Audio Hardware Design Discussion

It is time. I’ve started working on eagle (I swore I would use kicad but…) again on the developer kit for the future. But my electrical engineering skills are lacking especially when it comes to analog.

I want to use the SAMD21 for the basis of the new hardware, but I’m having some trouble trying to decide how to handle the audio.

The setup should have both input and output. Here is a list of options I’ve thought about:

  • Connecting the ADC and DAC pins directly to the mic and speaker. (Basically how things are implemented currently)
  • Adding some filtering to the above scenario
  • Adding just a single output transistor to the speaker with the filters
  • Using some class of amplifier circuit to deliver proper line level and quality speaker output
  • Using dedicated amplifier IC
  • Using I2S audio chip for input and output

In general I’ve never worked with microphones so I’m not sure what kind of levels if just tying it to an ADC channel would even work at all. I’m guessing you might need an amplifier going into the input as well?

So i could use help evaluating which on of these directions to go, and then once that is decided some help in developing the circuit. People are picky about audio I want to get it right, and I realize it was a weak spot on the last hardware.

I’d like to put in an actual speaker, and an audio jack also.

Audio techs, assemble!


I’m not sure if I missed something, but I think it’d be hard to make any suggestions without knowing more about where you want to take this as a product. Is it supposed to be something portable (like an arduboy) or hooked up to other equipment (like most audio gear)?

When you say you want a microphone input, is that a little built-in mic, or are you thinking of plugging in audio gear (real mics, guitars, etc) and supplying phantom power?

As for output, there’s also a wide range of options. Depending on who you’re targeting, simply plugging a speaker into a pin might be good enough… or various pins with different filters so the channels sound different… or you might need a dedicated DAC (like the one in a Pocket Operator).

I imagine people that are in to this tend to have a bunch of devices, each for a specific sound, so it might be handy to have a “raw” output: simply forward the amplified input without an A-D-A conversion, so multiple devices can be daisy-chained without too much quality loss.

Then there’s the matter of how you’re going to interact with the device. Lots of buttons (Pocket Operator)? Code? Plug in a Midi controller? Plug in an SD card with samples and patterns?

Would you like a screen with that?

Both in the sense that I’m trying to build this in the mindset it could be “semi-pro” or even professional quality music gear. Obviously there is going to be a balance to consider here on price versus performance. But the concept is elevate the level of the hardware to something that someone who is currently in the GBA chiptune scene or otherwise might take notice? I’d like to be able to plug in a microphone and a speaker and use the device to do some sampling and mixing. Your comparison to Pocket Operator is a fair one, and I would say something not quite as serious as that but would be a perfect companion to.

I suppose the real use case for the mic/speaker is plugging in an apple headset and being able to use that to it’s fullest potential. In my mind that means you need drivers on both the input and output.

That is kind of my feeling too, it keeps it similar to the original design aesthetic of the original device. It gives the user a greater directly manipulating the bits within the framework of working within one chip.

The point that the Pocket Operaterator has a dedicated DAC kind of points me in the direction we should probably have a lower level operation than that.

More to be seen on this soon.

Yeah, Midi over USB. And samples yes. A big part of my hope is to have enough fidelity and speed to transpose samples recorded on the device.

I hope I’m not opening pandoras box here microboard%20display


What a total wasteland! Does that mean nobody cares about sound?


Personally, I’m not too pumped about this project. It seems kind of meh to me. Im sure there are some people who would be more into this thought.

1 Like

I’ll be frank - my Arduboy is almost permenantly muted.
Sound is always last on my list when making a game because it’s the one thing I’m inept at.

I’ll chip in if the discussion moves to RAM, screen specs, microSD support or multiplayer support, but I struggle to take an interest in sound.

I’m sure there are some people who are interested in sound though.


We don’t have much info to go bye other that that it will be used with the intent of making music. If we had more information, I’m sure people would take an interest.

The right circle is a rotary encoder, the left circle is a potentiometer. There are a lot more buttons.

Looks good so far. That gives me a pretty clear idea of what this is gonna look like.
EDIT: how many buttons are there going to be exactly?

12 notes, 2 function keys, 1 click on the rotary encoder

so in total 15 “buttons”, but 17 inputs if you count “up” and “down” ticks of the encoder.

I think the potentiometer will be a simple volume knob in series with the audio output.

And, if it isn’t obvious this is the developer kit that will be an arduino sheild. Theoretically giving us the opportunity to play around with different platforms. One target is the arduino zero which is great because you can flash the bootloader on that. But other people could save money and get the metro zero from adafruit which ditches the dfu stuff.

1 Like

I care :smiley:

1 Like

Even though making chiptune music isn’t really on my list of things to do, if there is a kickstarter campaign, you can count on me backing.

Now this is the real key here, to try and make it really fun and easy to just pick it up and just start making something pretty catchy and cool within just a few seconds. Obviously the firmware it ships with will have a big impact on this, which is why I’m trying to get the developer kit baked.

The hardware will work great on it’s own, but again the concept is this can also be the platform for creating music for the next Arduboy… and in fact be fully compatible with it.

It will be it’s own format, with likely a 128x32 pixel display. But in addition to music I totally expect plenty of people to make games for this platform as well.

But, I don’t really know what I’m doing with chiptune. A few users here are already showing interest, and once there are dev kits to hand out… well… maybe I’m starting to share too much. :slight_smile:

1 Like

Well, I’ll be interested in seeing where this project will go. I have some friends who are into making music, and whenever I take a crack at it, it ends up sounding the way the color barf green looks. I think that you should just stick to the chiptune part, and look elsewhere on the web for people who would be into this. Most people here are really into game making and coding, which is why I say a kickstarter may help, along with making money to fund the project.

EDIT: also, as they say; don’t count your chickens before they hatch.

1 Like

Dear EE folks how am I looking?

Well, I think we only have mono output, for starters. Noob. Who gave this guy an eagle cad license?

If it’s going to be an arduino shield, why not use through hole headers? They’re better for mechanical strength. You could rotate the buttons 90 degrees (or 45 degrees for a fancy design) so they can positions a bit lower.

That one function button is a bit close to that rotary switch through hole via. It may be difficult to solder properly.

Just thinking out loud: When moving the function buttons above the ‘black’ note buttons. One at the left and one at the right. they could be (ab)used as D-pads :smiley:
In addition to that when using small (0,96") 128x64 display instead of the 128x32 it could be made Arduboy compatible. Heck call it Arduboy Piano :smiley:

Cuz they are ugly, and they make pokey spots on top that hurt my fingers. Aesthetically I don’t like all the shiny row of spots. Also makes it harder to route because it kills the top layer too.

Yes purposing keys as kind of “Defined” arrow keys will be part of the plan, more details on that soon. I’ll probably share the button topology next. Someone on twitter pointed me at a chip that could use the remote buttons on mobile headsets too that includes the mic amp, so… more input the better.

1 Like

Someone had suggested using this chip, the tpa6166

It has the ability to sense headphone button inputs, has the mic and headphone amps… it’s exactly what is needed but costs a bit more than a dollar… and it’s bga so the board layout looks like a nightmare. But, I can always get someone else to do that.

Any one else have feelings on this part?

1 Like

I haven’t had an in depth look but since it’s a BGA part. You’d probably have to move to a 4 layer PCB inceasing costs.