All modules in the DIY SYNTHI in place! Still tweaking left but it feels like an instrument now! 
6 Likes
home built? wow how long and what about the stick pin matrix brown board, love it wow great job all around
2 Likes
Thank you! 
Time? Hmmmā¦ Daydreaming, drawing and design a couple of years. 
Then I built other stuff, slowly got better. I could not have built this five years ago. 
But the actual building from first cuts of wood and metal to all modules in place took around three months, Working entire weekends a lotā¦
Matrix, partsā¦ Yes, I got really lucky finding a used matrix on eBayā¦ took quite some timeā¦ a guy in Ukraine or Latvia sold it from old army surplus stuff for around 20 Euros.
And all of a sudden even Amazon/Ebay sold these nice 10-turn vernier dials/planet gears, that actually look almost identical to the real thing (much cheaper on eBay).
https://www.amazon.com/Matching-Button-Linear-Digital-Potentiometer/dp/B0FLDWNSDJ/
Or search āRV24YN Series Button Linear Digital Knob SB360 RV30YN Potentiometer Switch KnPā
Then, I repaired a real EMS3, took measurements and decided what I wanted most (as I found a smaller matrix, the original is 16x16)
Designed on paper and blueprints in Illustrator, A friend helped cut the holes in the 2mm aluminum panel (at his place of work, a well supplied university metal shop); printing the text was a laser print (A3) laminated at a place in town. I cut a lot of holes in the print before laminating to protect it better. Enclosing a pic from mid build, where the laminated paper is easier to see.
Same friend helped me with laser cutting the wooden box.
I figured I needed to build well so I could open it and change stuff many many times without breaking it, thus the double-frame-on-hinges design, very much inspired by the Nagra 3.
I got the idea for cable lacing again from vintage stuffā¦ ordered very supple 28 gauge wire so I could fold away all connectionsā¦
5 Likes
Another from my backlog of modules - a modified version of the Nonlinearcircuits Triple sloth module:
Getting this working properly was a bit of a saga! I started off using this schematic (for a double sloth) which uses higher value capacitors for the slower sloth. It worked well for the faster sloth on a breadboard but I couldnāt get the slower one to work, even after creating and studying a CircuitJS1 simulation. Thereās also a schematic for the triple sloth using high value resistors for the slower sloths. At the time I couldnāt find a source for for the 68M and 100M resistors needed, so put this project aside for a while. Some time later I needed to buy resistors for another project and found a Chinese supplier of high value resistors on one of the usual auction sites. With these in hand I could get the slower sloths to run satisfactorily. With the help of my circuit simulation I tweaked some of the resistance values to get the maximum possible range of control over the oscillation behaviour of all three sloths, including adding a control potentiometer to the slowest (Inertia) sloth.
There was considerable variation in amplitude of the outputs, and I wanted all outputs at Eurorack levels ā Ā±5V, so I modified the circuit to allow either amplification or attenuation of the X outputs and tuning of the amplitude of the Y & Z outputs.
As usual the details are in my Github repository, including the full schematic, a PCB layout with all the alternative resistor positions for the output gain options, and a copy of the CircuitJS1 simulation to download for local use.
6 Likes
Nicely done, good work!
Iām curious whether yourā āFrequency Modā knob does much of anything? Mine seems to be completely dead but Iām not discounting the possibility that Iām just not using it correctly either.
I just finished a new module, Moritz Kleins Shapes VCO x4 with some added features. It has an octave knob and has a LMNC link input in the back.
More info on this Instagram post I made:
Hereās the schematic:
11 Likes
where does the octave switch fit in ? I really like this project, and just curious the therm type pots are special temp sensing components ? Parts list possibly please.
ā
I see in the instagram video the large knob at the top is the said āoctaveā switch ā , but where is in exactly on the schematic you made, just curious .
The octave switch is in the top left boxes, it uses a LM4040AIZ-5 shunt voltage reference (Whoops forgot to update the schematic) and a series of voltage dividers. The switch is basically a replacement for a coarse tuning knob in the Shapes VCO, it expects -5V to GND so I had to invert it with an op amp. It just connects via a 7 pin molex.
As for the temperature compensation Iād check out Moritzās DIY VCO series, its the same circuit:
Also.. once I get my files organized, Iām thinking about sharing the source files of all my modules via github. Just not sure about the legality of it.
3 Likes
Yes, Frequency Mod knob adjusts the amount the filter cutoff CV input impacts the sound. Therefore, it only does something if you are feeding it a frequency cutoff CV.
1 Like
Hi Ben , so what about the use of the 01x07 pin header on the switch voltage dilvider and ilnverter. He has a set of 10kās used and after each one there is a large M . He also shows the switch going up and down like a pot and varies the freq knob . Is this a special switch type or a pin header that you choose one pin at a a time ilike MFOS does in his noise corocorpia ckt . I see what overall he does just curious about that section. Any help would be greatly apprectiated thanks .
youāre talking about something entirely different
I was trying to get explaination on the 4 vco module , my bad !!! Lets move on. Now looking at the instagram post i see it is a rotary switch at the top .
Just completed, a module I hadnāt intended to build - a modified version of Hagiwoās 6-channel trigger sequencer.
However, having rewritten the code for the original Hagiwo version (see the discussion in the latter part of this thread) I thought perhaps I should have one afer all. Unsurprisingly, this one is based on my revised hardware for the Euclidean Rhythms module using a hardware SPI OLED display for maximum speed. Versions of my code for the 6-channel trigger sequencer (for both original Hagiwo I2C-based hardware and my modified SPI hardware) are on Github. I built this module using the PCB layout for my version of the Euclidean Rhythms module, also on Github.
I must just mention once again the advantage of using a SPI OLED display in an application like this which demands tight timing. It takes ~3ms for the Arduino to upload a fresh screenful to the OLED using hardware SPI compared with ~37ms using I2C. That ~35ms difference could easily be a very significant proportion of the total processing time.
5 Likes
Awesome module!! My issues with the 6chn gate sequencer is ancient history now. Even with the relatively slow I2C, im not having any difficulties anymore. I think im gonna build the euclidean module too.
Iām intrigued. Does anyone happen to have a spare boardset laying around for sale or trade?
1 Like
OK, progress over the past three weeks as Iāve taken up the project again as the weather grows coolerā¦ Iām now 4 for 4 with successful AO Hero/Sidekick builds, and have a few more at around 80% done. Iāve finished a rebranded AO Reverb that I got in a trade, a 1221, an 1183, a few of CrucFXās passive low pass gates, and two runglers. After building two of Abe/AI Synthesis āHigh Powerā supplies, my next step is to build a similar scaled cabinet as my current one, and take all of my voice modules on the road to a friends house to start exploring single voice opportunities together to really focus on learning all of the modules along a single audio chain. Iām hoping to treat the exercise like the DivKid single oscillator video here: https://youtu.be/KxsvH9Oub8I, except with some modules Iām even more excited to throw in like Richās Wave Displacer and Ringer modules.
Iāve got a few PCBs left over (they have silkscreen errors on a couple of pin headers) but sending them out of the UK would be problematic. It is easy enough (if a bit tedious) to build on prototyping board though.