Here is a handy circuit designed to use in conjunction with AR and ADSRs. It is a step and trigger delay. It can delay the start of a step or trigger pulse from the keyboard. The delay is from 2 ms. to about 2 seconds. When a key is pressed it will output a step or a trigger pulse after a preset time determined by a front panel control. If a key is struck again before the timing cycle has completed, it will re-set and start a new cycle on the last key down. Figure 1 shows the complete schematic.

A step or trigger pulse is inputted into J1 and inverted by Q2, its collector is high when there is no input. The first 555 timer, IC1, is a variable one-shot. R1 and C1 set its "on" time. R1 is the front panel control. With the 555 hooked up in this configuration, it will "one-shot" on a negative going pulse. So when Q2's collector goes low the cycle begins and pin 3 of the 555 will go from low to high and back to low again when the cycle is complete.

S1 on the panel is labeled trigger and step. When S1 is in the trigger mode, and a trigger pulse from the keyboard is supplied, Q2's collector will be the inversion of this and will fire the "one shot", IC1. When a step pulse is used the delay begins when the key is let up because the capacitor C1 is being held down by Q5. When the key is released, Q5 no longer shorts C1 and C1 begins to charge, completing the cycle.

Now, when you input a trigger pulse and hit the key before the timing cycle is complete, the cycle will start on the last key down. Why? Because of Q5. If you look at the location of Q5, you will find it placed right across the timing capacitor shorting it out and beginning a new cycle whenever another pulse is entered before the timing cycle is complete.

Now place S1 in the step mode, and input a step trigger from the keyboard. This now places a capacitor between Q2's collector and pin 2 of 161. The timing cycle begins on key down because Q5's output is integrated and begins the cycle. This also places Q3 and Q4 and IC3 in the picture. Q3 and Q4 are buffers for IC3. When you press a key down and hold it, you will get a pulse output from IC1. If you press a key, release it and press it again you get a pulse out. But, if you press a key, release it and don't press a second key, no output appears. You have just cancelled the instruction to delay with the aid of IC3, a TTL 74121, one shot. IC3 is wired in such a way that it will fire on the falling edge of the step. When it sees the falling edge, it fires. This output pulse goes to pin 4 of the 555's. Pin 4 of these ICs are labeled re-set. Normally this pin is held high for "normal use", but taking it low for a time re-sets the one shot. Pin one of IC 3 is normally high for "normal use" of the trigger delay except when cancelling the output pulse altogether.

Now we get to IC2 and IC4. IC2 is a fixed one shot that is used for the new trigger pulse output. It fires when IC 1 goes from high to low and then fires a fixed output pulse of about 2 ms. to J2. This is the delayed trigger. IC4 is a Quad NOR gate and is wired as a set-re-set type circuit. Pin 3 of IC2, the delayed trigger, is used to set the circuit, and latches pin 4 high. When the step pulse is released, pin 6 of IC3 fires and re-sets the latch and pin 4 goes low. What we have here is a delayed step which outputs at J3. This mode works only when S1 is in the step mode, and a step pulse from the keyboard is used. (NOTE: This trigger delay works fine with the PAIA D/A converter, use the "D6" output for the step input, this will give step and trigger delay.) This is very handy if you want delayed vibrato, see figure 2.

If all of this has you thoroughly confused, we'll add to the confusion with the timing chart shown in figure 3.

All four ICs and associated parts can be assembled on a 2" X 4" printed circuit board or perf board. With the values of R1 and C1 as shown in figure 1, 2 ms. to about 1.5 seconds can be expected. These values can be changed to expand the time. Use good quality capacitors for C1.

The photograph, figure 4, shows the orientation of the front panel. Notice that on this particular unit the circuit board is mounted on the top rather than the bottom of the panel. This was done on this unit purely as a matter of convenience in fitting it into a custom system configuration. The circuit board can be mounted in any convenient location.

Calibration of R1 was done with a triggered sweep scope, and the delay time read directly off the scope. See figure 5. Example: The time base is 500 ms./division and you press a key down, this starts the sweep. When the delay is over a pulse will appear. If the pulse appears after the scan has passed three divisions, then the delay time is 3 X 500 ms. or 1,500 ms. or 1.5 seconds. Or, you can guess at the time delay like you guess at the time of the AR, ADSR, Filters etc. The one thing lacking is calibration.

Figure 6 provides some sample patches to aid in familiarization and initial experimentation with this module. Like all new toys, you will be busy for hours... have fun.

This patch demonstrates the delayed step output and the key-up reset function. By adjusting your playing technique the string part can be made to swell in and out.

This patch demonstrates the delay period starting on the release of a key and may be cancelled by depressing another key.

This patch, demonstrates the "straight" delay function. The trigger pulse is delayed, but may be stopped by quick, successive key depressions.