I notice that the clone Minimoog contour envelope ADSR module, by AJH…has a simple switch that toggles from normal to slow, effecting all 4 stages. I have one of these and the effect is quite good.
I wondered how this is done and if it could be transferred to a modified 911 ? Some sort of switch adding a capacitor somewhere at a guess ? I’ve experimented with the 911 main 10u capacitor but all this does is lengthen or shorten the envelopes points and doesn’t actually slow down the envelope.
Are you talking about modifying the slopes of the attack and Decay/Release, (T1, T2, T3) and duration of these periods?
Like a linear vs exponential/logarithmic slope switch?
Sounds more like different timing. But this would only effect the 3 timing parameters and never the sustain.
Yup, the 10uf cap changes all T timing. Changing the slope would be a more than a single part change… but not much more. A 5U 911 should have room to add something like a lin/exp circuit, or some such mods.
Attack perception in then 1-10ms area has a lot of potential tonality in the slope/curvature/rate variation. Experimenting with a eurorack Tide module provides AR/AD envelopes with continuous variation between many different variation all with distinctly different flavors of … dare I say ‘snappiness’? :0)
I would rate this change would be DoublePlusNeat!
You may try by raising the value of the pots, or add resistors in series to them. This should lower the currents that charge (T1) / discharge (T2, T3) C1, then slowing the timings.
I am working around a clone of the 911, but at present I have no working 911 to test with.
Leonardo
Thanks for your replies ..
The AJH adsr module has a switch marked as ‘slow’ which slows down the attack time making the angle of attack quite slow and this is quite a good feature. I’m not sure of the effects on the DSR section as I’m mainly interested in the slow attack and yes in reply I suppose I’m looking for a
Linear / Exp type of range on the attack slope.
The original 911 has only one speed variable attack slope ..I guess its relatively easy to alter the circuit of the 911 attack section but alas it’s beyond my knowledge of the working circuit. A wild guess is the capacitors of which there’s 4-5 in the whole circuit.
If the AJH can perform the action described by a toggle on/off switch it must be something quite simple but as mentioned I simply dont have the
theory to apply it to the mini Moog contour circuit. I will try the adding resistors as mentioned ..thanks.
Ok, here’s there relevant passage and the URL to the device.
Slow mode is selected by putting the SLOW switch to the down position. It slows down the Attack, Decay and Release curves to around 20% of their regular speed, so that long evolving envelopes of well over one minute in duration are now possible.
The Slow mode can be used in all configurations, so it works in ADSR, Delay-Hold and Delay-Hold + Retrigger modes.
From the text of the device, it looks like they’re just adding 5X the capacitance. If you examine the curves with a scope, the slopes are not so pronounced they would introduce much difference in sound for the range of new timing(s) available with this mod - 10’s of seconds. I would expect no change is characteristic slope for common set of timings of slow and slow-NOT settings.
I think I once tried varying the 10u cap of the 911 but the results were not very good.
I may write to Alan as I’ve been in contact before and ask him if he’s willing to give me a clue as to how they did it. Don’t forget this is the Mini Moog contour circuit not the 911 he’s modified,
Yeah, I started to get that through my head towards the end. Lol thanks!
Since, I had this locked and loaded anyway. I substituted the 10uf major timing cap with a 2.2uf. As expected envelope slope shape remained the same, while timing sped up a factor of 4X.
Yes..my results as well ..!
I’m pretty sure the big 10u capacitor is not the answer, it’s somewhere else.
If I get a reply from Alan I’ll post it.
Indeed. The characteristic log slope is related to charging caps. Changing the cap isn’T going to change that.
I’m using a clone 911 that I’m trying just to work around the inferior pots they used while my good quality ones are coming from the UK. I already used a decade box to map the pot taper I need to get the range of timing on the front panel correct.
Changing the cap only shortened the range up past ‘the bad parts’ lol. I still have a poor clone (for now), it’s just faster!
Ok, looked at it. I see exactly the same circuit used in the 911 and 911A in the minimoog, down the 100ohm min-range resistor. So, we’re looking at the right thing there. Bet he just added more C 
The slope is pretty tame for log - the top 1/2 is where all the action is, and the Log slope is interrupted before it gets extreme. So, maybe double the C and halve the R get’s you a quicker ‘usable’ slope? Now I’m not so sure… this is fun!
Oooo and screwthefuckingcaptcha sheesh.
For all you folks following along at home today, I changed the timing cap on my 911 clone to 1/10 the original value of 10uf. I did this as a short term work around for difficulty finding a true-log 10% taper in this ‘snap -in’ vertical SMT packaging. ALPHA, and ALPS both offer a 2-piece linear pot that doesn’t match the timing on the front panel.
Using a resistance decade box, I mapped the correct R values for thee corresponding front panel timing - it’s a nice true-log distribution, just like the front panel. So… no mystery taper involved.
Anyway, none of these modern SMT pots taper the low R end of the element well, so fast timing is out the window.
Changing the cap moving the timing ranges to different parts of the dial, making some timing accessible, while other timing isn’t ( > 1000ms).
Having five 902’s, converting two to fast versions, presents no real lasting deficiency in capability. In fact, it makes these clones waaaaay more snip-snappy. A pair lets you mix all those wonderful filter/amp attack/decay timbres you can’t reach with the stock config.
Moog’s design still shines through. 
I received a reply from Alan of AJH synths. These are great authentic sounding modules true to Minimoog sound but with more versatility and I’ve always had great support and help from Alan.
So the answer quite simply is that he achieved the effect of slower ADSR
by adding a second similar capacitor in parallel with the main timing capacitor. The toggle switch switches it in and out of the circuit.
I haven’t done a trial yet and as previously stated this is the Minimoog contour circuit he modified but it’s similar to the 911 and should have the same effect.
Later that day…
added 10u cap = hardly any difference.
added 100u cap gives a marked slowing down of attack/ decay etc. Going to be quite a useful feature.
Indeed, 100uf would multiply the timing ‘constant’ by 10. A 2x wouldn’t be that noticeable on a log pot - 2X is always just a little turn to the right - finally the end only be double.
A 10x multiples every position by 10, which is equivalent a 1/4 of a turn!
I did the opposite on my module, replaced the 10uf with a 1uf. To give me a x10 reduction in timing. Now I have a lot of range to play with attack ensembles.
The ALPHA pot turns out to be a decent performer, especially relative to the stock pot.
I might put a toggle between 1X, 1/10X, and 10X.
