Interfacing multiple axoloti on one midi input

#1

i wonder if itt is possible to have multiple axoloti on one midi input (hard wired together)?

so it'll turn out like this:
midi in port > optocoupler > split opto-coupled midi signal for the required boards > every axoloti run individual patch
is this safe to do (like with voltage ratings and such)?

my goal:
so i can control multiple axoloti using only one midi in as i don't want to use midi thru (latency reasons)
like 2 axoloti for complex oscillator patches, plus one for a midi note triggered filter envelope, delay and eq's. smile
because if i run it all on one axoloti, i'm pretty sure i will have high cpu with just a few voices.

#2

just get a better midi splitter.
Using a single midi thru should not give any (noticeable) latency.

#3

Im using usb midi, with the axolotis connected on a usb hub.
i figure, usb midi is much lower than midi din.

if you need midi din, then Id have thought midi din in , then host the other board on usb to transfer data, that should be pretty quick. (1ms latency, if is full speed ... it would be 0.125 for high speed but i don't think it is)

you could then add some 'delay compensation' to get the timing correct, important if your playing from a controller/

there is hardware provision on the board for digital IO between boards, both audio and data, but that won't be available any time soon I doubt.. its quite a bit of work

#4

That is safe for a reasonable number of boards. Be sure to also connect ground together.

#5

thanks, this was the answer i was looking for smile

#6

Also, if you don't wanna go crazy, you have this already designed:

https://www.tindie.com/products/hotchk155/midisplit-1-input-6-output-midi-splitter/

Hotchk155 MIDI devices are pretty cool, and they perform quite good, as all my MIDI crazy friends say...

And he shares information on the design:

https://github.com/hotchk155/MIDI-Splitter/wiki/Hardware-Information

#7

That midisplitter looks nice. And good price too smile

I was wondering if anyone knows if it is possible to send midi out of both the USB and the midi DIN from Axoloti at the same time?

Anyway, going to test tonight after kids are put to bed.... If it works it is really cool. Then you have 2 midi outputs from Axoloti to start with smile

#8

yes, they are unrelated

#9

I just tried sending midi out through the the big USB port. Not the micro USB port for the editor. I used the "USB host port 1" output of the midi/out/note object and it is connected directly to Blofelds USB port. Blofeld blinks as it receives the midi notes, but not all of the notes play. It works fine when I use the midi DIN, though. ANyway, going to try again tomorrow.

#10

sorry for digging this old topic up again, but what exactly is "a reasonable number", is it like 3, 5, 10?
also, how many axoloti cores* can be powered upon one of the axoloti psu's before it's unsafe?

*=no usb device hooked up, no gpio used, just plain axoloti core boards

#11

I think powering 5 axoloti core boards from 1 of the switching power supplies on the midi/psu part should be fine, provided you're not using the USB host port with high-current devices.

#12

so ehm...
I got 5 boards, a Quneo midi controller and a 10000mA usb-powerbank, giving 5v at 2.4A over two usb-outputs...
Is the output of the powerbank enough to power this all?
I know there's enough energy in it as it could run 2 axoloti's and the Quneo for over 24 hours while still having half of the accu left, but not sure whether it could drive this all together at the same time.
I was planning to use an USB-hub to power multiple axoloti's from a single USB output from the powerbank and then use the other output of the powerbank to power the axoloti that also powers the Quneo. So that would be 4 axoloti's on a single usb-power-hub, while the other output is powering an axoloti and a quite power-hungry midi controller, because of all the LEDs.

Next to this, I'm wondering about how to connect the midi between the modules.
The plan here was to use a single axoloti as a general midi sequencer and midi-clock generator.
Then hardwire the midi-output of this board to the midi-input of 2 other boards (1->2 and 1->3) and have the midi-output of these boards hardwired to the last 2 boards (2->4 and 3->5).
Thing is, I read here about the use of opto-isolators and I'm not sure I got the right one here (got a bunch of old electronics that I can use as scrap)... is it actually necessary to use these when splitting the midi signal only into 2 outputs?

eg. wouldn't it be possible to just use diodes, driven by the midi signal, to switch the signal on/off?
and ps. I thought midi-signals were like "bitwise" signals, just sequences of on/off signals at a certain rate..? if that would be so, just re-amplifying would be enough I'ld say...

#13

nobody?
I'm getting close to finalising the case and connecting the axoloti's... would be a bummer if I'ld screw up my boards..

#14

btw, I've found this wonderful schematic for making a midi-thru circuit with lots of midi outputs.
thinking about 4 midi outputs for the 4 other axoloti's and one/two external midi output(s).
am I right to conclude that this circuit could be powered through the 5v socket on the axoloti?

#15

There are many really cheap all ready existing solutions you could check out first. Maybe you could take one of those apart and put it into your machine.

I got this one:

He also made this one, which is probbaly easier implementing into something than the round one:

#16

this is the schematic for one of these hubs and I thought, since I already have the parts... why buy it?

#17

@SirSickSik did you end up using a thru circuit?
my issue is this: i want to use midi-over-usb (for the higher speed?) coming from my controller to the first axoloti which does some hardware managegment (OLEDs etc) and most importantly, the polyphony handling for the 2-3 synth-voice-axolotis coming afterwards.

i want to keep the communication baud rate as high as possible because i am using a ton of 14bit NRPN messages, so lots of midi data. since there is only one serial pin pair on the GPIOs, and i don't want to chain the synth-axo's serially, i am not sure how to best connect all these.

would these basic "through" circuits also work for a different direct serial connection at a higher baudrate? ie. i use @Hugo's additional midi i/o object and just set the GPIO/serial object to 250k?

lastly, i understood correctly that i don't need the infamous optocouplers when i do a serial connection between axo PA2/3 pairs?

wait i just had an idea, can anybody comment on this? since usb over midi is so much faster (1.5mbit at least vs serial, at most 0.25mbit), wouldn't i maybe still be faster if i just chain my 3-4 axolotis via USB? output on the micro jack, going into the next axolotis big host jack.

#18

no, I still haven't made a midi thru circuit.
I've been trying to "fix" my midi problems by creating those preset/randomisation/pattern generators, so I only have to control some "live-tweaks" and keep midi communication to a minimum (next to this, the 5-axoloti case has been on a hold for a while as I was writing modules to use in a band). Though, I admit, this morning I've been thinking to take on the project again.

As for the higher baudrate.. I wouldn't know, haven't tried it myself.. theoretically, there shouldn't be a maximum baudrate as long as the components (both axoloti and optocoupler) can handle the rate and still get from 0 to 1 quick enough. In the end, the axoloti will set the maximum, as it probably won't be able to send any faster after some point. The optocouplers might get really fast I guess, if you're willing to pay for it.

Using the digital IO's for communication would also be an idea indeed. This doesn't need an optocoupler as you'ld use voltage to send over the data instead of the current. Though, when connecting multiple receiving boards, you might need a voltage buffer again to keep the voltage high enough when splitting the signal.

As for the midi over usb. I assume you should be able to send way more midi data without choking it. As long as you can make all the connections that are needed, this might work. My problem is that my midi data is coming over from usb as well, so I cannot also send it from the first axoloti over to the next using usb (at least not from the one that the midi controller is connected to).

It would save quite a bit if we could at least use an USB-hub to split the signal to multiple axoloti's, but for that we need to wait on johannes to make that possible. I think this would fix problems for a LOT of people.

Though, this is giving me an idea for some dedicated IO modules to send over data serially. Just like midi, but then with customisable data resolution and the possibility of using different IO's for different midi data types or channels. This will use up the digital IO's, but as long as you control it with midi, this shouldn't matter. Also, this way you can divide up the data to send it over more quickly, eg. using a seperate connections for midi notes, clock and controls to keep their latency as low as possible.

#19

now creating a control 2 serial converter module. Inside a patch it's working, haven't tested it yet between 2 axoloti's.
But what I've got working for now:
-channel: like midi, up to 16 channels code wise. Of course, you could use multiple modules and multiple GPIOs to add more connections.
-type (currently only control), but will have 0=note&velocity, 1=1x16 bit control, 2=2x8bit control,3=list)
-controller index: up to 1024 controls
-value being send (depending on type, bits are interpreted differently)

so, if using a different gpio for each axoloti being connected, you could use 16*1024 controllers for each axoloti, each 16bit resolution.

don't know whether it will actually work between two or more axoloti's though haha
and another update.. it's already far slower then midi.. at least within the patch using control rate and sending single bits at the time

#20

SSS - Iā€™m trying to picture what you are attempting. Are you sending one bit at a time over GPIO? How does receiver know when to read input?