State of the Arduboy

The State of the Arduboy is strong.

  1. Plans to do a SAMD21 based device have been scrapped, doesn’t make sense to create a new build target if it’s not IOT.

  2. I’d like to bake the flash chip idea from @Mr.Blinky on the board and include all of the games on it. That means the product probably needs a new name, but haven’t stuck on one. It would be a good time to create a lot of PR attention and bring new people to the platform because a common complaint from casual gamer is how difficult it is to load new games. I see this as the Arduboy hardware in it’s final form, and really should have been designed this way from the beginning.

  3. Bring the Emulator and Game Library to the front page. Make the user experience a bit like PICO 8 where you can come and play games online, but then also buy the hardware if you are interested. Final form would have code writing and compiling, bitmap, sound editing all from the front page. Don’t be surprised if ads show up if that ever happens.

  4. Bring the front page experience to mobile with dedicated iOS and Android apps. Thinking a one time price of 0.99 could get rid of the ads? Still be able to code, compile and play all on the device.

  5. For a “next product” I would like to do something music focused. The current idea is to take the 32u4 and Arduboy design and make it look like one octave of a piano keyboard. @JO3RI has put a lot of work into creating a synth and tracker for the chip so I’d like to celebrate that with bespoke hardare. The packaging would be next level, and would include a really well designed booklet and music lessons etc. Would probably come out at $100 and be a low volume product. But, this hinges on some industrial design problems like thin rotary encoders that can be used on the edge, what screen to use, and audio amp hardware.

  6. The Future: The Dream: The Mission: A $5 Arduboy. Or, something like Arduboy. It will need a new name. I want to take the cheapest electronic toys on the planet and find one to put an IOT chip and open source. The goal is to bring the Arduboy experience (with a new name) to the entire world, get it into developing markets. The vision I have is that a classroom could have 1 Raspberry Pi (or more) and then a Game Widget (working title) for each student, or group of students. If possible, the kids can take the units with them when they are done with the class. This will mean a total ground up redesign of the platform. But the concept is a durable minimum viable miniature game system that can be programmed with C++ by the user.

It doesn’t have a number, because it isn’t part of my plans but something that is already in the works behind the scenes but… The MicroCard has been licensed out to a larger toy company that is going to redesign the product and put other IP on it and release it into retail markets. I can’t tell you anything more until the New York Toy Fair in February, but stay tuned for more there!

So if you’ve noticed my absence it’s because I’ve been steering this ship for 5 years now. That’s a lot of work for one dude so I needed to take a step back and frankly just have a bit of a life for once. I learned many things about myself, spirit journeys were taken and now the batteries are recharged for a new mission. I think this number 6, about the new Game Widget thing is really the most exciting concept out there.

How do you think a cheap-as-cheaps open source game system could change the world?


That all sounds good. Putting all the games on the device is obviously the way to go, but I also think that will probably mean allowing much larger games. I think a lot of the charm in the Arduboy is the 28k limit. But understandable if that limit goes away.

I would love to see this happpen. Feel free to PM me with any questions. If you’re looking for name suggestions, Arduboy FLASH or Arduboy DX(DeluXe) come to mind :slight_smile:

That makes me think about AY player / AVR-AY even though I don’t have musical talent I’d llike to see that happen.

The $5 block game. I’m imagening those chinese LCD brick games but then programmable. It’s block matrix would be nice to program.

Exiting! a Micro Card Arcade :smiley:

100% agree!

As far as I am concerned that is the #1 enhancement the Arduboy really needs, and the main reason I still haven’t bought an official one (I went the homebrew route instead: ArduBigBOY (WAS: My Homemade: Pro Micro, SSD1309, with Flash Cart)).

The 2nd enhancement I would vote for is a larger OLED screen (the 2.42" ones are excellent), but I understand that that might be too big of an additional redesign and cost increase.

If and when the flashcart version comes out, that size constraint is theoretically gone from there too. In the emulators, I think there should be a checkbox the developer can select if they want to constrain or not. Then in the Arcade the entry can show whether its a 28k game or not.

The goal of this would be to make a device specifically for someone like you (or anyone really) who doesn’t have music talent but would like to make some beep and boops. There are some examples of music hardware that is kind of “fool proof” that keeps you on time and in pentatonic scales for example, I think there would be a mode for that. But the hardware would be built in hopes that some chiptune creators would pick it up and develop some masterpieces with it. Then the songs can be shared much like the games are now.

Exactly, more or less I want to gonzo style manufacturing where I basically go to one of those factories producing that and add an IOT chip thats open to the user. My goal would be to keep the 128x64 pixel monochrome matrix, but would be deployed on an LCD.

Also, doesn’t fit on a credit card :slight_smile: Thats a great place for the homebrew.

The flash chip does’t look like flash memory to the processor. It’s more like an SD card or thumb drive. Therefore, you still have a sketch size limit (but it may be an extra 1K due to a smaller bootloader).

We can also offload game assets (i.e. bitmaps, fonts, maps, tunes, etc…) to the flash chip, leaving more of the sketch for program code.

In fact, I have a proof-of-concept running locally that offloads the primary font to the flash card that saves over 1KB per sketch.

I was going to ask if this implied it would be entirely reflashable, so you could have an extended game or if it would be static.

I like that this is kind of the same line of thought as the “new 3ds”, where as its fully comparable backwards but with slight improvements. Even if the flashcart portion was isolated and unchangeable, as in you could turn on the arduboy and boot into the flashcart loader, or boot into a “normal” mode that was the same as how the current arduboy works would be a nice upgraded product.

The only thing I wonder is how would EEPROM be handled. I don’t recall how @Mr.Blinky’s cart handles that, but having “all” the current games baked in is a much larger scope than what the cart supports right now.

Have you ever used a Stylophone? or seen pocket operators? I think those products might be similar to what you have in mind.

Also Idk how feasible it is but it would be a nice feature to export a file from it and dump it into an arduboy game.

Sure, you can play tricks like this, but the flash chip doesn’t extend processor flash space, which is how I read what @bateske is implying.

Not completely. The screen resolution is different, so DS games end up looking stretched.
There’s a button combo you can hold to disable the resolution stretching, but it still doesn’t look as good as playing a DS game on an actual DS for some reason.

But I digress…

If we had an estimated date for the baked chip, we’d be able to plan out our current/next projects to make sure they are ready in time. :slight_smile:

1 Like

Still have about 4,000 Arduboys in stock and would want to sell through the bulk of those first. I’ll see where we are at after the holiday season, but I would expect middle of next year, at least for the launch. March or April time maybe.

I’d like to get prototypes made sooner than later.


Is a 2D capacitive touch surface (implemented using the same MCU, not a separate part), in alternative to rotary encoders, out of the question?

About the sound:

Programmable voice filters are IMO what gave old consoles and personal computer chip tunes their character and it would be a great feature to have. However, a single 32u4 can’t run the sequencer, simulated oscillators and do digital filtering at the same time. Programmable analog filters mean extra parts and complicate the design. More than one MCU could be used or Phase distortion (PD) synthesis could be implemented to have sounds that approximate some filter effects using a lot less CPU cycles. Before I got involved in the ATMlib rewrite project I had done some work on a phase distortion based synth on AVR but abandoned it when I realised the wave tables required would take up too much space to be useful for games. A product dedicated specifically to music should be able to fit sequencer, PD synth and wave tables in the ~28 kbytes program memory available on the Arduboy.

EDITED: clarified my question about 2D touch surface was in alternative to other discrete controls with moving parts.

For the longest time I didn’t want to sell the Arduboy in Kit form but it seems like a lot of people are doing that as a low cost option. I wonder if it’s possible for me to source and provide the kit cheaper than it is being bought on ebay?

I think I can sell the kit for cheaper, but honestly you can probably get it done with free shipping on ebay which is something I can’t offer.

Thoughts on an official DIY non-solder or yes-solder arduboy kit?


I would love to see a kit version, although you may want to change out the QFN based ATmega32u4 for the QFP version (or re-design one around an Arduino Pro Micro which is super cheap and where most things are already soldered). I find soldering kits together to be very therapeutic.

You forget that there are big advantages to having a kit version over having to build one from scratch:

  • Get all the parts at once instead of having to source them separately
  • No confusing “do I get the I2C or SPI screen model?” and/or “did I get the right screen model?”
  • Potentially a nice guide explaining how to assemble it
  • Don’t necessarily need experience with doing electronics

In fact, I’d say the popularity of the homemade section indicates that this might be something people are interested in.
Mayb it’s at least worth giving a trial run?

Personally I have no issues with soldering because I learnt how to do it in school and I was always pretty good at it*, I just don’t have a soldering iron.
Although, I think non-soldering would be easier for beginners, and less dangerous for children (no smoldering metal to worry about).

* Not so good at understanding how what I built actually worked because the teacher didn’t explain any of that - it was literally “here’s a photo of what the board should look like when it’s finished, go off and play ‘build circuits by numbers’”. The school eventually realised they’d hired a fraud when he suddenly didn’t turn up one day, and then they struggled to find a replacement for 6 months.


Maybe make it a both kit, where it can be assembled on a bread board, but also include a PCB where the same components can then be soldered on.

That has the added benefit you can learn with the device, assemble the bread board version, then once it works you can disassemble it and try soldering. Including the pcb is only a few pennies, the important thing would be making sure all the components can be used on a breadboard and a pcb, but that’s not too hard.


I like the idea of a solder free (educational) kit version, attracts an even younger audience.

There’s no through hole version of the ATMEGA32U4 so it has to be soldered on to a PCB but it could be a small module like a (Pro) Micro.

1 Like

I’m imagining pin headers/press fit connectors for a solder free kit version. It’d be really cool to have swappable different sized screens and case shells so the same guts could be configured for different sized custom Arduboys.

The economics may not allow it but otherwise, instead of an “off the shelf” module, you could include a pre-built custom breakout board containing the 32u4, crystal, USB connector, maybe the charge circuitry or a voltage regulator (depending on the type of batteries used), etc.

This would be similar to an “off the shelf” module but would bring out all the required pins and include only what’s required for an Arduboy compatible kit. It would get around the issues with a Pro Micro (missing pins), Micro (reversed LED polarity and kind of large) or Leonardo (large, expensive and not breadboard compatible).