SRFF Oscillator

In Uncategorized on August 2, 2011 at 6:38 am

Here’s an interesting little circuit for your entertainment.  It’s another one of my so-called reInventions, something that I dreamed up and tested working however it most likely already exists in the prior art.  I’d like to begin by sharing the creative process with you and then explain how it works.

I was participating in the Lunetta Challenge in which we each purchase a grab bag of 4000 series CMOS chips from Electronics Goldmine and then create whatever fun Lunetta circuit we can from the chips provided – and only those chips.

The first thing I needed was an oscillator and I was given no Schmidt trigger parts so I could not use a classic single-gate oscillator.  I could have looked up an existing circuit but that is not my way, I’d rather imagineer something if only for the fun of enjoying the creative process.

My thoughts turned to the familiar Set-Rest Flip-Flop.  Formed from either NAND gates or NOR gates in a cross-coupled configuration, an SRFF is kind of like a two-input pulsed inverter thingie in a way.  Applying a pulse to the Set input gives you a 1 output on Q and a 0 output on Q-, then applying a pulse to the Reset input toggles the outputs.  So we have two pulsed inputs and two latched outputs that invert in response to the inputs.

Going with the notion of the SRFF being a sort of dual inverter circuit, I wondered how I could make this thing oscillate.  The critical elements of this type of oscillator are an inverting device and an RC feedback circuit, and here we have a dual thing, so maybe I could use two RC filters in feedback to do the job.

Well there were two open inputs so I could see where to feed back the RC circuits, and just sort of graphically I imagined the input connections to the RC circuits from the Q and Q- outputs, as shown in the schematic above.  I wired up the circuit on a breadboard and surprise – it worked!

Actually the potentiometer in the circuit above is a later addition which provides a limited degree of frequency control.  For a fixed frequency oscillator just use a resistor there of equal value to the other resistor.  For full and symmetrical frequency control, use a dual ganged potentiometer.

The circuit has some interesting properties which you may find useful sometime.  It provides a clock and the inversion of that clock simultaneously, it offers duty cycle control by adjustment of the RC time constants, and it uses only half of a 4011 NAND chip, providing two free gates for other purposes.  Also the circuit works with 4001 NOR gate chips as well.

I guess that’s about all I can say about this interesting little CMOS creation.  You may choose to add it to your arsenal of music making circuitry or just enjoy reading this article for whatever it’s worth.

The most important thing I hope to leave you with is another example of the creative process – imagineering circuits like this is really kind of like an art form and be you an expert or a beginner, it’s something that you can do yourself.  So go forth and create!  It’s fun and surprisingly satisfying so I say go for it!


Guitar reInvention

In Uncategorized on July 25, 2011 at 8:03 pm

I’d like to share another reInvention created by myself and my buddy Are.  We imagineered this thing from scratch, built and tested it working properly on a real guitar, then of course discovered patents on the basic concept going all the way back to the 1930’s!  So it was one of the original ideas for an electric guitar pickup, but it was ahead of it’s time.

It took until the end of 2008 for us to reinvent it, and by then technology had evolved to the point that it became practical.  So practical in fact that I have the guitar in working condition sitting next to me right now as i type.  Although I say that the concept originated in the 1930’s – and also there were many other designs proposed over the decades – we put a little design twist on it that made it practical for use on modern guitars and other stringed instruments.

So enough about the origins of this reInvention, let’s talk about how it works.  It all starts when we put strong neodymium supermagnets in an alternating pattern under the strings.  I cut up a cookie tin to make thin little plates which i secured to the surface of the pick guard with double sided foam tape, right underneath the strings.  Then I stuck magnets under the strings.  See the photo below for a close-up of this arrangement.

Those are 9mm square magnets and the strings are 10mm apart, so the fit is close enough.  Note that the strips of magnets are snapped together as they run across the strings because the magnets are alternated like this:  NSNSNS.  That way the magnetic fields add up beneath each string instead of fighting with each other.  This is one of the keys to making this pickup system work.

What happens when you strum the guitar strings is that they vibrate perpendicularly to the magnetic field that they are located within.  According to physics, this creates a tiny electrical current that flows along the length of the wires.  These currents travel to the headstock where they end up at the tuning pegs and there they need a place to go so that we can collect them, boost them, and send them out to the guitar amp.  That’s where the following photo gets important.

Here I have shorted the tuning pegs together with lugs that were laboriously fastened out of tuna can lids with a Dremel tool, which tool forever and a half!  Once the lugs were created and soldered together carefully they formed a shorting block that electrically sums all six string currents together into a single combined current.

“Ah,” you say, “but how do you get that current back down the neck where it can be boosted and sent out the guitar jack?”  The simple answer is you use the truss rod inside the guitar neck as the return conductor.  That part is actually one of the main contributions to the state of the art that Are and I made.  Prior art in the patent record typically required an insulated bridge which is not standard equipment on guitars, so it is not manufacturable easily.  Our approach maintains the bridge at ground potential so no changes are necessary to the bridge.

So anyway, our tiny little electromagnetically generated currents have now gone full circle from under the strings, up the neck, down the neck, and back into the guitar body.  At this point we could just send the signal out to the guitar jack and put a gain pedal between the guitar and the amp to get our signal.  However that means super tiny currents traveling in the guitar cable which leads to noise issues.  So the solution is to boost the signal inside the guitar.

In our prototype we used a special microphone transformer with a mu-metal enclosure to accomplish the signal boosting.  The result is a signal strength about half as strong as a conventional pickup produces which is close enough for our purposes.  However we have since realized that an opamp gain circuit or similar would be the ideal choice here.

That means putting a battery in the guitar or otherwise delivering power to the guitar through the guitar cable or via solar cells with storage capacitors for playing in dark environments or other such complications.  This turns out to be worth the effort for the improvement in audio quality.

So how does it sound?  Clean, in the sense that it sounds like an acoustic guitar.  In tests with guitar players, I found that pretty much none of them like this sound, haha.  They are accustomed to the distortions created by the second order low pass filter that exists in conventional wound pickups and they don’t like the clean sound.  I tried many times to explain to them that we could introduce the characteristics that they preferred with additional circuitry and even make that adjustable without having to swap out pickups, but got zero positive response.  Not one guitar player likes this pickup system.

However, I feel that it’s great for certain styles of guitar play such as bluegrass or classical where the acoustic sound is preferred.  Also it would be useful for making an electric violin that was true to the violin sound instead of distoring it like conventional pickups.  It’s also great for any metal stringed instrument including custom instruments that people build for academic, industry, or hobby purposes.  So the guitarists do not have the final say after all.

One last thing I’ll mention is that this system should also be nice for hexaphonic guitar output where you capture all six string vibrations separately and send them out on a multiconductor cable to the amplifying equipment. That can be accomplished by not using the shorting lugs and simply sending six return wires down the neck in a separately machined channel in the neck.

So there you have it, from the 1930’s to today, about 80 years later an idea gets reinvented and made practical by the advancement of technology a a couple of hobbyist fiddling around.  Another reinvention bites the dust, so to speak.  Well, i should stay positive – perhaps some one will find this reinvention useful and actually tell me about it.  Enjoy your day.

Boolean Sequencer Basics

In Uncategorized on July 25, 2011 at 3:27 pm

There are many kinds of sequencers in the world of music creation, each with their own unique set of characteristics.  When I started creating music, however, I knew nothing at all about them not even what a sequencer was.  I could make lots of sounds with ChucK programs and make random notes and such, but I didn’t even know what the word sequencer meant.  All I knew was I wanted to make songs from these sounds.

To this day four years later I still have a limited understanding of all the fascinating variety of sequencers out there but there is one that I know very well:  the Boolean Sequencer.  That’s the name I chose for a technique that I later realized is as old as synthesizers themselves, maybe older.  Possibly a great deal older, historically, but let’s stick to modern day electronics versions.

So not knowing what to do, I got creative.  I remembered Boolean algebra and truth tables and Karnaugh maps and all that jazz from college, and I also remembered programming FPGAs with the Verilog language.  I imagined that there might be some way to take a logic truth table and step through it with a counter, producing a series of on and off signals.  This could tell me when to play notes.

I tried it and it worked, yay \o/ I was making songs – sorta kinda.  But these were one note songs but they had interesting rhythms, patterns of notes that caught my attention.  I was on to something good it seemed.  Then I wondered:  what if that series of ones and zeroes was actually a series of note frequencies?  How could I generate that?

Well long story short, I ended up with a very simple combined digital and analog architecture that I named the Boolean Sequencer, comically producing the acronym BS as a coincidental joke.  That thing is pure BS, I laughed about it!  Well the only real BS was that I thought I was inventing something when I was actually reinventing it, more on that later.

The architecture ended up as follows:  Tempo clock to binary counter to logic network to aggregation network, and out pops a Control Voltage (CV) that can be applied to any music making circuit that responds to an input voltage.  I played around with this BS thingie and learned a lot about it and I will share some of those findings with you shortly, but first let’s examine the stages:

The tempo clock is just an oscillator that produces a digital output.  If you want six notes per second you set the tempo clock frequency to 12 Hz.  Why 12 Hz instead of 6 Hz?  Because the clock will be fed into a binary counter which will divide it by two with the LSB flip flop.  Alternatively if you actually use the clock itself as the LSB you can make it be 6Hz for 6 notes per second timing.

The counter is usually one of the binary counters CD4020, CD4040, or CD4060.  These all count out a binary sequence and make most or all of the bits available on the output pins.  So we clock this counter and this produces a binary count N bits wide.

Next up we have the logic network which accepts the counter output bits as input.  It may be any logic function of the available bits from the counter, and may be any number of bits wide at it’s output.

Finally there is the aggregation network which is often just a set of resistors connected to the logic outputs such as a multiple input voltage divider or an R2R ladder.  I call it an aggregation network because it can be very aggravating!  Just joking!  Actually it aggregates the digital outputs into an analog signal, the CV.

Got all that?  Now let’s discuss some of the properties of Boolean Sequencers.  One is that they can have any sequence length that you like but normally you just set them for a binary whole number of steps like 8 or 256 or 65536 or whatever.  65536 steps?  Holy synthesizers Batman, that’s a lot of steps for a sequencer to have before repeating!  Yep, due to how quickly the power of 2 grows as you add bits, you can get amazingly long sequences.

And those sequences will be completely unpredictable!  Well actually they are very predictable according to boolean algebra and analog circuit theory, but the tendency is not to design for a specific sequence but rather tho just haphazardly hook up logic and aggregation networks until you like what you hear.  You can create by design or by exploration, that’s up to you.

I can say this, however:  AND and NOR gates tend to create sequences with gaps of silence especially at the beginning of the song and get far more active later in the song.  It’s the opposite for NAND and NOR gates, and XOR and XNOR gates tend to be active all the time, with no gaps at all or very few.

One last note about the characteristics of a BS is that it generates fascinating patterns within patterns within patterns.  Just stare at the counter part of the logic netowrk’s truth table (the binary count part) and your eyes will glaze over as you see all these fractal patterns nested at every level of hierarchy.

Well, the same is true of the logic output only in a more entertaining way.  As you listen to the song your mind will latch onto a phrase (pattern) and then hear it again a few times perhaps then again but a little bit different twice, then back to the first one just once, then another new one, and so it goes throughout the song.

I’ll finish up where I started and just mention that the Boolean Sequencer is not really a completely original idea.  In fact it has been created many times before as I have noticed diode arrays in step sequencers and other circuits that effectively implement a BS, and also I have heard what sounds like familiar old tunes in some of the BS programs that I have written.

In fact, I wouldn’t be surprised if the Boolean Sequencer was known well before electronic synthesizers even existed, maybe even centuries old dating back to ancient culture in woven patterns, celtic knots, or tiled art – who knows where binary combinations originally exited?  But that part is strictly conjecture.

Thank you for reading about all this BS stuff, heh, and I hope you enjoy listening to some Boolean Sequenced music or even make some of your own.