I've come to the conclusion that if you spend enough time reading blogs and forums that are peripheral to your scope of interest, it's likely you'll get sucked in. I guess this means that you outta read forums where stuff is cheap, lest you develop a penchant for $800 bottle of wine. Lucky for me (?), my poison of choice leans to that of analog synthesizers, a hobby which is relatively easy on the wallet. Because I've been bitten by the DIY synthesizer bug, which has led to my current obsession to build my own analog sequencer.
These days, the word "sequencer" usually connotates computer software that, among other things, allows the user to record and play back an unlimited amount of MIDI notes... polyphonically and distributed over endless numbers of individual instrument tracks. But rewind the clock about thirty years, and "sequencer" typically implied a crude analog electronic device with a couple of rows of eight or so knobs, such as the legendary Moog 960 and the ARP Sequencer. Each knob set a pitch (or other parameter), and when the sequencer started playing, it would endlessly repeat an eight-note pattern, and monophonically at that. Doesn't sound like much fun, huh? Actually, "old-school" sequencers can do a number of neat (and musical) things if employed cleverly. You candial up a pattern, then control the transposition of the entire thing with a keyboard controller, bypass steps to create more rhythmic patterns or use the voltages to control other parameters like filter cutoff, or volume... the possibilities are pretty vast, especially when used in a modular analog synth setup.
Being an amateur electronic tinkerer and modular synth owner, I've become more and more interested in the analog sequencer idea. I should make clear that, besides the lack of computer and tiny "note-capacity", the other big difference with an analog sequencer is that it doesn't output MIDI at all. Instead, they output control voltages and gate signals... that's what folks used to control analog synthesizers before the digital dawn of MIDI. Almost all older analog synths have this kind of interfacing on the back panel, as well as most newer analog synths (like current Moogs, and any modular synth, for example).
The key to most pre-computer analog sequencers were 70's era 4000-series logic chips. These are a well established series of standard "building block" computer logic chips. These chips are to modern computer CPU chips as Ford Model T's are to a rocket ship, but they're still readily available, under a buck, and easy-to-understand. And you don't need to program 'em! (Google "4000 series" or "4000 logic" for lots of info on them). With a little online research, you can quickly find DIY electronics freaks who have built themselves analog sequencers that proverbially slice, dice and julienne with esoteric features such as skipping steps, forward and reverse modes, pattern randomization and other craziness. But I'm a simple guy who just wants to rock some Kraftwerk, y'know? As a result, I've settled on the net-popular "Baby 10" sequencer, originally conceived in THIS article. The Baby 10 (named because it's a basic 10-step sequencer) uses a 4017 logic chip with ten stages. In other words, each time you send a quick voltage pulse to the "clock" pin, it sends a voltage to the next "output" pin. You can use an LFO or a dedicated "clock" chip (the 555 timer chip is the standard for making simple clocking circuits) to send a steady stream of voltage pulses. If you take the all those outputs and hook each one up to its own potentiometer (aka "pot") so you can set the voltage level at each stage, then sum all the pot outputs, you now have a basic sequencer. Since analog synths typically have a CV input for pitch, and a gate input for triggering envelopes, it makes sense to split the output of each stage and bus that to a jack for a gate out and send the other half of the split to the pot for a variable voltage for pitch setting (it's a good idea to use diodes on all outs as well; these let the voltage travel one way only, preventing all these mixed outs from fighting each other). Expanding on the idea little more, adding on/off toggle switches before each of the gate outputs makes sense, because this lets you turn steps off. Now instead of a never-ending stream of 16th notes, you can make far more musical patterns (well, an endlessly repeating musical pattern, that is).
One more thing; you may have noticed that our ten-stage chip implies ten steps, and ten steps doesn't work too well when the music is in 4/4 time signature (or 3/4 for you waltzing and swingin' types). Obviously, four or eight steps would be more suitable for "common" time music. All you need to do is wire the pin after the last step (step 8, for example) to the "reset" pin on the 4017 chip and you can use as many steps as you like. If you build your sequencer with independent output jacks for each stage and an input jack for the reset pin, you can use a patch cable on any step's gate out running back to the reset jack. Otherwise, you could install a rotary selector switch pre-wired to send various steps back to the reset pin. If you wanted to really keep it simple, you could build the sequencer for 8-steps only and hardwire the ninth stage to the reset pin (or really weird out your friends and hardwire it for an odd number of stages and claim it was designed for esoteric odd-time signatures only).
Finally, since the CV out pots generate a continuously-variable voltage, this can make it tough to set pitches accurately if you want the notes to conform to standard half-step temperament (i.e. diatonic scales). This can be remedied by sending the summer CV out to a "voltage quantizer" that will automatically adjust the the voltages correctly to discrete steps for even-tempered intervals on the fly. These are a little tricky to build from scratch, but the coolest solution I've found is a DIY printed-circuit board with build instructions from Ray Wilson's www.musicfromouterspace.com website. The board is affordable, and you could build it right into your sequencer circuit, and the site gives full instructions and a parts list. Thanks Ray!
I could go on and on with possibilities... this is the the beauty of DIY. And also the downfall, as you can easily get carried away. You can build this kind of thing on simple "experimenter board" generic circuit boards by hand wiring them, but I recommend getting a "breadboard" and trying it all out before you crank up the soldering iron. The neat part is that parts are pretty cheap... 4017 chips are 30 cents from futurlec.com! Oddly enough, the circuitry is dirt cheap, but it's the knobs, pots, switches and enclosures that'll add up real quick. It all depends on how crazy you want to go making it look snazzy. In the meantime, have fun... your friend the internet will tell you much much more about DIY sequencer fun (www.electro-music.com is a great DIY synth resource).
Post by Mitchell Sigman