393 VCO

About

Not to be confused with the VCO section of the TB-303, this there is an improved version of the core idea used in the Fluid Shape Complex VCO Pair utilizing the LM393 dual comparator chip, hence the name. This VCO is musically useful, and has both better tracking and lower part count than its precessor. The hi/lo switch sets the effective range of possible frequencies: "hi" is 0.3Hz to 5.8KHz, "lo" is very very slow to around 20hz (everything approximately, i'm no nerd). Hence, this device can both be a VCO, a VCLFO and in a pinch - even a clock generator.

The module features the FREQ knob for sweeping across the entire range and the FINE knob for adjustments inside about an octave. Two DC-coupled exponential CV inputs with attenuators are summed to FREQ along with the v/o (1 volt per octave) input for a keyboard or whatnot. The attenuated FM input provides DC-decoupled (or AC-coupled, but this one's not as correct) linear frequency modulation. It is quite potent, and the 'classic' FM tones lay on the first third of the knob rotation; past that it's screamy detuney danger zone. Finally, the PW knob sets the pulsewidth of the pulse output. An unattenuated CV input is directly summed up to it.

Under the knobs are the input jacks. The top row has CV1, CV2 and FM input, all going to respective attenuators. Below are the v/o, sync and pw cv inputs. The sync input generally expects a pulse from the outside source and harshly interrupt and resets the internal oscillation according to it. The input is sensitive to the incoming signal volume, e.g. louder pulse will produce deeper and harsher sync effect. To the right of the three inputs is the hi/lo switch mentioned before. Finally, the row of jacks at the bottom are the oscillator outputs. Left to right are: sawtooth, triangle, pulse and square. The sawtooth output is derived from the triangle core with a very simple and minimal circuit, and runs at double the frequency of the other outputs. The pulse output's duty cycle is affected by the PW knob and its CV input. The square output is always as symmetric as it gets, and, unlike the other outputs, is not centered around 0v. Instead, it oscillates between 0v and ~10v intentionally, so that it can be used as a clock/gate source for other modules. Obtaining a zero-centered square is as simple as turning the PW knob full counter-clockwise and taking the pulse output.

The oscillator tracks great 1v/o for a bit over 5 octaves of the most musically useful range (about 40Hz-3.5KHz), and does alright an octave above and below that. Outside of these ~7 octaves the tracking is gone, which is alright with me, since those frequencies aren't used a lot in the tonal musical sense. The design is simple and straightforward, requires just a few additional parts on top of "the usual", sounds great and provides great function for its low complexity. This is a great build for intermediate SDIYers who already built some 40106 VCOs and such, and are ready for more precise and full sound sources. Credit to Thomas Henry for the current source design i based myself off.

Schematic

Media

Basic sweeps triangle, saw and pulse (with pw demo). NOTE: this was before changing C3 from 10n to 3.3n, so the top cap frequency is lower than it ended up being.

lfo range demo by using the triangle out to control the fold amount on the SFP44. it gets reeeally slow at full CCW, so i decided not to include that part - just believe me it's over, like, 40 seconds.

volt-octave test

Little melody using the unit as the only sound source - tracks well across octaves, at least to my ears.

393 VCO synced to/FMmed by a different, worse-tracking v/oct VCO. Playing with sync and fm depth.

Pictures

Module
393 gutz
Module
board closeup
Module
terrible transistors+tempco heatshrink bud
Module
racked unit

This page was initially published 27 December 2022.