The Boolean Sequencer eChucK Design Project

[Inventor]

Well by now you've heard something about this thing I evangelize around here called a Boolean Sequencer. I think they make cool music and being one among many who are exploring it's possibilities, I've checked it out some for you.

I've written programs that you can use to run your own boolean sequencers and done radio shows on simple Lunetta BS's that you can build yourself. And though I've explored some of it's possibilities, I've got to set the record strait as usual and say that I didn't come up with it myself. It's a re-invention, not an invention, because people have been doing BS's since the early days of synths.

In fact, one of the early simpler songs that I created sounded a lot like the theme from Bonanza of all things, and another work sounded very asian thematic, so the BS concept in music may well predate synthesizers altogether but I only guess at that.

Anyway, this thread is about making a BS circuit board for modular synthesizer use. If at all possible I'd like to rethink the design and make it more of a community effort. Truth be told, the best fun of any project is the imagineering of it so tell me what you want on the board and we'll figure it out together.

Les

[Uncle Krunkus]

Pray tell, Uncle Les, what makes a Boolean Sequencer a Boolean Sequencer?

[Uncle Krunkus]

The reason I ask is that I'm still not exactly sure what you mean by "Boolean Sequencer", so before we get too far down the track of what we want on the PCB, I think we should define what we need on the PCB.

BTW, I was thinking the other day about an idea which would most succinctly be described as a "Klee Gate". It's basically a set of gate sequencers of variable step lengths with interrelated reset arguments, and a number of gate outputs which are derived from the different sequencers being routed through AND, NAND, OR, NOR and XOR gates. The idea being that you could radically change the outputted gates by just changing one step in one sequence, etc, etc. Is that a Boolean Sequencer?

[Inventor]

Uncle, my Aussie friend, so good to hear from you. What you propose is dangerously close to the BS area, though I have not played with logic on the gate signals. We could discuss it and decides if it belongs in a BS which is this:

Logic hung off a binary counter and R2Red together.

That's the simplest way of saying it. You start with an LFO at twice the BPM of the song. You feed that into the clock input of a CD4040 binary counter which gives you 12 bits of available binary count, or in other words a 4096 step sequencer. Then you attach logic - and this is where your creativity makes your song - to the counter outputs. You then combine the logic outputs with an R2R or other resistive network to get a CV signal. You OR the output of all the gates to get a GATE signal.

It's an age-old way that I've sort of examined and explored and defined somewhat to create this way cool long-sequence sequencer that people are getting interested in. So that's what it is.

Les

[Uncle Krunkus]

So you could reset the binary counter after any number of steps, say, and with a certain array of gates hanging off those, feeding an R2R and then used as a CV, you'd get a particular sequence based on the array of gates.
I'd say sequence length would be No.1 on my list of features!
I suppose you'd automatically re-scale the R2R based on the number of steps driving it? Or would you just incorporate a CV gain/offset of sometype after a fixed R2R?

[Inventor]

Thanks Andy, yeah i gotta feature sequence length more - and it can be 16 bits for a 65,536 step sequencer or as many as you want. We know how to chain counters, lol

OK, to the ladder. I have two strategies for it. R2R is one of them and the other is a fixed sum of resistors. R2R works but the tones get subdued too fast as you go down in bits. fixed sum of resistors works good but has limited number of tonal outputs. So I propose variable sum of resistors.

Take four logic outputs: A, B, C, D. connect them each to a resistor of different size, tie the resistor outputs together and buffer with a voltage follower to create your CV. This combines them and also weights them in some gradual way.

There are surely other ways to do it but this is what I have found out so far in actual programming and testing.

Les

[Uncle Krunkus]

But the difference between one bit and the next is only restricted by the overall bit depth isn't it?
For example, if you only take two gate outputs and put it through an R2R of two bits, you could get 4 semitones as long as you scale the CV correctly, yeah?
With 6 gate outputs and a 6 bit R2R, scaling across 5.xx volts would give you slightly over 5 octaves with one semitone per bit.
Does that make sense?

[Inventor]

Two questions are what type of logic gates do you want to use and how are you going to make them connect to the data bus? I propose XOR and AND gates, with a big OR gate for the GATE. I also suggest using rotary switches for the XOR gates and DIP switches for the AND gates. Here's why:

I've found in practice that XOR gates give you a more continuous musical quality, while AND gates give you dramatic gaps in the song (sometimes too many). It's good to have both of them to choose from in your song.

The reason for the rotary switches on the XOR gates is to create a more active pattern so that individual bits are not wasted, and the reason for the DIP switches is because that's how I've gotten AND gates to work well in the past. This way there are two types of controls as well.

What do you think?

Les

[Inventor]

[Uncle Krunkus]

Sounds good.
I haven't even begun to think about the array of gates. I'm still stuck in R2R/scaled-CV-land!
Do you mean a rotary switch on the inputs to gates so you can decide which of, say, 12 counter outputs goes to each side of a gate? (the inputs are interchangeable, so a 6 way on each input would work just as well in that example)

[Uncle Krunkus]

That 6th bit would be a whole semitone if you scaled the R2R right. Makes all the difference in the world!

[Inventor]

[Uncle Krunkus]

I get the feeling we're not exactly on the same page yet. But it's good to throw ideas around.
I think I should do a drawing of what I'm thinking of, but now, alas, poor Krunkus is off to bed!

[Inventor]

[Inventor]

For this project there are also two implementation paths. One is a custom board, another is this general purpose board that dio is working on. He thinks he's got the bugs out and is ready for an application to put his board to good use. We could code up a BS and use his board as the physical implementation. Then we could add flute, KS, whatever to it...

Les

[electri-fire]

Les is sending me two KS boards, one with and one without the Boolean Sequencer.

I mentioned in the chatroom I'd probably be modding them, leading Les to exclaim: "Are you going to cicuitbend them Mathé? Well, I'll be sending them to you in working order."

Well, the point is, as much as I like the Boolean Sequencer, I think it is not as versatile as it could be, the bottleneck being the 4040 clock divider / binary counter being driven by a single clock. As I'd be using the KS as a Lunetta sound source, I'm toying with the idea to replace the 4040 outputs with external clock inputs.

This mod is easy to do and requires no alterations to the PCB I presume. But for those that want to do this in a standalone module a bit of room to place a Hex Schmitt Trigger with six timingcaps and pots might come in handy, either to make six onboard LFO's or to schmitt trigger external LFO's into Boolean compatible clocks.

Also (I mentioned this before, but as a reminder) , you may want to keep PCB modability and expansions in mind by inserting two extra solderpads between in's and outs of the various functions. The trace between these solderpads could then be cut for modular functions "within" the module, or ignored by those that use the board "as is" .

[Inventor]

Mathe, I will provide patch points for important locations like the data bus and the r2r outputs, good idea. Making the board circuit-bendable by design is a good idea IMHO.

Les

[droffset]

Following along but i have a headcold so might have to come back to it.
Les maybe you can put the schematics/code/etc here as well? I do remember there being various versions as this project has developed.

[Uncle Krunkus]

I realised today that there is no need to incorporate any sequence length options. As long as you can control which bits are patched through to the gates, you have immediate length control anyway.
For example;
If you want a sequence which repeats every 8 steps, you only patch the outputs from the 3 least significant bits. 32 steps long, use the 5 least significant bits. I spose some method of forcing a reset at non-binary counts would be good. Just an extra AND gate with 2 switchable inputs would cover most situations. Still, it's not needed for the basic BS.

[Inventor]

Amen Andy, same thoughts here on that. Folks, here is a pdf of what i'm talking about:

Les

---

BS11a.pdf
Description:	Boolean Sequencer getting started schematic	Download (listen)
Filename:	BS11a.pdf
Filesize:	19 KB
Downloaded:	949 Time(s)

---

[Inventor]

I did a preliminary board layout and only got the two 8-bit circuits on the board, as shown below in the pdf file. I still need to add a clock and the R2R ladder plus buffer for it. Already the board is starting to get full.

This leads me to the notion that this board can be offered as a bitslice board. If we design it so that multiple boards can be daisy-chained together then the end user can decide and configure however many BS slices they want in their system.

Aside from that we'd have to go with dio's board.

Les

---

BS11b.pdf
Description:	The BS Board Layout so far	Download (listen)
Filename:	BS11b.pdf
Filesize:	22.83 KB
Downloaded:	914 Time(s)

---

[Uncle Krunkus]

I'm thinking of just four two input gates across the counter for the CVs, but with 5 rotary switches so you can decide which bits are used for which inputs, and four three position switches which actually let you decide which type of gate is in each position. Does this justify the extra hardware or not? I'm not sure, but I'd like to check it out!

[Inventor]

[Inventor]

Here Andy, your idea done up in schematic form. Worked out great and fits on the board nicely with four chips and four position rotaries.

Les

---

BS12sch_a.pdf
Description:	BS12 schematic	Download (listen)
Filename:	BS12sch_a.pdf
Filesize:	20.13 KB
Downloaded:	929 Time(s)

---

[Uncle Krunkus]

Yep that's really close to what I meant.
I think it would increase the possibilities alot. Some combinations may be more useable than others, but we won't know that until someone has a bit of practice "playing" it.