AR/AD Transient Generator

attack-decay / attack-release envelope generator

Courtesy/Thanks to: Ryan Williams

Description: This is an attack-decay or attack-release envelope generator. I call it transient because I like that name. The design was based on Ray Wilson's Retriggerable AR Envelope Generator. I chose his design due to its simplicity, and because it had the same features that I wanted. I have redesigned some parts of the circuit, and added a few things. Both, the trigger and gate inputs have schmitt trigger circuits added. The new inputs are active at about +2V. The peak voltage is set at +5V. The diodes in the attack and decay current paths have been replaced by CD4066 bilateral switches. This removes the diode 'on' voltage at the end of the decay stage. While the gate input is high and the attack stage is finished, the output voltage is held at +5V by an additonal switch. When the gate is removed, the decay stage starts immediately. When the gate is non-existent, a trigger starts the transient in 'one-shot' mode. Then the decay stage starts as soon as the attack stage is complete. The mode is chosen by using either the gate or trigger input. Both inputs can be used at the same time to re-trigger the transient if it is in the decay stage. LED indicators are connected to each EG output. This design could easily be converted into an ADSR. (see schematic below).



The prototype was built onto a circuit board made with a CNC milling machine. I was able to build and test the design in an afternoon. Initially, I blew a 4066 switch. This was caused by inadequate current limiting at minimum attack or decay settings. I have seen some designs that do not limit the current but in my case, the part was destroyed instantly. I have now added 1K resistors in series with the attack and decay pots to limit those currents. The maximum current through the attack switches is about 12mA. That is beyond the data sheet's spec, but it only happens for a very short time at the beginning of the attack stage. If this is a concern, one could use 1.5K resistors and the current would be kept within the 4066 maximum rating. I have also read that Fairchild 4066s have more problems with this than other brands and that is what I was using.