Synth DIY Stuff by (d)

This mini-blog is hosted on tumblr and serves as a quick way to document my synthesizer-based DIY projects. Follow me for updates on this DIY-type stuff and my music: Main site: http://glitched.org/ Twitter (as d01): http://twitter.com/do1 Soundcloud: http://soundcloud.com/d01

[glitched-diy]

SSSR Kotelnikov Osc

The Russian-developed Kotelnikov oscillator is a eurorack-format module, available as a fully-built module or as a DIY kit.  I opted for the kit because, well, why the heck else would I be talking about it on this site?

https://sssrlabs.com/en/store/sm042_kit/

At $200 USD, it’s not exactly a cheap kit, but I had a few bucks in my paypal from a return of a module that didn’t work out, so I said, why not.  Oh, also, I have a bit of an obsession with digital oscillators.  Having built 2 DIY Orgone Accumulators and owning one retail version, yeah...I’m in there.

The demos were enticing, but what was especially interesting was the integration with the on-board analog modulator (”transient-generator”).  It’s used to great effect in this video:

So, on to the build...

DAY1:  Tragedy Strikes!

After printing out the BOM and neatly arranging the components by size, value, weight, and style (nahhhh), I started with the surface-mount chips first, as suggested by the build-guide.

The regulator was no problem, thanks it its large V-OUT pad.  Next, the precision voltage-reference, REF5050.  I tacked one leg down, but it wasn’t straight, so I tried to adjust it and...

...Whoops, the darn pad came off!  Frantic that I had just tripped on the starting-line, I sped to my computer to email support.  Meanwhile, I combed the data-sheet and wouldn’t ya know it?  Pin1 is “DNC” (do not connect).  Well, ok then!  Tragedy averted.

Lesson learned: always read the data-sheets prior to freaking out.  Data-sheets are essential in troubleshooting and for sanity-checks.

DAY2:  Nothing to See Here

Day 2 was when most of the tedious soldering was done:  sockets, resistors, diodes.

DAY3:  Too Easy

Ok, maybe “easy,” isn’t the right word.  Soldering over 500 through-hole points is a marathon.  

Without any major issues, “all” I had to do was solder about a million bypass caps, a few pots/jacks, and switches.  I was nearly done.  Here is the Kotel at 99%:

The next day, I threw the rest of the chips in and stuck the 2 PCBs together with some tape and spit.  Using the headers on the boards was a better idea, so I did that.

There’s always a moment prior to power-up where, despite one’s confidence in his abilities, there’s a chance that it might not work.  I was especially concerned because of that solder-pad detaching from the PCB. 

Well, wouldn’t ya know it?  It worked on the first try!

The heart beats. Have to figure out a minor issue with one LED not lighting, then knobs go on. Blue? pic.twitter.com/Jw6vRqT4hM

By the way, the source of the LED issue was, uh, I kinda crammed one of the ICs into the sockets too hard and bent one of the legs.  

Conclusion

In spite of my early hurdles, this turned out to be a paint-by-the-numbers build.  No real surprises.  

As for the module itself, it’s a fairly-basic wavetable osc.  The fun stuff comes in when you mess around with the built-in transient generator in LFO mode at audio-rates.  

I would say, if I didn’t have 2 Orgones, this is an excellent digital oscillator to get.  Unfortunately, the OA offers what the Kotel has and much, much more.  Still, it makes it no less useful.  

Now, more pics:

(Russian IC!)

(This is the soldering-iron tip I have used for the past couple builds.  See, you don’t need the best, or even passable tools for synth-DIY!)

[glitched-diy]

Orgone Accumulator Retail

I never got to posting what the retail Orgone Accumulator package looks like, along with the manual!

Thanks to all involved: Kris Northern (Magpie), Jim Matheson (http://neutron-sound.com/), WMD, etc.!

[glitched-diy]

Orgone Accumulator: The Retail Manual!

By now, you must have heard that the Orgone Accumulator is getting a retail release!  If you haven’t, well, THE ORGONE ACCUMULATOR IS GETTING A RETAIL RELEASE!  Awesome!

What might be just as awesome is that I did the manual for this module!  See it, here:  https://drive.google.com/open?id=0B32_T8r1yEdtVWxTRFliSjVsMkk

Cool, right?!

I can’t wait to see the finished module and manual, in the flesh.

I’m also happy that those who were too trepidatious to get into SDIY will finally experience the most exciting module in Eurorack!

[glitched-diy]

Orgone Accumulator: Custom Wavetables

UPDATE: 10/20/2015 As with most tech-things, this article might become obsolete, or at least become the hard way to do things.  With the release of Orgone 2.0, came a visual waveform transfer-tool.  This will no doubt be handy for the non-programmers out there.  Take a read of the developments:  https://www.muffwiggler.com/forum/viewtopic.php?p=1912972#1912972

Anyway, please enjoy this article.  At the very least, doing this procedure will help you get comfortable with the Arduino programming environment.

So, if you’re like me, the first thing you wondered was how to update the Orgone Accumulator firmware with your own wave-shapes.  In my experience with wavetables, there’s a ton of variety in the tones you can get from single-cycle, looping waves.  In this case, however, I hesitated putting in the work to create an entirely new set of waves because the stock ones are so good and varied.  Still, there were a couple sounds (guitar, bass) that bothered me, so I decided to dive into the firmware anyway.  This guide is for those who want to update the OA firmware with their own single-cycle waveforms from the Adventure Kid waveform library.

Prerequisites

You have installed the Arduino IDE and your computer recognizes the Teensy MCU.

You have downloaded the official Orgone Accumulator firmware and extracted the .zip file. 

You are comfortable with editing a .ino (text) file, compiling it and uploading it to the Teensy.

For information on these steps, download and review the Orgone Accumulator build-manual (around page-4).

NOTE: I can’t be held responsible for anything that happens to you or your Orgone Accumulator while performing these actions.  It is assumed you have a working build and have followed the build-manual’s instructions properly.  These instructions are only valid for v.1.1-ish of the firmware.  They may become redundant or outmoded.  

Finding Interesting Waves

As you may have discovered, the stock waves on the OA come from a collection by “Adventure Kid.”  The page catalogs something like 4,000 distinct (and some not so distinct) single-cycle waveforms.   But--and here’s the catch--the Adventure Kid site only has pictures of the waves.    

The actual data correlating to each waveform is included in the OA firmware .zip (”AKWF_converted” folder, AKWF_signed_int_512.zip), stored in text files. 

How does one audition the waves and pick his favorites?  Well, one way is to look through the thousands of .gifs and cherry-pick the coolest-looking ones.  Unfortunately, the “coolest-looking” ones don’t always sound so great or they sound like a wave that’s already in the firmware.

I tried that and was disappointed, despite the jagged, glitchy, meandering wave-shapes.  You have to audition them.

💿 The first step is to download the actual .wav files from the AK website: http://www.adventurekid.se/AKRTfiles/AKWF/AKWF.zip .  

Extract that .zip to a folder on your hard-drive, maybe the same place you’re putting all your Orgone-related stuff.  

But how to actually audition the waves?  Enter FL Studio’s Harmless plugin.  (Don’t have FL Studio?  How about you download the demo: http://www.image-line.com/downloads/flstudiodownload.html )

In addition to being an awesome, synthesist’s dream of a soft-synth, it accepts...you guessed it...single-cycle waveforms as input!  

💿 Browse to the folder that stores the AKWF waveforms, on the left.  Click on the file and drag it on to the Harmless plugin window.  

💿 Play around with your MIDI keyboard and type (or write?) down the filename of those files.  There are 32 total waves included in the OA stock firmware, but there are some very useful, basic waveforms which I would not replace.  The best way to start off is to think of replacing only the waves you want to replace.

Trust me, this method is lot easier (and more fun) than simply guessing which waves sound good.

By the way, of course you don’t have to use Harmless/FL Studio.  There are many other soft-samplers out there that will do the job.

Editing the Firmware

So, you’ve got a list of cool-sounding waves.  Great.  Now the somewhat tedious part: copying and pasting the data that correlates to the replacement waveforms.

Where is this data?  As I mentioned before, it’s included in the OA firmware zip file, the one that you grab from Github.  (Yes, you have to download the whole repository.)

💿 Find the AKWF_signed_int_512.zip file and extract it.

Look at all the folders!  Directories as far as they eye can see!  Recognize the names?  They’re the same as the ones from the AKWF .wav collection.

💿 Now, you could navigate through the directories, searching for the waveforms you jotted down earlier, but that takes forever.  Just do a Windows (or whatever OS) search and open each .txt file as you find it.  Yes, have them all open.  You’ll see content like this:  

Editing the Firmware (for reals this time)

Before performing the crucial, copy+paste operation, follow these steps:

Open the Arduino IDE and open the orgone_accumulator_official.ino file.

Save this file to a new name in a new directory.

Make sure that you’re editing the file that contains the wavetable data (see screenshot):

The wave-data code sections have obvious names indicating what type of shape or sound it is.  Don’t like the “bassTable”?  Start with replacing that one.

💿 Go to one of the text files you have open, copy the stuff highlighted in pink (after the brace--copy only the numbers, that is) and paste it into the Arduino window (again, after the brace), under one of the “const int16_t xxxTable[] =“ sections.  Here:

Congratulations!  You’ve just replaced one of the stock sounds with one of your own!  

💿 A good idea would be to indicate which wave you used by making a comment in the code.  Anything following two forward-slashes (//) is considered a comment, for instance:  // AKWF_0183

Now, move on to the next section and the next, until you’ve replaced all the waves you want.  Close each text file prior to moving on to the next, so you don’t get confused.

💿 Having trouble finding the exact wave you want to replace?  The order of the waves is indicated in the code of the UPDATECONTROLS_CZ tab (or file, if it didn’t open automatically). 

The comments indicate which wavetable (low, middle, or high) you’re modifying.  

You can even change the waves that are enabled when “X” and “FM” modes are active.  (Look at the UPDATECONTROLS_CZALT, UPDATECONTROLS_FM, and UPDATECONTROLS_FMALT files.)

💿 Final step: save the files you have modified!

The Real, Final, Final Step

Ok, the real, final step is uploading the firmware to the OA, using the procedure outlined in the official build-manual.  It’s super-easy to do, assuming the Teensy is recognized and you already have a working build.  Basically: just connect the sucker to your computer and hit the “upload” button (the one with the arrow).  

Boom!  Done.

Enjoy some music made with the OA, and some not: http://www.glitched.org , http://soundcloud.com/d01

Thanks for reading.  

(d), a.k.a. “glitched”

[glitched-diy]

More Orgone Accumulator Sounds

...

More sounds from the Orgone Accumulator module (mainly).  This sounds more like a "real" track than a demo, I think.  Certainly, a taste of things to come (I've got over 38 minutes of cool sounds to cut down!)

No "filters" on the main sound, just a Makenoise Optomix used as a VCA/mixer.  The OA's output goes through the Elektron Analog Four's lush reverb. 

As for control, a Mutable Instruments Peaks in dual-"tap LFO" mode animated certain controls (position, frequency, and pitch); one LFO using a sin-like wave and the other, a random-ish wave.  (The latter went to a quantizer prior to hitting the pitch input on the OA.)

A click track from the A4 was sent to a RCD and SCM, which kicked off the envelopes, tap LFOs, and pinged the Borg Filter LPG. By the way, there's a noise-percussion track with a little saw-wave thrown in; this is the stuff that went through the Borg. Pitch controlled by RLS/Turing Machine, I think.

Anyway, there you go.  For more of my "real" music, visit: http://www.glitched.org

[glitched-diy]

Orgone Accumulator #2 Is Done!

Status: Recently Completed!

Yep, just like that, I'm finished with another Orgone Accumulator! 

(Pop-out, porno-style angle)

After completing the first one for Phaelam, it was my turn to enjoy the sometimes harsh and dirty digital sounds of the OA...but first I had to build it.  Easier said than done...but not that difficult, either.

I used a bunch of shortcuts in building this one because I was a lot more confident in what I was doing.  The "Teensy Fortress" (Teensy 3.1 MCU and necessary pin-headers/sockets) is a good example of this.

("Teensy Fortress," in the middle; "Teensy Gate," on the side)

I still followed the instructions provided by the build-guide, but I found that, after soldering some of the surface-mount pin-headers, soldering the sockets to the main board helped with overall stability and alignment of pins.  In other words, when you're done soldering a few pins of Teensy headers (and the SMD header), proceed to the main board.  There, you can solder some pins of the sockets to the main board, then solder the rest of the pins on the Teensy, with full confidence that everything will be aligned.

The other components that needed to be aligned were the switches, pots, and jacks.  I took a different approach to the jacks:  1) hold switch flush to board with one finger 2) take solder between thumb and forefinger 3) take soldering iron in other hand (I'm a lefty) and solder one pin from the top side of the board 4) repeat for the other switches and complete the job by flipping the board over and soldering the rest of the pins.  (As you can tell, I don't have a pana-vise; "helping-hands" are more annoying than helpful.  "Annoying Hands" is what they should be called.  Yeah.)

I really didn't feel like screwing/unscrewing the pots over and over again, so I basically just inserted the pots on the board, put the faceplate on, screwed a bunch of them in, and soldered everything in one go.  If you screw a bunch of the nuts in, this gives enough stability for soldering and will result in good alignment of the pots.

Jacks are similar.  Put the jacks in the board, screw in some of the pot-nuts, turn the sucker over, and screw in ALL of the nuts for the jacks.  From there, you can be assured of perfect height and alignment.  (I'm using Kobiconn jacks, by the way.)  

Testing went fine, calibration was a little more difficult for some reason, but overall, the whole thing went very smoothly.  Even knob-selection was painless.  For some, choosing the right knobs is actually the most important part of a DIY build.  I guess I'm one of those guys.  

The pink knobs just hit me as the ones to put on the waveform controls.  Although the cream-colored button-caps clash a bit with the white Davies, I think it looks great. 

So, there you go.  I've already made a bunch more sounds with this unit.  It's such a beast.

[glitched-diy]

Orgone Accumulator Build: Final

Satus: Complete!

Day 4 - END: 

I was determined to get this thing done on the fourth day.  All told, I can't remember how many hours I put into the Orgone, but it seemed like a lot, actually.

On this final day, I started with the eight switches.  Not a big deal, but centering them and putting them flush on the PCB is essential if you want clean "click/unclick" functionality.

Next were the pots.  Again, nothing too out of the ordinary, but one must be patient and make sure those pots are flush to the PCB, same as the switches.  The best way to do this is to go by the build document and use the faceplate as a guide.

The density of the board once again made this soldering segment more challenging than a typical synth-DIY PCB.  I don't have the best tools (just a standard consumer-grade Weller), so I had to be extra careful.  The pictures don't accurately depict how tight it is.  In some spots, you have about the space of a dime to work with--the thickness, not the diameter!

(The business-end.)

I guess the jacks weren't too bad.  Pretty standard stuff the rest of the way.

Calibrating the -3/+3V with the trimmer was easy, the 1V/Oct was not, at first.

I tried to do it several ways, but the best method I discovered was this:

Hook up a MIDI keyboard to MIDI-to-CV gate interface.

Pipe 1V/Oct output to the input on Orgone.

Play the lowest C-note your interface allows (usually equates to C0/0V).

Play the same low C-note ("C0") on another synth/keyboard, using a simple tone.  I used my Monomachine's SID machine with a TRI wave.

Attempt to match the tone of the "control" sound (Monomachine) with the Orgone by using the "Tune" knob. Try to minimize "beating."

Now, the actual calibration part: Create a simple, whole-note chromatic scale sequence on the control-synth, starting with C0 and ending with C1 (C0, D0, E0...C1.)  Send this sequence to the Orgone via your MIDI-to-CV converter as well.  Now you have the same sequence running on both the control-synth and the Orgone.  They won't match, which is what the calibration is for.

Use a tiny screwdriver to adjust the V/Oct scaling trimpot, located below the "Fine" control.  Continually twist to the right or left while the sequence is playing.  You may have to adjust the "tune" parameter to maintain the base note.  Eventually, the note spread will sound more accurate.  If it's difficult to determine with the current waveform, switch to something like a saw wave.

After the first octave sounds good, move up an octave:  sequence C1 through C2.  Adjust settings, as above.  Continue through to the highest octave your CV interface allows, then go back to one of the lower octaves, just to check that nothing drifted.  

Your Orgone Accumulator is now calibrated to a 1V per octave scale!

Another type of calibration is centering the attenuverters (the CV input pots), so that no modulation is applied when a CV signal is inserted into one of the jacks.  This was almost impossible, as the "0" value is almost literally the width of a human hair.  Even more difficult:  putting the knobs on so that the line matches up perfectly with the centered value of the pot.  I did my best.

Here's the completed module, built for my buddy, Phaelam (what an amazing friend I am!):

(Purple/black/cream is the scheme he went with.  Cool!)

What's left to say?  It was an arduous, but supremely rewarding, then extremely heart-wrenching experience (because I have to send it away!)  

The sound?  After playing with it for mere hours, the Neutron-Sound Orgone Accumulator is an (ironically) digitally-organic, lively, nasty, wavetable/FM hybrid oscillator, the likes of which I've never heard in Euro-world.  Shape Shifter who?  Megawave what?  

"Wanna hear it, here it go":

If you're in the mood for dirty, wavetable synthesis with extensive modulation possibilities, and you have some experience with a soldering iron, there's nothing else more worth your time!

[glitched-diy]

Orgone Accumulator Build Day3: Orgonna Go Outta My Mind!

Status: In Progress...

Day 3: 

(All caps done!)

(Teensy 3.1 = complete!)

(Teensy daughterboard attached to its mother...awwww...)

Although there was much talk of the Teensy portion of the build being most formidable, I was able to get through it without wounding myself...much.  I won't say it was easy, though.  The space is tiiiight.  I had to shave off a few mm from the inner pin-headers (see the second pic) and both sides of the 5-pin header row.  It was still a struggle getting everything to line up.  

My advice for the Teensy portion of the build: go by what the build manual says!  It made things a lot easier.  If you're the reckless type and just solder everything pin-by-pin, it'll be super-difficult to align the pins to the main Orgone board and even more difficult to extract the MCU, if you ever need to.

At least I can see the finish-line. 

Left to go:  pots, jacks, switches, and misc.  Then:  testing and calibration.  

[glitched-diy]

Orgone Accumulator Build, Day2

Status: In Progress...

So, here's where I am at the start of Day 2: 

(All the resistors and 70% of the caps are done!) I should be through all of the caps tonight and I might start working on the Teensy pin-headers.  

Thoughts so far?  There's a million points to solder and they are really tight, but I haven't run into anything "difficult," per se.

Keep it locked to this Tumblr for more.

[glitched-diy]

Neutron-Sound Orgone Accumulator

Status: Working On It!

(Orgone Accumulator panel+PCB)

http://neutron-sound.com/noa.html

A new project to work on, just in time for the holidays.  This one is actually not for me, but a build for a long-time friend and collaborator, Phaelam.  As payment, he offered to buy me a panel+PCB set, as well as a Teensy 3.1 (the micro-controller at the heart of the Orgone), so I couldn't pass it up.  It'll be tedious, but not too difficult.  

As you can see, it gets pretty dense in some spots, but the Turing Machine was darn dense as well.  The difference here is in the number of points to solder, as well as the Teensy 3.1 mini-PCB (they don't call it "teensy" for nothin'!)

("You are my density...")

Here's a pic attempting to demonstrate the diminutive dimensions of the Teensy (then again, you don't know how big my finger is; I'm actually a giant): 

(The heart of the Orgone, the Teensy 3.1 MCU)

Updates, as soon as I get some of this done.

[glitched-diy]

Random Looping Sequencer (a.k.a. "Turing Machine") #2

Status: Recently Completed

(Naked electronics)

This one took a lot less time than the previous two because I had all the parts on hand, including the spare, black panel.  (The "previous two"?  Yes, the "randomness" wasn't working on the first one and all the soldering/desoldering ripped up a bunch of pads.  I said screw it and just got a fresh board.)

(Black on black.  Pretty ridiculous, no?)

The panel is all blacked out, which automatically makes it look cooler.  In reality, I was just sick of my previous design and decided to start over with a blank slate.  I'm not sure if I'll even bother getting a "grayscale" panel for this one; I kinda like it!

One thought you may have is why one would need two TMs.  Think of these possibilities:

1) Use one as a random/looping sequencer and the other as an oscillator (driven from another osc's pulse wave, at audio rates).

2) CV output of one TM into the CV input of a quantizer (uScale, for instance); CV out of the other TM into the "key shift/transpose" input of the quantizer.

Now, you want two, don't you?

Here are some lessons learned and tips for would-be builders:

1) It's a packed PCB, so plan your component placement procedure appropriately.  Start with the lowest profile things (resistors) and build your way up (receptacles, LED, box-caps, ceramics, jacks, switches).  Some parts will challenge the steadiness of your hand and you may even melt the edge of a capacitor, but just be patient and you'll be fine.

1a) If you leave those two electrolytic caps for the end, you might wonder how to solder them in, due to the close proximity to the other capacitors, but it's easier that you think.  Just put the leads of C5 and C6 through the bottom of the PCB, leaving some slack.  You should easily be able to solder the pads on the component side (which is actually the bottom of the PCB).  

2) Most capacitors are just there for decoupling, so it makes little difference what type you use.  For instance, the BOM recommends ceramic 100n caps, but the box types work just fine.

3) Pot values don't matter much.  I used 100k for this build, but 50k for the last one.  Just make sure they're linear and can fit through the panel (and stick out far enough for a knob).

4) Transistor selection is important!  Not every 3904 (or whatever you decide to use) is the same.  Compare noise "power" by socketing a few.  (Heck, you could even just push the transistor down as far as it will go; the tension/friction should be enough for the noise portion of the circuit to work.  If it does, just solder the sucker in there!) 

Any questions or comments?  Tweet to me @ d01. Here's the Random Looping Sequencer thread on Muff Wiggler, for further help.

[glitched-diy]

MusicThing "Turing Machine" #2

Status: Working On It

This will be the second "Turing Machine" (or, as I like to call it, the "Random Looping Sequencer") that I've DIY'ed.  The first one was simple (and cheap) enough, so I figured I'd add another one to my mostly-DIY rack.  To mix it up, this one will have a green LED bar and use a white version of my RLS panel.

I plan to eventually replace it with those classy-as-hell Grayscale panels:  http://grayscale.info/panels/turing-machine-random-looping-sequencer/ .  (I thought my panel was decent, but there's a certain minimalism that I was trying to capture that Grayscale did much better.  Plus, it's metal, which looks more "professional" than acrylic.)

Updates in a couple days, assuming I still have all the parts.

[glitched-diy]

Manhattan Audio (negativespace) CP3 Mixer

Status: Recently Completed

Link: http://www.muffwiggler.com/forum/viewtopic.php?t=91461

(It's the one in the middle.) For a while now, I've been wanting another STG .MIX.  It has this clipping distortion that sounds really good to my ears.  It's not saturation or "overdrive," but something else.

Anyway, a .MIX is about $130, but it's such a simple circuit.  I often thought of just putting it on a protoboard, but I never got around to it. 

Well, thanks to "negativespace" (Manhattan Audio), I didn't have to lift a finger to design my own layout or panel.  Yeah, yeah, sometimes I take the easy way out and just opt for a pre-made PCB and panel.

The build was extremely easy and required no exotic parts.  The schematic suggests matching transistors, but the random 2904/2906's I picked out of a pile worked perfectly well.  I didn't even use the specified pots; in fact, they were all different: 100kA, 50kA, 100kA, 25kB.  No ill effects came with the haphazard component selection.

Testing went fine, but I was a bit confused about how to zero-out DC offset.  The measurements I got were ~/+100mV on either side of the output pot (the purple knob).  I later found that these stats were fine and that the main thing was to get 0mV on the meter, at the middle position of the output pot.

Easy, fun build.  I should put up some sounds, no?

[glitched-diy]

SDIY Initial Mega-post (from synth.pen.io)

This is the initial "mega-post" that includes all of the information that was contained in synth.pen.io .  

Special Note

Word to the wise: always document your progress on any unfinished projects! I added a couple things on the 4069 OSC, but de-soldered certain connections for experimental purposes. Months later, I've forgotten what connections go where and why I did certain things. It's vexing, but now I have to re-trace my steps. It can be difficult, even with the aid of the schematic. Again: documentation, like the type found on this site, can be extremely helpful!

René Schmitz's "Korg MS-20" Filter

Status: Maybe Later

http://www.schmitzbits.de/ms20.html Schematic:

A very simple filter, based off of the "late" MS-20 -style sound. Right now, it's on the breadboard, about 95% complete, but I haven't looked at it for a while, so it might take me a bit to figure out how to finish it!  Then again, I might not have the motivation to commit it to circuit board due to the recent acquisition of yet another analog synthesizer.

EDIT: I'll tackle this one at a later date.  As simple as it is, I don't have a couple of the necessary parts; namely, the CA3080.  

YuSynth ARP 4072 Filter

Status: CANCELLED

http://yusynth.net/Modular/EN/ARPVCF/index.html UPDATE: In 2011, I had a strong desire to build this filter, using components I hoped to purchase from Bridechamber, a well-known synthesizer DIY retail operation. Long story short, I never received the items I paid for and the guy gave me my money back. Since then, my knowledge of SDIY has grown immensely and don't feel I need to rely on prefab PCBs as much. Don't get me wrong; I prefer PCBs--they're obviously quicker and manufactured better--but if I have the time and patience, usually opt for truly doing it all myself. I have put this one on the back, back, back burner.

PT2399-based delay:

Status: Maybe Later

There are many flavors of the circuit based around this delay chip. I like this approach because it includes a soft-clipper before the input to create a warmer sound:

http://www.schmitzbits.de/pt2399.html

UPDATE: Looks like BugBrand has used a similar circuit:

http://matrixsynth.blogspot.com/2011/04/bugbrand-pt-delay-overview-demo.html

UPDATE: Debating whether to build this one or not. I mean, it's "just" a delay and I already have some really capable models in the Virus TI Snow ("tape" style) and Monomachine (dirty digital).

Everything Below: Complete!

Hexinverter 909 Clap

Well, almost complete.  Everything works as it should but I don't have a panel for this one, yet.  Still anxious to hear some sound?  Fine, listen to this mind-blowing demo!!

Here are the naked electronics:

CGS52 -- Simple Wave Folder

http://www.cgs.synth.net/modules/cgs52_folder.html

What a simple project! Great bang for the buck.  Nothing but a couple op-amps, resistors, and transistors. Took me all but 30 minutes to go straight from the schematic (below) to circuit board.

Tom Whitwell's "Turing Machine" (Random Looping Sequencer)

http://musicthing.co.uk/modular/?page_id=21

This is probably my favorite DIY project of all-time.  Not only does it duplicate (and sometimes surpass) most of the functions of the lauded Wiard Noise Ring, but it's about ten times cheaper, when all is said and done.  If you're able to get your hands on a PCB or kit, definitely build one! The original panel, by Whitwell, is artsy and  very creative, but a little "off the wall" for me and my current modular design style.  I perfer a simple, straight foward look to my panels. Besides that, I just wanted to have an excuse to create another design and try out some different materials at Ponoko.

By the way, if you want to make one of your own, I have put the .svg (for Illustrator or Inkscape) up, for free, on Ponoko.  Go here.

The design turned out great (some say it looks like HAL or KITT from "Knight Rider"), but the "matte black acrylic" does it no favors.  The problem is that the porous nature of the material; if you make a mistake painting, the panel is almost irrevocably smeared.   You can see that there are a couple spots that could use some cleaning up, in the picture above.

There's also a white version, seen here, which has no "data display."  I wanted to see if the LEDs would shine through enough; they do, but they're way too diffused to tell which data "bit" is on or off.  In the next Ponoko run, I'm definitely going to do a white version with the data port, as well as a standard black (glossy) acrylic one.

Random Looping Sequencer Expansion

On the Muff Wiggler Music DIY forum, somebody posited the idea of using the individual "bit" outs as pulse triggers.  In other words, the data "pattern" that is converted to control voltages by the DAC can be used as triggers (or gates).  Forget about the technical details; all this means is that I can use a CMOS logic chip to create several different pulse sequences, based on the "random" sequence that is spit out by this module.  Sounds cool to me! 

Designing this expander was actually very easy.  Essentially, all I had to do was tap the outputs of the clock and various "bits" and apply some logic via a CMOS chip to get some different pulse patterns, out.  By "tap" all I mean is connecting the output of one bit to input "A" of the CMOS chip and output of the clock or another bit to input "B" of the CMOS chip.  The resulting output will be different, yet related to the original "random" pattern.

Final unit with expander:

Music From Outer Space PSU (#2)

http://www.musicfromouterspace.com/analogsynth_new/WALLWARTSUPPLY/WALLWARTSUPPLY.php

The MFOS wall-wart PSU is a cost-effective way to feed your modular's ever-expanding power needs.  I bought the first PSU directly from Ray Wilson, the second, I made myself, on a strip-board because it was so straight-forward.

Here's the schematic:  

Even someone who is not well-versed in reading schematics can see that it's really just a couple of circuit "lanes".  To tell you the truth, though, I did cheat, a little.

There is a great strip-board layout, on electro-music.com, by a guy named "thundarr."  It looks like this:

(From http://electro-music.com/forum/topic-35373.html&postorder=asc )

Here's the result:

Nothing too interesting.  I simply housed the units in a small, Radioshack project box and maybe even a Radioshack-sourced DC jack, too (but that might have been from Newark.com or Mouser).  Super easy.  Oh, and I use 12v, 1.5 amp wall-warts, found at Jameco. 

SSM2164-based dual-VCA

http://www.sdiy.org/philgallo/mgbvca.html (Something similar to that.)

This is the schematic from which I worked:

If you're a beginner, it may seem a little confusing, with all those lines going everywhere. That wasn't my problem; I was more perplexed by the "why" of it all. What's the purpose of those .1uF and 10uF caps? The 560pF caps and 500 ohm resistors?

Well, the .1uF and 10uF caps on the power rails are for...power supply de-coupling. Basically, smoothing out the ripples in the current coming from the PSU. Once I learned this, I realized that 50% of the components on the schematic are devoted to de-coupling. Cool. No big deal.

But what of the small cap and resistor attached to the signal inputs? Simple: a lowpass RC filter network. Essentially, this little filter dampens the frequencies above ~20khz--freqs that humans can't hear, but could cause unwanted effects on the VCA. (Calculate your own values for a RC lowpass filter: http://sim.okawa-denshi.jp/en/CRlowkeisan.htm )

All I have to do is connect the inputs and outputs. I'll report back with results!

UPDATE 11012011: After a couple false-starts, it works! Two linear VCAs from one SSM2164 chip! It's really smooth and quiet. Although I like my modules to imbue some sort of "color" to the signal, it's good to have a more surgical-sounding element in the mix.

I used tiny wires in a more point-to-point style, this time. I liked the way it worked, but I'm going to be smarter about the front-panel hardware, next time. It's just too messy...but it works!

UPDATE 11252011: The front panel has arrived from frontpanelexpress.com and it's awesome! See the pictures, below. You may be asking, "what's the WASP filter doing on there?" Well, I decided to combine the panels, to save a bit of dough. Functionally, I perhaps gave up a little, but I figure that I'll be using both the modules in most patches, so it doesn't bother me that they're permanently joined together.

I think the separate panels would have been about $30 and $25, so I saved $15. I wouldn't advise this for someone who has a modular setup that's constantly in flux.

The above shows what FPE sent me, prior to the in-filling.

You can really see the fine detail on the engraving. Very deep, no burrs. FYI, the Front Panel Designer font used for the VCA was "Bold, 7-stroke." I wanted the WASP to be distinct from the VCA, so I used "Helvetica Medium, Outline" for the text.

The final product. I decided to leave the 2164 alone because the bare aluminum offers so much contrast to the anodized face. Of course, the WASP filter had to be yellow; an homage to its forebear. The yellow Davies-clone knobs are questionable; they're a bit too bright for my taste, especially the filter mode switch. It's just a little off. Still, I'm extremely happy with the end product!

The rear. It may seem a little cramped, but that's because...it is. Nothing is touching anything it shouldn't, though!

CMOS (CD4069)-based VCO:

It's completely functional, I just need to calibrate it to 1V/oct. The PWM sounds really cool, but the saw is pedestrian. I could easily build another copy on the same board, which is what I'll probably do next.

UPDATE 04062011: Complete! Well, all I have to do is glue the two transistors together and put some heat sink compound on them (er, reverse that), but I'm considering this done.

Look at the difference between the two VCO layouts! I really went freeform with the second one (left) because I misjudged how much space I had left on the board.

EDIT 03/2017:  Sorry, some pics were lost due to an image-host change :(

Now, that's a disorganized rat's nest! I really have to figure out how to solder in an orderly manner. Still, you can't argue with the results: it works!

UPDATE 04122011: The tracking on this oscillator is pretty bad. I can't get any more than a couple octaves out of it. Disappointing, but this is still a very usable and cool sounding modulation osc. Eric Archer (ericarcher.net) states that this is probably due to not using matched transistor pairs. Yep, that's probably it. Plus, I didn't really do a great job of sticking the trannies together. They're sort of off-axis.

I did read something on Thomas Henry's site about the quality of certain manufacturers' 4046 chips. I wonder if this goes for the 4069 CMOS chips as well. To test this theory, I bought a few 4069s from various manufacturers. I'll report back.

Update 04262011: Well, the chips from the different manufacturers didn't make a lick of difference. The only thing that really matters is the matched complementary transistors (NPN + PNP). I probably won't want to put forth the effort to make a matching device, so this may be the end of the road for the VCO (but I'm still very happy with it!)

Here are a couple sound samples, from my soundcloud: http://snd.sc/fcil4E EDIT: Ok, just one sound clip. I didn't like the others. More, soon.