4093 NAND MiniModular

I got the idea to do a mini modular based on the 4093 NAND chip.  I decided to use each of the gates as an oscillator.  The three knobs across the top control frequency, control voltage, and output amplitude respectively.  The first switch puts some lag on the input control voltage, the second switch changes the oscillator to LFO mode and the third switch switches between continuous and gate mode.  These controls are repeated four times for  each of the NAND gates on the chip.  The top control is a starve connected to the supply voltage so I can reduce the overall voltage supplied to the device.  I made it 1/8" so I could connect to the upcoming MS20.

This video demonstrates the device in operation.

I also tried connecting this to using a keyboard by using the Cv and gate outputs from my Pulse synth.  This worked out well for the gate but the pitch does not change much with the output voltages, about one note per octave.

NAND Mini Modular

Making progress on the NAND mini modular synth.

Planar Crystal Live Rig

Demo of the rig.

Mini Modular NAND

I have been thinking about the NAND synth and giving it some additional capabilities.  This is my plan for a mini modular based on the 4093 NAND.  Each of the four NAND gates would be made into a separate oscillator.  Each oscillator would have three switches three knobs/pots and three jacks.

Switch 1 On/Gate: Determines if it is on all the time or if it can be controlled by the gate input.
Switch 2 Square/Saw: Switches a capacitor on the Cv input (I need to add this cap to the circuit shown above.)
Switch 3 VCO/LFO: Switches the feed back to a larger capacitor to make it into an LFO.
Knob 1 Pitch
Knob 2 Volume
Knob 3 Modulation amount (pitch)
Jack 1 Gate input
Jack 2 Cv Input
Jack 3 Output

This is the basic schematic for this.  Assume the gate starts in state Hi Lo then the 1n cap charges until it goes Hi at which point the output goes Lo.  The rate of charge and discharge is controlled by the pitch pot or by the transistor depending on the base current.  The LED monitors the output and turns on or off based on the output.  The LFO switches in a larger capacitor so it takes longer to charge.  The wave shape applies a low pass to the input signal.  I may have made some mistakes in this drawing but this is basically how it works.

This basic structure is repeated four times.

Next comes the filter which I described in a previous blog.  This consists of three simple resonant low pass filters and a clipping op amp differentiator.

This one does not have the voltage control transistor in it yet so I am still working on that.  It also shows a QF connection on filter 2 which should be bypassed.  The way this filter works is to have both a resonant low pass (with assymetry switch) and a low pass (Q turned all the way down).  Then subtract the two signals and clip them with the op amp.  This creates a complex waveform (supposedly).  Finally feed the result out through the final filter with resonance.  Need to add an additional input for external signals as well as transistors so I can modulate the cut off frequency.  The filter will from the base of the module.

RS is a rail splitter.

Planar Crystal 020

The planar crystal is a device which can transport someone spatially, temporally, and inter-dimensionally upon touching.  Those who touch it accidentally may wander the multiverse searching for the way home.


Motif ES7 with PLG-150DX (used for Skrillex sounds, Waldorf Pulse, Dune (Plug-in), mixed in Reaper, Kicks from Pulse and Little Phatty, some effects using Lexicon Vortex.  Drums using BFD Eco and MPC500.  Sound effects include home made electronic drones, filter modulated bulldozer samples and glitched effects.  Audio damage Replicant, Michael Norris Drone and glitch plugins.  Final noise sweep is a Thing 1 Noise generator modulated through a Korg EMX1.  Vocals recorded with DP004 digital recorder and modified using Michael Norris FX and Replicant.  Voxengo Elephant used for mastering compression.  EHX Stereo Talking Machine used on arpeggios near the end to give vocal effects.

Very Bizarre Filter Revisited

A while ago I came up with a design for a very bizarre filter which attempted combine a fuzz distortion with resonant low pass filter.  Unfortunately the prototype did not pan out.  This is a second attempt at a bizarre filter.  The idea is not to create a certain type of filter but to deliberately introduce distortion into a filter. 

This design is based on the new filter Ray Wilson is using on his Noise Toaster.  The Noise Toaster is an excellent sounding box which gives you a lot of options in a small space.  I was looking for a simple filter design so that I could try out some experiments without using a whole pile of chips.  The advantage of this filter is that it uses only one op amp.  I had been experimenting with Ray's Weird Sound Generator filter since it is very simple.  This new filter is an improvement over the WSG filter.

The first modification I decided to try was putting in some diodes.  I found that putting a diode and a potentiometer that could be switched on and off in the resonance section allowed for a lopsided resonance which should give some unique character to the sound.

The second modification I wanted to do was to subtract a low pass signal from the resonant low pass signal. That should result in just the resonance from the signal.  Next I wanted to add some hard edges to the signal so I passed it through the an op amp at high gain.  Finally I added another resonant filter at the end to reintroduce resonance on the newly created peaks.  Leaving in all of the controls gives nine different controls for messing with resonance and distortion.

The left trace shows Ray's original filter, the next trace shows the diode modified filter, the next trace shows the clipped resonance signal and the next image shows the re-filtered signal.

I have not built this device yet, I will have to wait to see if it works out.  If I build it I will post a video.  I am hoping to add a filter to a NAND synth so I may use this one.

Analog Sequencer

I finished an analog 10 step sequencer based on one of Ray Wilson's designs with some modifications.  For one thing I made it based on 9V so I could use my standard power supply.  I used a 555 for the clock instead of Ray's clock.  I ended up leaving most of the gate generating part which I later regretted but I would need to add a daughter board to fix it.  I used ten different colored LEDs on this device.  The 10 knobs control the voltage sent to the left output.  The knob by the left output controls glide, the knob by the right output controls step time.  The right output is supposed to be a gate but it only works if you have every other switch up and down.

Another issue is that it outputs a voltage from 4.5 - 9 V rather than 0 - 9V or -4.5 - + 4.5 etc which is problematic with most other synths.  In the next revision I would try to change that as well as a few other things.

Anyway, not too bad for a first prototype.  The videos demonstrate how it sounds:

And another:

Accompanied by a Yamaha CS01 analog synthesizer using a breath control to modulate filter cut off and processing both through a Lexicon Vortex.  The sequencer is modulating  a modified Atari Punk Console that is part of the Thing 2 device.  That device also has a NAND synth, noise generator and starve features.

I added a resonance knob to the CS01 as can be seen in the video.

Here are some pictures of the sequencer in progress.

Thing 2 vs. Steam Drone

In this video I do an improvisation using my Thing 2 NAND/APC/Noise drone box and Steam Drone box.

The Steam Drone is on the left.  The system combines a NAND synth with a resonant low pass filter based on the LM13700 OTA.  The LEDs use two transistors and capacitors to do flashing lights (yellow and red). The knobs control the frequency of the three NAND oscillators as well as the level of audio sent to the filter.  It turns out the filter performance depends somewhat on the input level.  the next set of knobs include Q, filter cut-off, and output volume.  The knob on the far left is a potentiometer that the input voltage passes through.  When this is all the way to the right it passes the voltage unaltered.  As you move it to the left the input current has to pass through more and more resistance essentially starving the circuit of voltage.  Starve is one of my favorite effects because the circuits get into a mode where they don't behave like they are supposed to, before they shut down completely.  Three of the switches control are used to link the NAND oscillators directly to the output, the other switch turns on the power.

The box on the right is the Thing 2 system.  This is a board I made by combining the following circuits: NAND synth, Atari Punk Console, low pass filter, zener diode noise source, blinking light circuit.  I had a bunch of layouts drawn up so I decided to combine all of these on a single board.  The noise part of the circuit was not making any noise so I kept increasing the resistor value over the op amp to increase the gain but it still would not make noise for me.  I noticed however that the whole circuit was producing noise so I settled for that.  This box has additional inputs for the control voltage of the 556 which basically controls mod and pitch.  I also added an input for the filter to process an external signal since the noise wasn't working.

Something interesting occurs on the Thing 2 when you starve the circuit.  The noise starts to become enhanced and the LEDs start having a big impact on the sound.  I was originally disappointed with the flashing light part of the circuit because the flashing of the lights is not related to the sound the device makes.  As you starve the circuit though, each time the lights flash they make a sound.  So I ended up really liking this circuit.  It was sort of a whim to combine all of these circuits on one board but I like the results a lot.  I ended up with two other indicator LEDs, yellow and green, in addition to the flashing ones, blue and red.

I had photo cube from Michael's which was big enough to hold the circuits and the knobs and I had used a photo cube for the Thing 1 device.  I came up with the idea to cut pieces of Fresnel lens and mount them on the inside surface of the photocube.  This gives a really cool effect when the lights are out.

Hadronistics

Hadronistics involves the study and application of Hadristors; gravitational devices which which play an essential role in FTL (Faster Than Light) transportation.

Took this one down.

This piece was done by improvising four different long sounds using home made boxes including the Thing 2 device depicted in the image.  The Thing 2 device combines an Atari Punk Console, a NAND Synth, an Odd Filter and a noise source in a single box with four LEDs, two of which are blinking.  The walls of the box are covered with fresnel lenses to give the impression of something which look larger on the inside than it does on the outside (TARDiS).  Other instruments used are an APC expanded with a capacitor array and odd filter used as input to a Bass++ drum trigger for rhythm patterns.

The Thing 1 noise source, NAND Synth 1 and 2 with VCF LPF 1, some signals were processed through a Motif with chorus, reverb, phasing and delay.  The entire sound was processed in Reaper with Apple reverb and Voxengo Elephant for compression.

Thing 2 and Steam Drone

This video demonstrates my Thing 2 device with a Steam Drone device.

Phaser, echo, reverb, and chorus added with a Motif.

Steam Punk Drone/Noise Box

I found this box at a thrift shop a while ago and was trying to come up with a circuit to put in it.  I finally decided on a circuit made from a NAND synth combined with resonant low pass filter.  The controls are as follows:

Top row (left to right) Power on, NAND Osc 1 to output, NAND Osc 2 to output, NAND Osc 3 to output

Middle row (L to R): Starve (10 kOhm Pot on supply V), Q, Fco, Output volume

Bottom row (L to R): NAND1 Fqc, NAND2 Fqc, NAND3 Fqc, NAND input level to filter.

Back (not shown) Ext. Audio Input, Fco Cv input, Output.

Note: Below the glass plate are a couple of blinking diodes.

The control voltage mix op amp is to the left.  It turns out this section did not function because apparently I forgot to tie some stuff to ground and -V.  The NAND gate chip is on the right.  In the middle is the VCF based on LM13700.  This is the filter schematic.  Note this has an error the transistors on the right need to be grounded.  I think there is another error too.  (The NAND synth section is not shown)  The TL972 I am was using did not seem to work in this application.  I am replacing it with a TL082 in future versions.  Not quite sure why.

Updated circuit diagram and PCB layout.  IC pin spacing should be .1".