Building a Buchla 200

A build log for my Buchla 200 DIY clone system. Maintained by Trevor Wilson.

[source-of-uncertainty]

Hey, it’s been a while. I haven’t built much Buchla lately, but I’ve just received a new pile of CGS Serge boards that I’ll be building into a custom case and panels. How about I post about that here, eh?

[source-of-uncertainty]

The finished 285r. The balanced/ring mod section sounds fantastic, but the frequency shifter section leaves something to be desired. I can’t put my finger on it. When I use the internal modulator, there’s a grit to the output that is unpleasant, but when I modulate with an external signal the outputs sound cleaner. I think there could be some distortion at work here, so it might help for me to make the internal oscillator’s signal a bit quieter. I’ll figure out how to do that sometime soon. 

[source-of-uncertainty]

281r is assembled and working great. It just needs trimmers installed, which I’ll add tonight and then calibrate it. This one was a little tricky because the footprints for the 16mm pots were slightly vertically misaligned from the drill holes in the panel. So I bent the legs of the pots in order to get them sitting in the panel straight, and made sure to reflow the initial solder joint I’d added to each pot. They’re pretty much perpendicular to the panel now.

I’ll try to get some audio up soon. 

[source-of-uncertainty]

Work in progress on the 281r. Everything is populated on the PCBs except the trimmers (which are in the mail). I need to insert the ICs, wire up the jacks, and build the power cable. Then testing, and then I’ll add knobs and switch caps.

The panel for this one was a little tricky, because the drill holes for the LEDs were just big enough to allow the LEDs to slip out of place when I turned the module over for soldering. I got them to stick by bending kinks in the legs when the LEDs were in the right position. This is a bit weird, because the LED drill holes were perfectly snug for 3mm LEDs on everything I’ve built so far - 292, 207, 245, 266.

This module just received a new board revision and a new BOM to go with it. Some of the components on the BOM didn’t have footprints on the board. So I wasn’t missing anything, but I did end up over-ordering parts because of that. 

[source-of-uncertainty]

Work-in-progress photos of the 285r frequency shifter/balanced modulator and 281r quad function generator. The 285r is actually finished now, and I’ll get pictures and impressions up soon. The 281r is on the bench right now. It’s a dense build and will have a lot of wiring for its panel size, but I’m excited to finally get it done. I started this project while the 281 clone was in the middle of a revision, and it felt like ages for the PCB to be released and for the BOM to be published.

You can see some fine surface-mount soldering work on the 285r than I’d done so far. I used to be really afraid of surface-mount work because of what I perceived as my poor fine motor skills and big, clumsy hands. But thanks to some practice with Nonlinear Circuits builds, which use quite a bit of SMT, I’ve practiced and built my confidence to a point where SOIC Spin FV-1s like in the above picture are not a problem. 

[source-of-uncertainty]

The finished 266r. Definitely the densest build I’d done up to this point in the format. I was missing a few of the CMOS ICs when I’d finished the rest of the module, but once I added those everything worked great. 

The Electric Druid white noise IC used in this build produces noise digitally, rather than sourcing noise from a transistor or a more primitive shift-register digital noise source. Analog noise circuits pick up some nasty popcorn noise in my apartment, but the noise from this IC is free of any of that and has a clean spectrum. 

[source-of-uncertainty]

The finished 207r mixer. This was a pretty easy build overall, though getting the panel-underside nuts on the switches in the right positions was a little tricky. Went together without a hitch and worked on the first powerup. I did add the SMT resistors to make the slider pot curves logarithmic, and I’m glad I did. I matched for gain the pairs of transistors on the channels with VC panning. 

I wish those sliders weren’t so expensive to get for non-Europeans. They feel really nice. 

[source-of-uncertainty]

Current status of the 266: most components stuffed, sockets populated, vactrols still need to be added. Wiring, pots, and final assembly need to be done. I’ll set aside a weekend for this - maybe the Sunday after I get back in a week. I ran out of open-frame sockets, so I dipped into my stock of the other style. I took a photo of the PCB so I could have the reference designators on hand before I blocked them out of sight. I’ll have to make a bulk order of a lot more open-frame sockets.

My delayed shipment of boards for the 285 frequency shifter and 257 control voltage processor should be on their way soon, and the revised boards for the 281 quad function generator should be available soon, too. 

I’ve had a 208r/218r build on my mind lately - aka a Music Easel - but I’m going to hold off on thinking about that any further for a while. The case would be the biggest hassle with with that project, because I’d have to either track down an official, current BEMI case for it or adapt a larger Halliburton Zero suitcase at some expense. Definitely going to wait on that. 

[source-of-uncertainty]

I made a short recording with the 258 and 292 last week. Everything is controlled by a Blippoo Box, but I think the 245′s pulser is in there generating an envelope as well. 

[source-of-uncertainty]

I started on the 266 Source of Uncertainty last night, since the BOM for this revision has been released, and I’d ordered parts based on the previous revision, making guesses at what would be different based on what I knew. I wasn’t 100% accurate, but I only had to order a few more parts (so far). This revision (which has two boards instead of three) omits most of the on-board voltage regulators and uses surface-mount capacitors mounted on the solder side to save space. It also swaps the many tantalum capacitors on the previous revision for electrolytics. One part I was missing was the 10V voltage regulator, which is a surface-mount component in this version (note the big copper pad visible on the right-hand board). 

This is another big build, and it’s more dense than the 245. I don’t think I’ll finish it before I go on vacation this Friday, but that’s okay. Yesterday my sliders for the 207 mixer arrived from Germany, so I’ll be able to finish that up soon, and then I’ll be at or a little past the halfway point for the modules I want to build for this system.

[source-of-uncertainty]

The 245 sequencer is finished and working. In the process of debugging, I found that I’d managed to put the LEDs in backwards, so I pulled the control board far enough apart from the panel to desolder, remove, and replace the LEDs. I had to reduce the specified LED resistor values to get them up to a brightness I liked. 

Everything seemed to be working, except the pulser wouldn’t run by itself. I could use it to manually advance the stage, but its internal oscillator wouldn’t start. When I was playing with the start/stop switches, I deduced that the stop switch was likely wired backwards, which is weird because for both switches I had the same colors of wire soldered to the same pad positions. This would seem to indicate that the stop switch’s PCB footprint is connected in reverse. I switched the wires for the top and bottom contacts of the switch and all was well. 

This was a long build, with a lot of resistors and a lot of panel assembly, but it’s handsome and feels sort of luxurious to use. It’s easy to use it to tweak knobs and subtly change a sequence, making some steps louder and softer and tweaking their timbres.

[source-of-uncertainty]

The UA726 chips I ordered a month ago finally arrived... and they're fakes. Not one of them worked in the 258 oscillator I bought them for, and in fact one of them ended up damaging some neighboring transistors on the board. Thankfully I had replacements for them at hand. The 258 will still work, as I've got pairs of 2N3904 oscillators installed in the UA726 sockets, but they just won't be temperature-compensated. A company called Xvive is reviving production of the UA726 soon, so I'll probably buy a pair when I can. Meanwhile, my 245 build is delayed while I wait for switches that will fit the panel I have, and my new 266 boards are still waiting for a BOM to be released so I can start on them.

[source-of-uncertainty]

The 292 is working! I didn’t take pictures the night I finished it, but it went together without a hitch and sounds great. I tested it out with the 258 and Euro modules to help out. I keep meaning to record some audio, and I’ll do that soon.

Over the weekend I finished a Eurorack adaptation of the 281 Quad Function Generator, which will also help out in testing, but I’m still waiting for the Buchla clone PCB to be back in stock. The 266 Source of Certainty board has shipped, and I’ve got boards and a panel for a 245 Sequential Voltage Source. 

Pictures show the panel assembly I did last night on the 207 mixer, which is now waiting for its sliders to arrive from Europe, and when those are in I’ll wire the jacks to the board. 

Last night I also started populating the 245, and I added resistors and diodes to the top PCB. The bottom PCB will take a lot of stuffing, but at least it’s not particularly dense. That one will take a lot of panel assembly, too, as it’s got 20+ pots on top and a bunch of banana jacks. 

[source-of-uncertainty]

Started building the 207 yesterday, and realized I’d missed a few capacitors from my order, and that I should probably get a few more PN3565 so that I can match the required transistor pairs more closely. The preamp transformer (which goes in the big hole at the lower left) arrived on Friday, and it’s smaller than I expected.

I also added some stabilizing caps to the opamps on the 292 to prevent self-oscillations, per Dave Brown’s solution over here. My Sputnik Quad VCF/VCA, which is a clone of the same circuit, had problems with distortion caused by oscillations, and I’m betting that it was fixed the same way when I sent it in for service.

I scored boards and a panel for the 245 sequencer on Friday. I spent a while figuring out how to handle the start and stop buttons, because the mounting hole in the current panel and the current board’s clearance don’t allow for any illuminated pushbuttons I could find. I put in a parts order yesterday, and I decided to try an NKK bushing-mount non-illuminated pushbutton that has optional colored caps and metal collars for each button. If that doesn’t work, I’ll just use momentary toggles like what’s suggested in the BOM.

This project feels a lot more piecemeal than the Serge box I made, thanks to the currently spotty availability of the boards and the greater variety of exotic parts required. My final module setup will end up looking something like this:

The 261e is a stand-in for the 1-space version of the 259 dual oscillator that Roman Fillipov is working on, which may or may not be available in a few months. That spot is the least certain of any spot in the rack, as of now. It might end up being the 259, another 258, a fuzzy 158 oscillator from the Buchla 100, or a 285 frequency shifter/ring mod, which can sort of act like an additional oscillator module.

While I’m waiting for boards and parts, I’ll be building a Eurorack clone of the 281 function generator for use in one of my smaller cases. But since it outputs 0-10V, I can use it with these Buchla modules I’m building, via adapters.

[source-of-uncertainty]

Started on the 292 tonight. Now that I know what to do with the panel wiring, it's a lot easier to do it first thing. Unfortunately I managed to mess up my parts order, so I had to reorder some resistors and caps, but I can wait to finish the rest of the initial stuffing. My parts order for the 207 looks like it was a lot more accurate, so I can start on that soon as well.

[source-of-uncertainty]

More panels are in! Left to right: 292, 207, 281, 266. I'm ready to start building the 292 and 207, and I'll start the 281 and 266 in a month or so.

[source-of-uncertainty]

Matching Vactrols, in order to get equal-sounding channels in the 292 and similar responses from the smooth voltage sections of the 266.