Synthesis Technology Morphing Terrarium

Quantity: 1

Here is a Synthtech e350 Morphing Terrarium DIY project.   Paul Schreiber’s description, taken from the AnalogueHaven product page.

the synthesis technology e350 morphing terrarium is a unique wavetable vco with 2 independent waveform outputs. unlike similar designs, the e350 is a fully self-contained, 1v/oct vco. what sets the e350 apart from previous wavetable oscillators is the voltage-controlled morphing.

the module comes shipped with 3 banks of 64 ‘basic’ waveforms (192 total). each bank of 64 waves is arraigned in memory as an 8 x8 array (like a chess board). the ‘rows’ are called ‘the x direction’ and the columns are ‘the y direction’. the 2 panel controls and the 2 external control voltages (morph x and morph y) are used to “point” to where in the array we use for the data that is the xy out. the third control voltage is called ‘z morph’, and it determines the waveshape at the z out. the z control voltage ‘scans’ then wavetable bank from wave #1 (think as row 1, column 1) to wave #64 (row 8, column 8). for example: if x morph is set to 2, and y morph is set to 5, the signal at xy out is the waveform stored at row 2, column 5 (the 13th wavetable). the z out is completely independent of xy out, so if the morph z knob is at 48, then the 48th wavetable ( row 6, column 8 ) is the output.

the wavetables in the 3 banks are very carefully ordered by rows. in general, scanning in the x direction will generate a smooth progression of timbre. each row, however, is different! there are sine banks, pulse width modulations, filtered noise and even male/female vocal formants. so, then you can see that scanning in the y direction (in a column) will be extremely varied. and the beauty is that you can scan in both directions simultaneously if you want.

but we were not content with just indexing through the table 1 wave at a time. instead, the e350 implements a proprietary smoothing/anti-aliasing algorithm that allows the 2 outputs to be a continuous blending of one wave to another. the algorithm calculates 128 “in-between” waveforms, so that the 192 ‘basic’ waveforms are really expanded to over 24 *thousand* individual timbres! so as you adjust the panel controls, or add external modulation (from lfos, eg, etc) the e350’s outputs are “glitch-free” and will provide you with a near infinite range of tonal possibilities. for people that like the glitches of older wavetable vcos, there is a jumper option on the pc board to add these back in.

but, we wanted to do more, so we added a 3-position range switch for using the e350 as the world’s best lfo. at the lowest setting, the e350 will cycle once every 13 minutes and be repeatable within 20usec. the sync signal can be used to reset both outputs to zero (‘hard sync’) like any other vco. the ‘c’ bank of waveforms are specifically designed to be used in lfo settings (although they will output audio timbres as well). the lfo waveshapes are everything from ‘chirps’, sequences of individual notes to noise gates, s&h effects to exponential/log sweeps. like quadrature outputs? the e350 can have variable phase outputs under cv control.

there is nothing like the e350 in any modular format. in many patches a vcf is not needed: the timbres that are generated are already smooth and harmonically “correct”. from the precise tracking to over 24,000 waveforms possible, the e350 sets the standard in modular oscillators.

Panel Design

You can download my

Front Panel Express Morphing Terrarium Panel

Yes, I just had to fit this into a 1U panel with small knobs.  I located all the knobs at the top, the jacks at bottom, and the two switches in the middle.  I admit the downside of the closeness of the knobs, having immediately discovered that fiddling with the MX (Morph X) knob risks nudging the coarse frequency pot to its left.  But I got the 1U sinewave VCO I wanted, plus 24,000 other waveforms!  For this application I used high-quality Bourns potentiometers, Mouser part 652-91A1A-B24-B18.


The dual PC boards with SMT components fit well on a Bridechamber 3-jack bracket, cut down in length. Wiring was made harder than needed because I was out of thin wire and had to use #22 with rather thick insulation.  This resulted in a bunch of wires running under the bottom board and required that the jack wires be soldered to the top of the board, contrary to instructions.  But it all fits.  Here’s a mid-assembly photo.

And all together.


Leave a Reply

Please use your real name instead of you company name or keyword spam.