it is set to.

some voltages on ic1D

Pin 12 1v without gate, about .08v during attack phase, 2v with gate after attack.
Pin 13 0v without gate, 5v during attack phase 13.3 with gate
pin 14 5v without gate, -15v during attack phase - 13 with gate.