It’s really pretty much a step sequencer, but with the design optimized for using it to create envelopes. There’s a trigger input with an internal clock (or it can be clocked externally). It has retriggering and looping modes. It’s Eurorack, not Kosmo, and it’s based around a PIC which comes from Barton preprogrammed and AFAIK there’s no access to source code, so I probably won’t build one but maybe consider coming up with my own Kosmo/Nano implementation. (Or one of you could beat me to it and save me the trouble…)
And yeah, it’s an actual printed front panel, not just a piece of sheet aluminum with holes in it, because it uses sliders and Barton didn’t want to have to deal with hand cutting slots, I guess!
I’d say that one can get a very workable Kosmo module just by using the existing PCB set from Barton behind a 10cm Kosmo panel.
Barton’s control panel appears to be about 8.4cm, so it should be only a matter of using the same slide potentiometer footprints and making the Kosmo panel just 8mm wider on each side. Of course, someone might want to put more space between the sliders and have them behind, say, a 15cm panel, but I don’t think it will look as nice. These LED sliders do not take a knob so they will look kinda “lost” and “puny” if spaced out. The ergonomics of having them close together also makes sense in this case, as you can flip them across to set the desired waveform with a flick of your finger. The jacks are mounted independently anyway and there are only four of them, so there is not much wiring going on. Unless, of course, one wants to flywire the potentiometers and three bottom toggle switch a bit further apart to make it a bit more balanced against a 10x20cm Kosmo panel.
Again, it’s only my personal view, but I’d say that with the cost of two PCB and the preprogrammed PIC at just $20 is far more worth designing a 10cm panel than considering the trouble of coming up with the code and everything from scratch.
Yeah, you probably wouldn’t want the sliders much further apart — though you could want more of them. Barton says the original version (in a larger format) had 10 stages but it was reduced to 8 for the Eurorack version. Presumably the PIC is programmed for 8 stages only, though.
I’d prefer a bit more spacing on the knobs.
Mainly it’s just that I’m averse to closed source firmware. And this being essentially a step sequencer, starting with the code for Sam’s sequencer would get you a good deal of the way there.
There has been similar designs around. I think MFOS had something like this and I think there can have been a Maplin module as well. But they were all analog.
Anyhow, when going to USA for Xmas I will bring home parts for this module and some other - it’s absolutely crazy.
If you’re interested in PIC microcontrollers but have no reference point because of the closed code, you might want to have a look a similar project by Tom Wiltshire of Electric Druid as a starting point.
If I understand correctly, the ED project takes a rather different approach — using an ADC to read a CV for each stage and I guess PWM for the output. This allows you to vary the output levels using control voltages, and to interpolate rather than step between levels, but you need an ADC for each stage. The Barton design just uses digital outputs going to attenuator sliders, a simpler approach using only one ADC for the speed CV.
For a tiny Eurorack module a PIC maybe makes the most sense, but for a Kosmo module a Nano seems fine. There are a lot of I/O pins used in Barton’s PIC design
but I think for an 8 stage version a Nano would have enough I/O. One pin to spare if I’ve counted correctly. A shift register for the switches would reduce the number of pins needed and allow for more stages.
Looks pretty similar to MI Stages, which is open source, though that is only 6 stages instead of 8. I didn’t dig into the docs, but it may be easy to expand.
It is definitely a complicated beast. Though, looks like two channels of the DAC are unused, so it could be simply expanded to 8 stages. ADCs also have two extra slots each for the pots. LED controller has some extra slots. Where you run out of pins is the buttons, gates, and stage LEDs. Though the design could be simplified if only the sequencer mode was desired. That would drop the need for mode buttons and LEDs. That would also drop the LED controller IC.
The more I look at it, maybe it isn’t so complicated…
Another thing is: With a Kosmo module you have more vertical room, so you can use 45 mm instead of 30 mm sliders. Gives you that much more fine control.
