Electrolier (September night in Virginia)

​I've been working around the clock to finish my Hackaday Prize application and pack for Hacking Nature's Musicians in Mexico. As closure for my recent chapter, the various circuits that I've shown you over the past few weeks have migrated from my bench and into a sculpture titled "Electrolier (September night in Virginia)," 2018. Here​ is an informal video:

The video quality isn't great... (shot with my iPhone under bad lighting and extremely messy studio)... but hopefully you get the idea. The sculpture contains one instance of my Mother Nature board design, recognizable by the dense tangle of white wires connecting logic gates, and six sound generating "animal circuits" each with its own speaker. The colorful graphics are spray painted cardboard, and everything is hung on a tree structure under a moon (globe pendant bulb). Oh, and that large moth is made out of silk velvet that I dyed and embroidered with an old industrial "Ultramatic" machine​. I would have used small molex-style connectors to connect everything, but there was no time.

It's interesting that so many long wires didn't screw up my signals. I credit this to my use of common emitter amplifiers to buffer the signals, and the fact that I'm not drawing much current for anything you see (or hear). The whole sculpture is powered with a 12VDC / 1.5 amp power supply. I also used multiple 0.1 uF ceramic capacitors between power and ground on most of the individual perfboards. It's critical to remember power-ground capacitors when you're dealing with a lot of amplified signals banging on your power rails. I'm careful to avoid really small gauge wire, and I add many pathways to ground ("let ground abound.")

So... now I am packing my electronics bench into a suitcase and praying that security does not detain me at the airport. I'm throwing in packets of desiccant and crossing my fingers that jungle humidity doesn't zap everything. Making analog electronic circuits in the jungle will be interesting to say the least -- stay tuned for the sounds of los músicos de la naturaleza​ en Quintana Roo!

View fullsize

View fullsize

View fullsize

View fullsize

Mother Nature in perfboard

I've been busy migrating my "Mother Nature" controller circuit out of my breadboard and into perfboard... and yes, it's insane but no, I don't have time to produce a printed circuit board because I'm leaving in six days for a fellowship in Mexico. For the record, I do not recommend soldering so many components and connections in perfboard because the risk of error is high, either from bad solder joints, signal interference, or just plain confusion. I plan to design a printed circuit board for future embodiments of Mother Nature. Stay tuned.

As you can see in my previous log, I'm surrounding my perfboard circuits with spray painted cardboard to make them look cool -- and by the way, you could use this quick-and-dirty strategy to make a "starving artist's badge" for the Hackday Superconference.

As for the design of my "Mother Nature Board," aka random pulse generator to trigger various events, I have some additional technical tips to share:

• Don't have logic ICs on hand? Build discrete transistor gates. This approach has the advantage of common components (NPN transistors, resistors, diodes) and you can add multiple inputs to the same gate -- which is useful if you discover that an event is triggering too often... just add another input to the gate and the outcome will become less frequent. Not triggering often enough? Remove or change an input to the gate. You can even tie an input to ground (or power) and let the other fluctuate. I use 2N3904 transistors for as many things as possible, but you could also build these cool light gates described by @Dr. Cockroach 

There's a limit to how many things you can drive directly with a signal such as the output of a logic gate. Good engineers read data sheets and calculate voltages and current at various locations in their circuit. Impatient engineers add a generic common emitter amplifier between the signal OUT from a logic device and the signal IN to whatever you're driving. This hand-waving approach to buffering will not work in all cases! But it will probably work for most slow logic applications where you want to transform a signal multiple times and drive some light loads. I'm an artist who prefers prototyping to math, so I work a lot with "try it and see" circuit design. Plus, I am not building a nuclear reactor. Note that a common emitter amplifier will invert your signal. Pay attention to whether you need your signal to be active high or active low. Note that a 555 timer in monostable aka one-shot configuration is looking for an active low input. If necessary, invert the signal again to get what you need.

Mosfets make great electrical "on/off" switches because they don't draw current on their gate (==the mosfet equivalent of a transistor's base). They just need a voltage. I use mosfets for the last step in my Mother Nature circuit, or the point at which I want to turn power on or off to a particular sound circuit (the load). Make sure you have a gate resistor to ground or voltage will "sit" on the gate even when the signal is low (and the mosfet won't turn off).

I hope these tips are helpful. To end this log, I give you a video showing my Mother Nature circuit in perfboard with two sound circuits hooked up. If you watch the LEDs carefully, you will see how certain combinations of logic are triggering the sound circuits. The cricket is wired up to chirp most of the time, whereas the Katydid is triggered less frequently. For this demo, I hooked up only two sound circuits because it's already hard to understand what is happening, and more circuits becomes a sort of natural chaos... which is my goal, as you will see in forthcoming logs.

Same signal, many sounds

I've been migrating my circuits out of the breadboard and into soldered form. I'm also hooking up speakers so that my sound generating circuits are ready for installation in a sculpture of Virginia's nocturnal ecosystem. Because most of my circuits use a custom piezo electric speaker (built with my own amplifier board), I have to install each piezo element into a physical housing... and there's some interesting "play" here. One electronic signal can generate many different sounds depending upon the speaker's physical design. Here's a video of me testing various installations of a piezo element for an insect that makes a steady background noise.

It's cool to experiment with the interface between electrical signal and material properties. It's also critical to get it right, or else the piezo element will make a horrible rattling or screeching. Even a tiny drop of glue (to fix the piezo element in place) will alter the sound quality --usually the pitch-- and superglue can sound different from rubber cement or hot glue or tape. All of this is no news to a human musician, who knows that a note ill-played (even slightly) is a fail, and teensy subtleties often make the difference between average sound quality and genius.

It is interesting to wonder what makes animals of the same species sound different in nature: is it their electrical impulse or their physical form? (Is it their schematic, their CAD file, or their bill of materials?) Members of the same species inherit the same circuit design as well as physique. External physical factors, such as climate, have some degree of influence over material properties of the body --for example, a dehydrated cricket sounds differently than the same cricket wet with rain; a fat frog sings a different tune than the same frog skinny; and so on and so forth. We also know that "animal circuits" are sensitive to electromagnetic fields, and capacitive coupling undoubtedly affects individuals in close proximity. I suspect that physical factors play a greater role in the variability of an individual's song if but for no other reason because birds sound the same whether they are sitting on a tree branch or an electrical transformer (at least, I think they do). PS: Eventually I will tackle the challenge of bird song, but their vocal complexity requires more computation... which is why I have started with insects.

To end this log, I leave you with one more video clip of my "background noise" insect. Here it is with a plastic spool installed over the piezo (now painted green, black, and white). The soldered perfboard is embedded in a cardboard cutout that is painted to look like a forest creature sounds. Soon, this mixed media object will be joined with other embellished animal circuits to build a vignette of nocturnal musicians.

lesser angle-winged katydid

Watercolor and analog electronic study of a Lesser Angle-winged Katydid, 2018. I plugged in a couple of crickets to give this fellow some natural context. To read more about how I made this, visit my project “Hacking Nature’s Musicians” : https://hackaday.io/project/161443-hacking-natures-musicians

Follow me on Hackaday

I have started a project to describe and share my analog electronic circuits designs so that people can better understand my electronic practice. My project, located on Hackaday.io, is called “Hacking Nature’s Musicians.” You can find it here: https://hackaday.io/project/161443-hacking-natures-musicians

In a recent project log, I describe how I create chirping crickets using discrete transistors. Check it out: https://hackaday.io/project/161443-hacking-natures-musicians/log/153312-crickets-natures-favorite-astable-multivibrator

Owl surgery

Paper owl looking onto its electronic eye circuit, 2018

Transitions. Sometimes it feels like the holes that we have are equally -if not more- beautiful than what promises to fill them. Here, a paper owl contemplates the circuit that I designed to fit into its eyes. The addition of electronics to a static object adds more than functionality and aesthetics - it changes the identity of the object. Non-electronic things live in a physical world with thousands of years of creative history, while electronic things are very new. What was once an owl then become a robot - perhaps more robot than owl in our estimation. Does the owl stand to lose more than it gains?

Machine-centric intelligence

We struggle to relate to machines on their own terms, despite the fact that we created them. I suppose there are people with fluency in some machines, but the body of knowledge in computer science alone vastly exceeds anyone's capacity to understand. Moreover, there are subtle and often surprising effects that arise from even basic electronic components - instruments for manipulating electricity in ways that have yet to be discovered. 

However, if you add human features like the eyes in this video, suddenly we connect. But with what do we relate, really, besides our own reflection? We must push ourselves beyond human-centricity to see things for what they really are.

Shift registers

I am in the process of building a control system for the insects in my latest Electrolier. As a starting point, I will use shift registers with (or possibly without) linear feedback. Adding linear feedback involves "tapping" two or more of the logical outputs, evaluating their state with boolean logic (usually an XOR gate), and feeding the result back into the start of the shifting sequence. In my video, you can see two 8-bit shift registers in the center (TI part number CD54/74AC164E). I put LEDs on the logic outputs Q0-Q7 to visualize what is happening. As you can see, the LEDs of the two shift registers are not synchronized, a property of an undefined starting state that I will exploit to get randomness for "free" -- at least that's what I'm thinking.

More about my breadboard: on the right is a 555 timer in astable mode which provides a clock pulse of about 1x per second. On the left is a 5 volt regulator - this part is irrelevant if you have the correct DC supply voltage, but mine is 12 VDC.

Good general overview to answer the question “what is a linear feedback shift register” (LFSR):
https://zipcpu.com/dsp/2017/11/11/lfsr-example.html

On choosing taps for a linear feedback shift register:
https://cs.stackexchange.com/questions/1121/choosing-taps-for-linear-feedback-shift-register

Mini project on how to implement a LFSR: https://www.slideshare.net/KishoreChandrahasVanam/lfsr

Another helpful paper on using an 8-bit LFSR (i.e. Texas Instruments chip CD54/74AC164):
https://pdfs.semanticscholar.org/320a/8b2e781ac6165b400eca96047489685fd1f7.pdf

One final note: because shift registers are comprised of flip flops, an option to add true randomness to this otherwise pseudorandom bit generator is to exploit the metastability characteristics of a flip-flop as described here by @crj11: https://electronics.stackexchange.com/questions/394557/old-school-pseudorandom-pulse-generator-requesting-assistance-with-hardware-des?noredirect=1#comment965703_394557