Modular Synth Blog Archive, Part 2

Unfortunately, the dates for the following postings were lost in the shuffle of recreating this site. They start in late 2004, I believe.

The Modular Synth Diary Returns

In my own defense, until recently there hasn’t been a whole lot of news about my modular project. I’ve been kind of busy with other stuff (a major understatement if there ever was one) with the founding of Audio Damage, Inc. and the acquisition and reestablishment of Paper Jade, not to mention a trip to Italy and the usual demands of stuff like eating, sleeping, mowing the lawn, etc.

Anyway, enough whining. Here’s a quick summary of what I’ve done since my last entry, which I am startled to see was made just under two years ago.

I completed a MOTM 190 VCA module, which filled the last available space in my cabinet. Realizing that I wasn’t going to be quite satisfied with the contents of that cabinet, I was luckily able to get my cabinet maker to make a second cabinet to go on top of the first. It holds only one row of modules, but has special spaces for my Expressionist MIDI/CV converter and its breakout panel. The breakout panel I haven’t built yet; that’s the next project on the list.

The next two modules I built from PC boards I bought from Oakley Sound Systems. I built the Equinoxe phase shifter and the Sample and Hold. The latter I built to complement my MOTM S/H module because the Oakley version can trigger on either the rising or falling edge of the control signal, which offers some interesting possibilities. In the long run I may replace the MOTM module with another one of these, although that would mean having to obtain another noise source, too. I was impressed by the quality of the Oakley PC boards and instructions, and am quite happy with the performance of the modules. I have photos of both modules and will try to put them up here soon.

At some point I realized that another mixer would be handy, and after some cogitation I converted a Blacet Quad VCA/Mixer module kit into MOTM format. I had the panel fabricated by Front Panel Express. This was the first time I tried making a panel this way, and I was really pleased by the results. It doesn’t look exactly like an MOTM panel, since it’s anodized and engraved rather than painted, but it looks very nice and is certainly more attractive than anything I could cobble up myself. I replaced the Blacet’s pots and jacks with the same parts that MOTM modules used. I’m happy with this module also, and in simple patches typically end up using it more often than the MOTM VCAs and mixer because it conveniently combines both functions.

Next I converted a Doepfer Fixed Filter Bank to MOTM format. This one I sort of half-converted. Instead of completely changing it to match the MOTM format, I fabricated a bezel (or rather Front Panel Express fabricated a bezel) of MOTM size with a rectangular hole in which I mounted the Doepfer module. I also mounted two 1/4″ jacks in the bezel, and connected them to the input and output jacks of the Doepfer. Finally I made a power-connector adapter to match the Doepfer’s connector by soldering wires onto a 0.1″ header-pin block and putting blobs of hot glue over the exposed connections for insulation and stability. The filter bank itself creates exactly the sorts of “scooped mids” and adjustable formant filtering that I was hoping for. I’m considering converting the Morphing Filter in a similar manner, but it would be a considerably more complicated adaptation because of its plethora of connectors.

For my most recent project, I adapted a Cyndustries Quad Low Pass Gate module. I started with a stuffed PC board I purchased from Cyndustries, and again used a Front Panel Express panel. It took me quite awhile to figure out how to cram all of the controls and jacks for this module into a 2U panel, but it came out well. Again I have pictures of this one which I’ll post soon. I had a little trouble with this module, thanks to a soldering error. When I first powered it up, one gate did not work. Cynthia was very helpful with my troubleshooting efforts, and eventually I found that I had bridged two of the connections at the edge of the PC board with solder. This unfortunately burned out the input LED of one of the Vactrols, so I bought another Vactrol from Cynthia. Once I installed that, everything worked fine. What a great-sounding module this is! I’ll have to make some samples to put up here. It has a weird resonant quality to it. It’s hard to describe, except that it sounds very “analog”. I’m told that it’s a familiar characteristic of Buchla synths, and that you hear it all over Morton Subotnick’s old recordings.

That’s the news to date. The next projects are the breakout panel for the Expressionist, a Blacet Time Machine, and two more MOTM filters. I have the kits for all of these; just gotta find some time to break out the soldering iron again.

Photos of Converted Modules Now Online

The photo gallery now has photos of my Oakley Sounds Equinoxe and Sample & Hold, and my Cyndustries Quad Low Pass Gate. You can find them by clicking here.

Breakout Panel for the Encore Expressionist

Recently I finished a breakout panel for my Encore Expressionist MIDI/CV converter. I started this project well over a year ago, but tended to put it aside in favor of more interesting projects like building new modules.

When I had my second cabinet built, I designed it so that it could accomodate the Expressionist and a breakout panel for its connectors. Since the Expressionist’s connectors are on its back panel, mounting it directly in a synthesizer cabinet means that there is no easy way to plug cables into it, particularly if you can’t walk around behind the cabinet, which is the case in my studio. I found Dave Bradley‘s page about his breakout panel, including schematics for adding buffers and LEDs to the outputs. Buffering the outputs made good sense because there have been a couple of occasions during which I’ve thought that the Expressionist’s gate outputs in particular lack sufficient oomph to drive several modules at once.

I used Front Panel Express to fabricate a panel. While I was laying out the panel I first thought of adding a simple multiple in the unused space. Then I came up with the idea of adding additional jacks to some of the outputs, wired together. This seemed even more useful than a multiple since typically I end up using a multiple to connect the Expressionist to several EGs and oscillators. But I didn’t need extra jacks on all eight pairs of outputs, nor did I have room on the panel for them, so I just added extras to the first four output pairs.

The first real challenge was coming up with a way to attach the PC board to the panel. Ever since Paul Schreiber threw down the gauntlet with the invisble PC board attachments in his MOTM panels, people have come up with various ways of doing the same thing without access to his secret process. Larry Hendry’s Stooge Brackets are the usual solution, and I’ve used and imitated them several times. That approach wouldn’t work for my breakout panel because of the dense packing of the jacks, and because the only aluminum stock I had on hand was too thick to fit over the shafts of the jacks and still leave enough thread to attach them firmly to the panel.

After a bit of thought I decided to cut some L-shaped brackets from aluminum angle stock and glue them to the back of the panel, and glue threaded standoffs to the brackets to accept screws for attaching the PC board. My father said that he’d successfully glued aluminum to aluminum using slow-cure epoxy, so I went to the local hardware store which has a vast selection of glues and found some epoxy specifically for use with metals. I used a wire brush tool in a Dremel to remove the anodization from the back of the panel, thinking that doing so might create a firmer bond.

I ended up first gluing the standoffs to the brackets, letting that glue dry, and then gluing the brackets to the panel. This let me attach the PC board to the brackets first, thus ensuring that I glued the brackets to the panel in the right position. It worked very well. The slow set-up time of the epoxy made it relatively easy to get the parts aligned properly before clamping them, and the whole assembly is very solid. There are photos of it in the photo gallery of this site; you can see them by clicking here.

Once I had the mechanical assembly completed I put together the electronics. I decided to use point-to-point wiring on a clad prototype board. In retrospect I wish I had fabricated a PC board instead. The circuit had enough parts and connections to make the point-to-point wiring job rather tedious. Had I used a PC board I would have learned how to make PC boards, but instead all I learned is that I don’t want to do another big point-to-point wiring project. 🙂

But patience prevailed.

I’m nearly done with another project I began a little over a year ago, a MOTM conversion of a Blacet Time Machine analog delay. Hopefully I’ll finish that either this weekend or the next.

My First CGS Module, and My First PCB

I recently completed a module based on one of Ken Stone’s PC boards and one of my own. Ken sells a number of PC boards for DIY modules. His website is here. I ordered a couple of PCBs from him awhile back, and recently completed the Digital Noise module. I realized that since this module has a relatively small number of jacks and controls, it wouldn’t occupy more than about half of a single-unit MOTM-sized panel, so I decided to add a simple sample and hold module to the same panel (since noise sources and sample and hold modules go together so nicely).

I decided that this was also a great opportunity to try my hand at PC board fabrication. I decided to use Express PCB because their software looked like the easiest to learn, and because their prices are quite good if you can confine your design to a relatively small area. (Since then I’ve learned that Custom PCB is probably a better way to go in the long run, but they don’t have software of their own. It looks like CadSoft Eagle is the least-expensive full-featured PCB layout software, so I’ll probably use that for future products.) I borrowed the circuit from another module I’d built in the past, making a few modifications to suit my needs and to make it fit in half of a 1U panel. [Update: this is a fairly old blog entry. I came to the conclusion that Eagle sucks, despite its popularity. I use DipTrace for all of my PCB layout, and BatchPCB for fabrication.]

Capturing the schematic in the Express PCB software was pretty easy. Laying out the PCB itself was considerably more difficult. At first it seemed almost overwhelmingly complex (even for a relatively simple circuit like this one) but as I spent some time with it I started to rather enjoy the process. It’s sort of like solving a puzzle. Over time I found more ways to simplify the traces by rotating components, moving stuff around, and so forth. It was really fun to receive the finished boards in the mail–just as satisfying as receiving my first Schaeffer panel.

I used a Schaeffer panel for this project also. This time, to mount the PCBs on the panel, I used epoxy to affix long threaded standoffs to the back of the panel. Then I used small angle brackets and screws to attach the boards to the standoffs. This method worked quite well, and I will certainly use it again for future projects on small panels.

The CGS Digital Noise module was easy to build. I had a little trouble finding the 4000B-series CMOS ICs, but the Patron Saint of Synthesizer Builders came to my rescue. (Thanks again, Chub!) The module worked on its first power-up, and has a really interesting sound. It sounds quite like old arcade games, which makes sense since they probably used a similar pseudo-random-number technique for generating noise.

My own PC board was a little more stubborn. It didn’t work at all at first, but I soon discovered that I had put the three transistors in backwards. Once I corrected that mistake, and then corrected a minor wiring error on one of the jacks, it worked perfectly.

There are photos of the completed module in the photo gallery here. Here are some sound recordings in MP3 format:

White output, frequency knob rotated from high to low and back again

Pink output, frequency knob rotated from high to low and back again

Frequency CV input driven by output of sample and hold, which is sampling the white output; VCA envelope and sample and hold triggered by LFO

D 1 and D 2 outputs connected to two FM inputs of MOTM-300 VCO to generate random pitches. Output of VCO connected to MOTM-440 filter, followed by VCA and Blacet Time Machine delay

Another CGS module: Synthacon filter

I finished a CGS Synthacon filter recently. This project didn’t go as smoothly as my previous CGS-based module. This module is based on the filter from the Steiner-Parker Synthacon, an aged and obscure analog synthesizer. I saw one once, many years ago. It was not a very good synth; it had laughably bad temperature stability. You could fan air into the back of the cabinet and hear the oscillators drift quite dramatically. However, it had an unusual filter. The filter has a 2-pole multi-mode topology, but the modes are selected at the input rather than the output. It has three inputs, low-, high-, and band-pass and one output. If you plug one signal into one input, the filter behaves like a standard filter. But if you plug more than one signal into the inputs, the filter acts as a sort of frequency-dependent cross-fader. It’s a rather odd effect, but it sounds wonderful if you run a couple of oscilators tuned to some interval into it and sweep the frequency back and forth.

The CGS PC board recreates of the original Steiner filter circuit which was published in some magazine in the 1970s. (You can visit the CGS site if you want to read the details.) The instructions for the module include a lengthy message from John Loffink describing how he changed resistor values on the PC board to improve its frequency control and 1V/octave tracking. As he described, if you build the filter with the parts as specified by Ken, the frequency knob will sweep the filter through its entire range in about half of its rotation. Aside from that design flaw, it wasn’t entirely apparent to me how one should wire up the PC board to give it a standard (or at least MOTM-like) complement of frequency CV inputs. I wanted a front-panel knob for setting the nominal frequency, an unattenuated CV input with 1V/octave frequency control (or at least something reasonably close), and an attenuated CV input for mixing in a CV from an envelope generator or LFO. It became obvious that setting up this filter in this manner would require adding a CV mixer rather than altering the exising CV summing circuitry, since (among other reasons) the existing circuit doesn’t really decouple the CV inputs. Ken confirmed this after I asked about it on the mailing list.

So, after assembling my filter, attempting to bring it up to snuff with John’s modifications and then giving up on those, I partially disassembled it and grafted on a CV mixer/scaler with a bit of perfboard. I made the CV depth panel control a reversing attenuator, a feature whose usefulness helped to offset the hassle of soldering, unsoldering, and resoldering the panel pots.

I didn’t expect to have to change the original circuit this much, but I’m happy with the filter in the end. It has a very unusual sound–rather rough and chunky (for lack of a better word).

Rest in Peace, Larry

A tragic loss: J. Larry Hendry and his wife were killed while riding his motorcycle.

It’s a weird thing to discover how much someone you’ve never met in person means to you. It’s a sad thing to discover this because they die.

“Stooge” Larry was a prominent–arguably the most prominent member–of several overlapping online communities of synthesizer enthusiasts. He was known in MOTM circles partly for his Stooge panels, MOTM-lookalike panels for modules from other vendors, and patch cables of both high quality and low price. He was known more widely for his unfailing kindness and good humor and generosity with fielding questions from fellow synthesizer builders.

I never met him in person, and I exchanged email with him only on a few occasions. I knew him mostly through his postings on the MOTM mailing list. In my 20 years of participation in online communities, I can’t think of another person who has been as consistently kind, warm, and generous as Larry was. My own synthesizer has five Stooge panels and dozens of Stooge cables.

We miss you, Larry.