So if one was inclined to put 5 shruthi in one enclosure for a 5 voice poly chain, is there a better way to connect them other than Midi? While on the subject, slave units can be run without controls etc correct?
afaik you still need the controls on the slave unit unless you never want to change anything. you’d need to bridge the pots so they’re not floating. but there wouldn’t be a way to change eg the matrix (and other settings not accessible via CC / NRPN) if i got that right.
about the midi connections, that would be good to know. i know some people did wire their gm5 inside say a sid, but i’m not sure whether they needed the opto then.
- The LT1054 based supply used in the current filter boards cannot be used to power more than 2 filter boards. If you want to power many filter boards from the same +/- 5V supply, use something more robust with a transformer or an AC wallwart.
- You can connect the bare TX pin of a micro to the bare RX of another. This removes the need for MIDI sockets, cables, opto etc.
- If you don’t need to change the MIDI settings and chain #, the slaves can be built without controls. The mod matrix is accessible by NRPN and is correctly sync between a polychain master and slaves.
- With a bit of firmware hacking you could “overclock” the MIDI out of the master and the MIDI in/out of the slaves to 62500 bauds or 125000 to reduce the latency.
I don’t think any of these would be very good ideas beyond the “one off” hack. The result would be a half-baked polysynth. The real thing would have to be designed using very different principles…
Sorry, should have been clear on the power issue, each board would have each own +/- 5V but a bigger supply is no big deal if there are any benefits. I could use the master to set up whatever is needed for each slave and just move the AVR to a different location (or is it EEPROM?)
What are the obvious drawbacks? This was something I was thinking about building for myself but I have yet to do a poly chain so I dont really know much beyond what I have read
MIDI settings are not stored on the eeprom.
- No note sequencer / arpeggiator (they do not play well with the polyphonic note forwarding thing)
- No LFO and step sequence sync (each voice will have its LFOs / step sequences freely running, out of sync with the others)
- Extra latency when propagating a knob move, program change or any MIDI message down the chain