2 – ‘Blue Eye’ Synthesizer One (B.E.1 Synth)

After completing the extremely easy to construct Quick and Dirty DIY Studio Headphones, I was left feeling somewhat satisfied with my handiwork, and began to crave another project. Maybe one that couldn’t be done by a trained chimpanzee. Maybe not.

Either way, I started scouring the net for DIY  projects, and found a surprising number of ‘projects’ that were so incredibly simple and pointless as to make my basic headphone modifications look like they were straight from the notebooks of DaVinci himself. The flip side of that, is, of course, that there are also a staggeringly large number of projects that, at least at first glance, appear to require a spiffy lab coat and masters in electrical engineering (the lab coat’s really the key item in that pairing, I think). The trick became to find something that could be put to some use when completed, and that I, with my pathological abhorrence of mathematics and neophyte status in the crafting of electronics, could tackle without, say… exploding my face or setting fire to a cat.

What I landed on was this: a page that detailed, start to (sort of) finish, exactly what needed to be done to create this bleep-blooping-questionably-useful-square-wave-generating contraption. We’re a band that’s somewhat fond of strange noises (without venturing too much into Mars Volta territory, although they -along with Radiohead, Massive Attack, and Portishead- are perfect to listen to while staying up all damned night, building such a device as this), so I figured with some modifications here, some tweaking there, this could end up being something we’d actually use. Worst case scenario: I was out about $25, and gained whatever experience would be gleaned from the construction of such a device. While researching parts suppliers, I stumbled across Tanner Electronics, a great Carrollton-area family owned small business that has a pretty solid inventory of electronic parts.

Materials:

  • SPST (single pull, single throw) toggle switch
  • momentary push button, normally closed
  • momentary push button, normally open
  • three photocells (I don’t remember what values I used, and haven’t bothered to go back and test them. Play around with values near 100k, it’s entirely up to you.)
  • LED holders, to mount the photocells on the outside of the enclosure
  • soft rubber grommets (optional, to cover the LED holders and protect the photocells while diffusing the light coming into them)
  • 9v battery clip/battery
  • 12V blue indicator lamp (running at a max of 5V, it still shines brightly enough to serve the purpose. Use an LED if you like, just remember to resistor the anode appropriately so it doesn’t burn up.)
  • 100k linear potentiometer
  • 1M linear potentiometer
  • 40106 Hex Schmitt Trigger/Hex Inverter (I didn’t use an IC socket on this, which I now realize is pretty much inexcusable, since they cost all of $0.30/ea and prevent you from damaging the chip when soldering while allowing you to easily replace the IC later down the road.)
  • IRF520 MOSFET Transistor
  • 7805 5V voltage regulator
  • .01 uF capacitor
  • 1.0 uF capacitor
  • 3.5mm audio jack
  • project box
  • circuitboard

Tanner provided all the components listed in the bill of materials, inexpensively enough that I picked up multiple copies of each, to counteract the fact that I fully expected to accidentally fry the hell out of at least one item. While I was there, I snagged various items that seemed as though they might be useful, and a plastic project box roughly the size of a guitar pedal to hold the electro-screamer-whatzit when it was done. All totaled, I got out of there to the tune of about $35, and amassed enough parts to make a couple of the devices.

To begin with, I set about prototyping everything on breadboard and followed the directions to the letter, receiving similar results to those presented on the website. There was some variance due to my not getting the exact parts listed, but it still worked quite well, and nothing caught fire (added bonus).

From there, I added some photocells and played around with the wiring a bit.

I left the actual circuit alone on the breadboard for a few days, while I decided how best to mount everything in the project box. I was also still pretty wary of beginning my first foray into circuit board soldering, so I held off until I was a bit more anxious to get the thing done. In the meantime, I figured out how I was going to arrange the components on the board, and realized that the board itself could be much smaller and still fit everything. I marked a line at the edge of the area I needed, then scored it with a knife and made a clean break by holding it against a straightedge, which worked quite well, and left me with a narrow piece of leftover protoboard to use later in another project.

With most panel mount components in place, starting to populate the protoboard.

Once I couldn’t wait anymore to get it finished, I broke out the dremel tool and started putting panel mount components on the box. Quick tip #1: own a garage (or just… try not to be an idiot). We’re apartment dwellers, and I didn’t relish the idea of drilling outside on the iced-over patio, so I began the dremel-ing in the living room, where I had to quickly improvise a solution to the “aw-shiiiit-there-are-tiny-flecks-of-plastic-flying-frigging-EVERYWHERE”  issue that immediately arose. In hindsight, that really should have been obvious before I started, but… oh, well. At any rate, I ended up holding the enclosure in a plastic garbage bag while I drilled out the holes, and it sufficed to prevent the ensuing shrapnel from embedding itself in the carpet.

Yes; "nut" and "shaft." Stop giggling.

That bronze nut at the base of the shaft? Fairly important.

Quick tip #2: make sure to get the proper mounting hardware for components. Again, this should be fairly obvious to everyone (except, apparently, me). I ended up purchasing a handful of potentiometers, but got out of the store without any mounting nuts (which are free, clearly labeled, and in a bin on the same shelves as the pots). Not having any prior experience with electronics, I assumed that the screw threading on the potentiometer shaft was how it mounts to the enclosure; so I carefully drilled out a hole that was just the right size, and screwed the pots into place then hot-glued them down so they wouldn’t turn. This was just one of many things (as it turns out) that I did wrong, but in a way that ended up working. This process also took entirely too long, since I couldn’t find our power drill, and had to manually bore out the holes to the right size using the tiny dremel drill bit, in between recharging the battery, which leads me to Quick tip #3: know where your power tools are before trying to use them, or just suck it up and ask your wife where she’s hidden placed them.

The power switch, indicator lamp, and one potentiometer mounted in the enclosure.

After mounting the pots, photocells, the switch, and the power indicator lamp, I clipped everything from the enclosure into the breadboarded circuit to test it. Everything worked and nothing exploded or burned, so it was finally time to move on to the circuit board.

This is pretty much a 'before' picture of bad soldering. Oh, and it's wired wrong, but I'll get to that later.

Note: If you haven’t soldered before, you need to know to be careful of solder fumes. Make sure you’re in a well-ventilated area, and have the airflow manipulated so that the fumes blow away from your face. I use a small case fan attached to a 9V battery and mounted on my soldering station.

I’d already arranged the components on the board, and bent the leads to hold them in place, so it was just a matter of wiring and soldering them into place. This is when I discovered Quick tip #4: they sell ‘helping hands’ for a damned good reason. In the interim between breadboarding and soldering, I had dismantled an old CD-ROM drive, for a variety of reasons, and ended up using part of the drive case as a soldering station, with alligator clips attached to the internal mounts. Since then, I’ve re-jiggered it a bit, and it works fairly well (if I do say so myself), but at the time, it was extremely frustrating and time-consuming to solder anything, anywhere, especially when a small PanaVise and set of helping hands will set you back less than $30, and are infinitely more useful.

As it turns out, years of not soldering anything, ever, doesn’t really prepare you to be a wizard at soldering. Strange, but true. So, I ended up with a few decent solders and a ton of off-colored, bulbous mounds of solder, with flux having run everywhere from re-working the solder joints. I went back through, removed some particularly heinous connections with desoldering braid, and redid them in a slightly more acceptable manner, but was still left with a bright, shiny board covered in flux. I didn’t have any solvent or appropriate cleaning alcohol, so I scrubbed off what I could, and ended up scratching dividing lines into the circuit board with a knife to make sure that the conductive flux wasn’t shorting any leads. Again, I’m aware that it’s not really an approved or endorsed “correct” methodology, but I didn’t know better, and it worked, so shut it.

Not exactly a pro job, but for a first project it was coming together nicely. Except for the ugly soldering.

Once I was done clearing flux, I checked the leads with a multimeter to make sure that the current would flow properly, and that none of my solders were bad too bad.

Now, the project box I used did have mounts into which to screw the circuit board, but only if the board took up the entire inside area, which mine didn’t. So I took part of an old plastic ballpoint pen shaft, cut it to the right length to act as a standoff mount, then screwed it into a corner of the board and screwed the opposite corner into one of the enclosure mounts. Now, it was held firmly in place in one corner, and protected from any possible stress by my makeshift ‘leg.’ This worked out well, because the component side of the board faces into the enclosure, nestled gently amongst a rat’s nest of thick wires (in my zeal not to have any wires too short, I may have left them all too long). The bottom side of the board, with all of the soldering, faced out to the screw-on bottom panel of the enclosure, which turned out to be conductive, so I put a thin sheet of insulating packing foam between the board and the panel to prevent any shorts.

By this point, it was dangerously close to sunrise, and my back was killing me from hunching over a makeshift workstation for several hours without moving or stretching. Which is probably why, in all of the double, triple, and quadruple checking of the circuit that I did, I was absolutely certain that I’d done everything correctly. An assertion that wasn’t shot down when I closed it all up, plugged it into some speakers, powered it on and was rewarded with a blast of noise. It worked!

Note the timestamp on the image title. I was a bit delirious by this point.

The first time anyone picks it up, within thirty seconds of playing with it they all grab the power indicator lamp (the glowing thing at left) and start furiously trying to twist and pull at it, to the extent that I have to tell people not to do it before I hand it to them.

Functionality:

  • The knobs control the pitch of the two oscillators. Since the potentiometers have completely different values, one acts as an LFO (low frequency oscillator) and the other is much higher frequency, ramping the sound up beyond audible (at least to my rock-drummer-deafened-ears) levels. Since they’re tied in together through the transistor, the result is more like one highly variable oscillator than two distinct tones.

  • One photocell (the one by the output jack) acts as volume control. Exposure to light increases the volume, while darkness almost silences it. This can be manipulated by exposing it to a controlled flashing light in order to create accent notes, or covering it with a hand and moving carefully to crescendo or decrescendo.

  • The other photocell serves a dual function. When the light to it is diminished, it creates a high resistance, and limits the overall power that flows through the circuit, which drops the output frequencies. This effect can be manipulated similarly to the other photocell; by applying a controlled flashing light to create a pulsing tempo, or ‘playing’ it with a hand to control the light that strikes it. The other effect of this photocell is that, since it reacts differently to various wavelengths of light, controlling the color of light that hits it will modify the tone of the sound.

In later days, I’ve gone back and made some modifications, including replacing knobs, covering the metal on the power switch with heat-shrink tubing, hiding the 9v clip mounting screw with a black grommet, and adding:

  • a momentary kill switch (large red button, near the screw), to allow the user to cut the output when desired. It’s faster and easier than toggling the power off and back on with the main switch.

and

  • a photocell ‘bend‘ of the original circuit, with a momentary button to enable the bend (the button is directly below the killswitch in the picture, with the photocell next to it, in the hot glue. Yes, I’m going to clean that up.). When activated, it raises the pitch and alters the sound of the tone created, while the amount of light striking the photocell adjusts the pulsing frequency, from a solid note in darkness, to a very lethargic pop in full light.

Since then, it’s been used extensively by several people, and it’s functional, stable, and consistent in its operation. When I decided to write this, I started drawing a schematic of how it’s actually wired up, since I deviated from the plans, and made modifications. This is when I realized that, as far as I understand the basics of electronics (which, admittedly, isn’t saying a whole hell of a lot), this circuit probably shouldn’t work.

Something I probably should have consulted a bit more carefully during my marathon build session.

More on all of this later. Consider this under construction, but for the time being, here’s the schematic (below), based on my tracing all of the wiring on the actual unit. See pin 14 on the 40106? According to the 40106 IC datasheet, that’s a power lead, so it’s sort of important, yet I don’t have it connected to anything.

But I do have power going into pins 8 and 13. 8 is an input, with nothing coming from 9 (the corresponding output), so that should have no effect on the overall sound, and could be removed entirely. 13 is an output, with nothing going into the corresponding input, so as far as I can tell, that should do nothing, but it does. [Edit: I’ve since learned that when you don’t resistor out the unused IC leads to ground, they oscillate on their own, which is considered poor design since it pulls extra power and makes the circuit less efficient. So, as I now understand it, connecting output pin 13 to the power is actually feeding an oscillator’s output through the entire circuit, creating a feedback loop, which also explains the effects it creates.

The LM7805 voltage regulator, as it’s soldered into the circuit, has pins two and three backwards, so the output pin is effectively connected to nothing, while the ground pin is wired up to things that should be connected to the output.

The whole thing works, and nothing overheats or reacts erratically. Still, though, the lesson learned is this: sleep is good. Don’t spend eleven hours straight working on something and expect it to be perfect when you finish at sunrise. Always make a schematic of what you’re building before you do it, double check the damned thing, and follow it exactly.

I'm planning on cleaning up the schematic and making corrections. When I do, I'll build another unit and post the results and new schematic.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
%d bloggers like this: