[sdiy] the white whale - modular synth patch memory

[rsdio at audiobanshee.com]

For a project like this, I'd start by looking at what's been done.

I don't own an ARP 2500, but it has quite an extensive matrix, doesn't it? If you look at the number of bus channels on the ARP 2500 and make sure your design has at least that capability, then you should be on your way to a workable design. Rather than physical pins, you'd have electronic control.

What about the Buchla systems? They seem to have some sort of optional patch memory that can be added to any module. I can't justify the cost of a Buchla modular, so I don't know the details, but you might want to look at the capabilities there. Do they only handle routing? Only pots? Both? Again, if you can match the capabilities of an existing system, then you've probably got a good design.

I just read Sarah's comments, and I think that's similar to what I was thinking. Your bus can have fewer lines than the total number of CV sources and destinations, but it can still work. Using a bunch of 8:1 or 16:1 analog multiplexer chips should allow you to assign a CV source to a specific bus, and then route that bus to a specific CV destination. Kind of like the Oberheim Matrix-12 can create up to 20 virtual patch cords from just about any source (27 total) to any destination (47 total), and that's in addition to the 12 "hardwired" modulation routings.


On Mar 18, 2015, at 9:14 AM, Richard Moore <rkm31415 at yahoo.com> wrote:
> Excellent!  I was dreading building a messaging protocol on top of SPI.  It's good to hear there's something better to use.

Hmm, I don't see the difference between SPI and CAN in terms of protocol. With CAN, there is no set protocol. In the auto industry, where CAN rules, you can't decipher the messages on a CAN bus because there's no standard protocol. You're just getting bytes, and there's no rule as to how to interpret these bytes. Granted, CAN has shorter packets (8 bytes?) than SPI, which can be any length.

The benefits of CAN are the multi-master aspects. There's not really any set messaging protocol. SPI would require a single master, multiple slaves, and some way to select one (or more) slaves for each packet.

I'm not saying you shouldn't use CAN. I'm just pointing out that you're still going to need to invent a messaging protocol. I suppose you could implement MIDI over CAN. That would be very interesting, but would still require assignment of controller numbers and channels. However, with 8 bytes per message, you could have a much larger number of channels, notes, controls, etc.

Brian Willoughby