Assembly is complete for the Slightly Nasty Model 1011 Discrete Oscillator. In this post I’ll describe the build, providing a few specific part numbers used, and give a couple of example sound recordings.
Features
- Sine/Triangle wave output with manual mixing between the two
- Saw/Pulse wave output with manual mixing between the two, plus sub-octave mixing
- Sub-octave square wave output
- Coarse and Fine frequency knobs
- Initial pulse width (symmetry), plus CV input with attenuator
- 1V/Oct CV input
- Log CV frequency input with attenuator
- AC-coupled linear FM input with attenuator
- Sync input
The Build
Most of the components of this all-transistor (discrete) oscillator are mounted on a main PC board. Three 10-pin headers join it to a panel PCB that holds the pots. The two PCBs are held together with standoffs at the corners. The panel is held by the pot nuts. Banana jacks mounted on the panel protrude below the panel PCB and connect by wires to the back of that board.
I departed from the recommended standoff hardware. I’ve been using header sets 11mm high, matching them up with 11mm standoffs with M3 threads. This is easier to work with than the non-threaded 10mm standoffs with washers and one long screw and nut, as given in the user manual/assembly guide. Once the jacks are soldered to the back of the panel PCB, the panel cannot be removed, making it hard to reach the screws at the corners of the board, underneath. My method uses a screw at each end, into the threaded standoff. Removing the main PCB is easy, just by removing the four screws on the back.
Power header and trim pots go on the back side of the main PCB. Four capacitors also mount on the back, eliminating any worry about clearance. Notice the white lines separating the sections of the circuit. The sections are also separated on the top side, with components numbered by the section they’re in, also corresponding to the schematic page. Very convenient for locating parts during assembly. (Please, save me from iBOMs.)
Some specific parts
- PNP transistors – Mouser 637-BC557C
- NPN transistors – Mouser 637-BC547C (recommended)
- NPN transistors that I used – Mouser 512-BC547BBU
- 11mm threaded standoffs – Mouser 710-970110324
- 10 pin header sockets – Mouser 710-61301011821
- 20 pin header pin strip – Mouser 649-1012937892001BLF
- MTA-156 power header – Mouser 571-6404454
- Trim pots – Mouser 858-67WRNKLF where N is the value
- C206 – Mouser 23PS310 Styrene
- 1K Tempco resistor – Thonk LT16S102F33
- Bold – Small Skirted – Davies Plastic Knob – Thonk
- Bold – Large Skirted – Davies Plastic Knob – Thonk
I used BC747BBU for the NPNs because I had a lot on hand. I also chose the 23PS310 Styrene capacitor for the integrator because I had one and I like using Styrene caps in critical spots.
Cosmetic changes
Finished module
I made a few minor changes. Green jacks are used for outputs A and B. I stayed with the other jack colors. The sub-octave square wave output jack red color already matched my Benjolin oscillator square wave outputs. I use blue jacks for CV inputs.
FM attenuator knob is Dark Red. The CV attenuators Dark Blue. I would have used regular Red and Blue knobs, but they weren’t available when I was ordering from Thonk. The Dark Blue are barely distinguishable from the Black knobs; I may replace them later.
Two Sound Tracks
The first track demonstrates the waveforms. It starts with the sine wave on output A with manual pitch sweeps, then the same with the triangle. These can blend for adding subtle harmonics to the sine. At around the 40 second mark there’s a jump to output B with the pulse wave. I do some manual PWM. Then the saw is blended in, followed by adding in the sub octave.
Around 2 minutes in I do some linear FM on the sine wave, using a triangle wave from another oscillator. At 2:45 I switch to the Log FM, starting with audio rate and then LFO rate modulation. After that I do some syncing to the square wave of another oscillator. This is in mono.
This second recording is just a playful patch that uses the Rungler from the Benjolin, plus an LFO on the PWM, and then blending various wave forms. Left and right channels get different waves.