Big MIDI mixingconsole project

Status
Not open for further replies.

WillieA

New member
Hello. I have no exprerience when it comes to Arduino, but after a lot of google i ended up here:) i wanted a mixingconsole for my studio, a real analog one costs millions, so i want to build my own one, looking like a real one but just a MIDI controller.
With the tiny knowledge i have i NEED some help😅
I want a 32 channel mixer with somewere anround 24 pots per channel. = 768 pots😅 can i use the teensy 3.6 and just use a lot of 16channel multiplexers? Or is there some vetter way? Thank you! /willie
 
Okey, i have spend the last 6 hours on this forum and now understand that many guys have tried to build big MIDI controllers, is this even doable? One aspect that i thought would make it easier was that there is no hi speed reading of pots. Any delay smaller then 500 milisec is not an issue. i don’t want to repeat anyones thred but i haven’t found the answer yet. I know i am a noob at this but i work a lot with music and this controller would really speed up my workflow. There is many MIDI controllers out there to buy, but they are all to small. Be able to adjust all the audio plugins with real knobs would be fantastic!

I’ve guess that running over 700 pots is not easy and that the USB powersuppy is not enough. Can i just connect a 3.3v or 5v or what ever i need to run transformer with enough amps to give the pots what they need?

A lot of dumb questions i guess but hope you understand them😄😄 all the best / willie
 
I highly recommend you build at least one fairly simple MIDI controller, wiring only 1 pot to each analog input pin. Use the example code and get it working. Yeah, 10 or 12 pots is nowhere close to your goal of over 700, but in doing 1 the simple way you'll gain valuable experience.

While connecting a huge number of pots to a single microcontroller is theoretically possible, doing it reliably in practice is harder than it sounds. The toughest parts are often the mundane things, like simply connecting so many wires. Many people make the mistake of focusing only on the cost of the electronics, only to later realize the cost of wires, connectors, front panels, nut & bolts and other physical hardware adds up fast. Often spending more on that stuff makes this whole project much easier and more reliable, and skimping on those mundane things turns the whole thing into a painfully frustrating experience.

So please, do yourself a huge favor and build a small scale controller first. Gain experience with Arduino and wiring up the pots. Then start planning the big project once you've had those first learning experiences. Then you can make decisions and ask more informed questions about how to scale up based on your first hand experience getting a fairly small controller working.

Also consider, one way to scale up is to build many smaller controllers and plug then all together with USB hubs. While that costs more, it has 2 huge advantages. First, when something goes wrong, it probably takes down only 1 small controller which is much easier to diagnose and fix (or replace, if you make spares). Second, building that way lets to have a working system right away and grow it as you have time to add more. If you go for a complicated all-or-nothing approach where you have to build a tremendous amount before you can use any of it, the odds of ending up with nothing working are much worse.
 
A big thank you for reply. Yes i was going to build i channelstrip to try everyting out. The idea of using multiple teensys an link them together with USB hub is great! That was my first intention (spelling?) but then i saw the mux and was thinking that would be easier. But that is some great advise. I have a lot of experience building analog equipment so the soldering feels like the easy part😅 thank you again! Worth so much to just have someone to answer ones beginners questions!
 
I thoroughly concur with paul. Steps to your goal -

- get a breadboard with mutiple pots working just wiring them to analog pins.
- get a breadboard with one MUX and one pot working
- get a breadboard with one mux and multiple pots working (as many as your breadbaord practically allows)

at this point i'd strongly suggest you consider designing a PCB. Simple two sided PCBs come from china super cheap these days, and you could make life alot easier. Particularly if there's repeatability in your 3D design (which I guarantee there is) you can make assembly soooo much easier! You could design a PCB per channel, with it's mux built in. All of a sudden you're only wiring up 4-8 wires per channel (depending on the protocol you use). You could/should aim for another intermediate goal (like paul says), of perhaps 6 or 8 channels. At this point you can decide how best to scale up.

I'd also stress that when I say working I mean working in your application, or as close to this as possible. Many people (including myself) have made the mistake of getting some numbers printing to the screen that seem to change when a knob is turned and moving on straight away. It depends on the software you're using, but get something as close as possible - ie get your first knob to control the master volume. then lets say you've got 10 knobs on your breadboard. Make that control two channels - volume, gain, 3 band EQ.

Good luck!
 
I agree with what's been said. This box started out as one of those white breadboards and two pots. Then one encoder was added. Then I wired up the first perf board proto. It had an audio shield strapped on a Teensy, with 3 buttons and 3 pots. It was electrically identical to the board used in the audio library examples so that I could do all of those experiments to learn this stuff. FirstSynthBreadboard_x.jpg

Then I added more pots, three encoders and Jacks for CV and gate inputs. All of these were added one or two at a time, then the code was written and tested before going further. AnalogSynthBreadboard_2_x.jpgAnalogSynthBreadboard_x.jpg

Once all of that was working I decided to make the jump to lots of pots. In my case 48 pots, 4 encoders, and analog CV and gate inputs. The black chips are 8 to 1 muxes. They get 48 pots into 6 inputs and require 3 outputs to drive them. I made 4 VCO's, two VCF's two ADSR's and all the mixers needed to make a stereo synthesizer into a T3.2. Later on I rebuilt the main board to work with a T3.2 or a T3.6. NuCPU_front_x.jpgNuCPU_back_x.jpgNewCPU_3_x.jpgNewCPU_2_x.jpg

There has been some discussion of the "controller" and its 32 inputs with 24 pots per channel. I believe that it's doable, and probably best done as 8 X 4 channel boards running through a USB hub.

I assume that this must control a "real" mixer somewhere, probably in a DAW like Ableton Live or FL studio. Some serious consideration must be given as to how you are going to get 32 analog signals INTO the PC for the DAW to work with, and the processing power needed to handle 32 audio streams while doing all the typical DAW stuff, like running several soft synths. This is not an easy or cheap task, and again can best be broken up into small tasks. My old setup used a pair of very old M-Audio Delta 1010 boards, each with 8 channels in and out. It has been working great for 10 years, but it requires a PCI slot (NOT PCIe). New PC's do not have PCI, and Windows 10 does not support it. I have recently switched to a Focusrite USB 8 channel box with outboard 8 input analog mixers on two of the channels to mix the outputs of my synths.

Again you can build one channel strip out of the 4 and get it all to work with the DAW or whatever you intend to control before going further. It would also be wise to keep a look out for used analog mixers. I have seen several for sale cheap now that the trend favors the new $$$$ digital stuff. I got a 32 channel Mackie for $100 at a ham radio show a couple of years ago. I haven't decided yet whether to rebuild it or gut it and use the pots for a controller like what you want to build.

One more thing. I used 1 meg pots for my synth because I had a bunch. A value this high can lead to inaccuracies when muxing them all because of the input capacitance of the A/D converters. I changed to 10K pots and simply wired all of them to the 3.3 volt output on the Teensy. Each 10K pot eats 1/3 of a milliamp, so there is enough current for a lot of them.
 
You have given good advice! I've tried connecting many many pots to Teensy in the past and it was a nightmare.
I made my way around this by learning Eagle and its made things much easier.

I recently made an Arduino midi controller that has quite a few pots. I used a pcb board manufacturer for this project and it worked like a charm. In a previous version they were all connected using ribbon cables...nightmare....my newest version simply connects the pots to my Arduino using a some nifty little bridges...which also keeps things together nice and tight ..here's a video of the prototype i made which wasn't as physically stable (I messed up the board measurements...live and learn!) - https://www.youtube.com/watch?v=nLoY2vf10Oc&t=5s


I'm switching back to teensy from arduino because teensy seems to be much faster and am waiting for my 4.1 to arrive in the mail.

I have already completed my designs for a controller that has about 75x10k sliding pots (https://www.aliexpress.com/item/4000536592351.html?spm=a2g0s.9042311.0.0.73b94c4dA2TqWn)
1x ribbon pot (https://www.sparkfun.com/products/8681)
about 20 switches and
and 1x oled display https://www.ebay.ca/itm/IIC-0-91-0-...284864?hash=item263f365e00:g:6QYAAOSwgdZe-zH4

....Im worried about the power consumption though and am not sure if the teensy will have the juice to power that many things.

Is there an easy way to calculate the amount of power the pots/screen will consume and weather or not the teensy has the juice to handle all them?
 
Each of your 75 x 10K pots will be across the 3.3VDC to GND, so if I am doing my math right, they will together all draw a constant [ 75 * (3.3 / 10000) ] = 24.75 mA of current. You will get slightly more current draw because of the input pin(s) being used to read the pots, but the 24.75 mA is a start as to what you will be expecting from your 3.3VDC regulator. You should be able to get the power draw of the display/screen from the datasheet.

In my TeensyMIDIPolySynth project, I had 48 x 50K pots + 48 high-efficiency extra-bright LEDs + 2 x 7-character LED displays. All of these, plus the Teensy 4.1 running everything results in a little over 100mA total draw on the 3.3VDC regulator, which is very reasonable.

Good luck & have fun !!

Mark J Culross
KD5RXT
 
Status
Not open for further replies.
Back
Top