Quad Divide

My own design. This is 4x CD4017 counter ICs. I ran out of the correct color banana jacks. This is also a reused panel which is why the layout is a little funny plus there are extra holes where the old circuit board mounted.

Perhaps a better name for the module is "Simple Pattern Generator."

The outputs should all be orange. They are GATE outputs. The green is a GATE input.

The column of switches on the left choose the division from 1 to 6. The column of switches on the right choose between outputs 0-5 of the CD4017 (0-5 = 1-6 for practical purposes... the first one is just labelled "0"). The output goes to the jack on the right and lights the LED.

Each channel is a row. So there are 4x 4017 counters that can divided and arranged to do a variety of simple patterns.

Unlike ripple counters, patterns like "1-2-3-4" are available. You can program 6/8 time and things like that easily.

I only stopped at 6 because I only had 6 position switches available. It would be simple enough to increase this with switches with more positions.

Use as a Sequencer

The 4 gates are extremely useful for creating interesting drum patterns. The gate outputs are connected to envelope generators which control noise and sine waves tuned like drums for classic electronic drum sounds.

The gate outputs can also be made into CV for VCOs or other modules. I run the outputs in my CV mixer, then run that summed output into a VCA, then feed that into a VCO. Patched this way I get several options to adjust the sequence in ways a conventional sequencer doesn't allow. The most interesting is how the pattern rearranges when changing the reset controls. Since I'm summing the CVs instead of sequentially using them, the number of pitches available are numerous because each output is tuned different, and then they create more pitches when multiple gates are active. The CV mixer controls allow each gate to be tuned to a different pitch. The offset control on the mixer functions as a "range" control, and the VCA functions as a "scale" control.

The next version

-Increasing the available steps to 8, 10, or using different or multiple chips to go beyond 10 are all options I am considering.

-A CV output. Each gate output will internally connect to a manual attenuator, which will then feed a CV mixer. This provides a sequenced CV output without using other modules (see above).

-Replacing the manual rotary switches with CV control of reset point and output selection. A block diagram and sketch of the panel controls are below. LEDs would be required to give a visual indication of what the current selections are. The linear CV would be presented to a chain of comparators. Each comparator output would control an analog switch replacing the old rotary connection. Both a manual and a jack CV input would be available. For fully automatic CV control, the manual control should be fully CCW. An attenuator may be necessary for the CV input so that large signals can be set to move the selections in small increments (1, 2, etc, when the full signal may cause large leaps only).

Several solutions to the CV control of switches come to mind:

1) using a µC and programming the logic into a single chip. This is probably what most people would do.

2) (for 8 switches/steps) using 7 comparators to drive analog switches (2x CD4066) that will be connected to a diode + Schmitt trigger ("Mickey Mouse Logic") multiple OR gate.

3) (for 8 switches/steps) using 7 comparators to drive logic gates to drive 1x CD4051 (8 to 1 MUX) to connect output lines to reset pin. The logic gates convert the 7 comparator outputs into a 3 digit binary that addresses the MUX. My best solution so far is to use 6 XOR gates and 4 OR gates, shown in the schematic above.

4) After looking into the above ideas, I became aware of the "8-bit priority encoder" IC, such as the CD4532. This is exactly what I was going for with the logic gates. This idea is shown in the diagram above that says "9 chips/channel." Quite a lot of chips. Perhaps there is a way of reducing this other than using a µC.

5) LM3914 LED driver IC appears to have all of my comparators, plus I can get single output using the "dot" mode. Sadly, the outputs don't cleanly switch between each other in "dot" mode, to be more pleasing visually. I experimented with the outputs driving resistors instead of LEDs.

Audio Demos

The mp3 files can be downloaded with the arrow buttons below and to the right.

Quad Divide Demo

This patch includes the Quad Divide, the Dual LFO, CV Mixer, PH-3 Phaser, Dual Delay, Ring Modulator, Quad and Dual EGs, and a typical VCO, VCF, VCA voice setup.

The first 3 channels of the Quad Divide are summed by the CV Mixer, passed through a VCA (for scaling), and fed into a Dual VCO module. The last channel is triggering an envelope for a VCA that has the ring mod as an input. The clock is provided by 1/2 of the Dual LFO, and this is also driving the Dual EG for the VCF and VCA of the dual VCO. The two VCA outputs ("normal" voice + ring mod) are summed and sent through the PH-3 Phaser and Dual Delay. The other 1/2 of the Dual VCO is doing a slow triangle sweep of the VCF.

Sanskeys

This patch uses no keyboard or MIDI input. The Quad Divide is working along with a 4024 binary counter to provide gate signals to 5 EGs, and 1 of those EGs triggers a 6th EG when it finishes its cycle. One dual VCO has two pulse wave outputs feeding a VCF, 1 EG is provided PWM. Into the same VCF is a ring mod signal, from another VCO sine out and the ring mod's internal sine. Both VCO inputs to the ring mod are being modulated. The VCF is the state variable kind, and the low and high ouputs are panned left and right. An aux send to the VC delay module in also in the right mix, and in parallel to the main signals. The delay time is under voltage control. Signals from my dual LFO unit are modulating the VCF cutoff and the master clock speed. They are gated by VCAs, under control of the EGs. This patch mostly played itself; I simply turned things down at the end to give it a finish.