Arduboy 2 My Ultimate Hardware Wishlist

I have only 2 wishes:
More Memory (maybe outsource to SD card) or SD card Support to load the Games
and Color screen. That´s it.
Have anybody called “Colecovision” or “Intellivision” ?

1 Like

I agree with just about everyone here. SD so we don’t have to worry about overall scope of the game, and a couple more colors would be nice. If it had the capability of making something like this I’d be pretty excited. Not being able to implement a new feature because the code alone takes to much storage space is kind of a drag. Restricts the potential to a little mini game player.


I’ve pledged for the Tau board on Kickstarter. It’s Arduino Zero compatible. 32 bit ARM M0+, 48MHz, 128KB flash, 16KB RAM, DAC for audio.

When (if) I get it, I might try putting it on my breadboarded system, in place of the Arduino Micro and see if I can port the Arduboy library to it. I would be interesting to see how many Arduboy sketches would “drop in”.


I would like a color screen, either an SD card slot or a Meg or two of flash, and start/select buttons.


I am hyped for my arduboy for longer now, since the ks, I am following it and now finally had the money to pre order one, will cost me basically 70€ to get one, so i hope you are appreciate that @bateske :D.

I read through the thread and I also have already a wishlist for the next gen of arduboy, but also I think its important to keep the “trademark” of arduboy. I think things like bigger screen should be limited, I mean to recognize the product, the core of it.

I agree with that totally.

oh wow, excited to see what it look like.


SD card is almost a given, the question is if the processor is an M-0 or if we add bluetooth maybe?

And also, I’m looking pretty hard and something that can support pico-8. But there are some serious challenges in doing color and faster processor, namely power consumption. Assuming I could fit everything on the board (serious challenge) the battery we use now would only last for about an hour, maybe 2.\

Once we get this arduboy all up and running, it will go to retail and once sales are moving well enough that I can work on something else then expect to see some limited editions of some new developer kits. I’ll probably make like 50 or 100 units of something and see how people react to it, then try again with another set of kits until people really like it.

There is a lot of activity here in the community of people building their own stuff, especially @davidperrenoud so if one of these hardware gets enough traction we can support it and find a way to get the people involved in on the piece of pie somehow.


I’ve stated a lot of this before:

IMHO I think the Arduboy is such a success because it didn’t strive to be the ultimate portable gaming platform. One main appeal was the challenge of coming up with decent games given the simplicity of the product. Another was the large code and support base provided by being Arduino compatible.

Although there have been some desires for color, communication for multiplayer, and other enhancements, these things start to get away from the goal of being an easy, fun, basic thing for beginners to learn with and to test the abilities of seasoned programmers without much effort.

To me, two main complaints are the limited code and RAM space available, and only allowing one game to be resident at a time. To some extent, the lack of speed may be a minor factor.

With that in mind, if it were me, for the next “Arduboy” I’d just try replacing the ATmega32U4 processor with an ATSAMD21G18 and add a MicroSD slot. Keep the display, buttons, case, battery and most other things identical to the current Arduboy. This would mean:

  • Flash increase from 32K to 256K
  • RAM increase from 2.5K to 32K
  • Clock speed from 16MHz to 48MHz
  • (However, none of the above will be precisely comparable, since we’re going from an 8 bit AVR architecture to a 32 bit ARM Cortex M0+, but still a big improvement)
  • The processor would probably drop in fairly easily, although a small 3.3V regulator would likely have to be added.
  • The total cost difference from the current Arduboy would probably be minimal
  • The processor is the same one used by the Arduino Zero so the goal would be to make it appear as an Arduino Zero (much like the current Arduboy is a Leonardo). We would still be under the Arduino environment so there wouldn’t be much of a learning curve for developers and porting the Arduboy Library and other aspects would likely be quite painless. Many sketches would probably port over with minimal effort.
  • The processor has a true DAC built in, which could be wired to the speaker for improved sound capability
  • With maybe a little extra circuitry it’s possible to run USB in Host mode, meaning one unit in host mode could connect via a cable directly to another unit in device mode for two player capability.
  • Battery life would be similar to the current Arduboy
  • Hopefully, the case would only require minor modifications to expose the MicroSD slot on one edge
  • Since the processor runs at 3.3V, it would interface to the display in a safer manner.
  • The time to develop and release this new product would be relatively short compared to other more elaborate proposals

To get an idea of what might be required to use the ATSAMD21G18 and be closely Arduino Zero compatible, the Rabid Prototypes Neutrino is a good example.


In one way, I love the SD card idea (specially as “catridge like gameboy”). On the other side, I also really like the coding on minimal conditions. at first i wanted arduboy because it seems to make fun programming for it. Later I decided it is also a very good way to increase my low programming + efficient skills. I’m currently working on a ASM2C# converter and it already have shown, how much i still have to learn.

I have much planned and cant wait to get it into hands, maybe i will try to start just with the software. So I will try to make as much as possible for the current model and later maybe something more with the next gen.

well, I tested the software, it really seems useful, if you have a plattform whe you are able to use it. Lua is also something, I want get more into, but would we have to pay for the software to be able to work then on the next arduboy?

This is definitely a real problem, what do you think of charging it through phone? is that even possible? would it make sense, i mean a even smaller device getting charged from the phone

I am sure, this time much more people will be interested to dev for the next gen :stuck_out_tongue:

I agree with @MLXXXp.

If screen update, coming hardships when make game.
The period of making games has a better short.

pico8 is 128x128 16 color. I feel good barance.
256 or 65536 color is not good.

GB and GBA homebrew programming is exist.
but It has passed the golden age. i dont hope to make GB or GBA clone.

new gimmick… MicroSD slot is good.
RPG or novel game need it.

speaker is arduino tone.
a long period of time. not suitable to hear.
if sd slot added, probably i use wav sound.

memory. i want x2 screen buffer. ATSAMD21G18 is nice!


One thing to keep in mind is that a color OLED display would require a SPI speed of:
128 x 128 x 60 fps x 16-bit = 15.73 MHz

It seems that the SPI clocks are limited to 4 MHz for the nRF51822 and 12 MHz for the ATSAMD21G18.


It’s possible to write the color OLED with 8 bit color as well.

I’m looking pretty heavily at the new atsamd, atmel themselves have asked me to try this out… It has a DAC, so yes wav files from sd card are pretty easy. I’d be looking at possibly including an actual speaker with amp.

The question is, how tied to the credit card format are we? I’m kind of thinking passport sized for the next one would make designing it much easier.


passport size/colour screen would no longer mean “arduboy” to me
more pixels, processing power and memory would still fit with the ethos and allow games with larger scope but some limitations to stimulate creativity
you mentioned earlier about custom oled screens, would this still be possible?

1 Like

You also may have to look at the speed capability of the chosen display as well, which could turn out to be the limiting factor. For instance, the SSD1306 based display currently used is rated for a maximum clock speed of 10 MHz.

I don’t know much about the nRF51822, but the SAM D21 Datasheet, in section 27.2 for SPI master mode, states:

Serial clock speed up to half the system clock

Which for a 48 MHz system clock would be 24 MHz. But maybe I’m missing something. @davidperrenoud, where did you get the 12 MHz value?

Something else to consider with the ATSAMD21G18 (I’m not so sure about the nRF51822) is that the SPI interface is DMA capable. If the display contents were double buffered in processor RAM, the display could be being updated in the background via DMA while the processor was rendering the next frame, thus providing some additional processing horsepower.

1 Like

Speaking as only one customer, I wouldn’t want something bigger. A little thicker would be OK. non-metal back would be OK, color screen would be OK (16 color prob fine), and SD is almost a must. But beyond that, I wouldn’t add too much to it, or you’ll risk losing the Magic! (IMO).


My point is game developing times.
Arduboy is very simple. I and any developer can make 1-3 weeks.

No offense, 8bit(256) color is opposition.
4bit(16) color or 4 mono(GB) is limit.
2bit(2) color is good.

128x64 size. not bad. i feel near normal.
But super mario(action game) difficult.
160×144(GB size) is limit.

Passport(physics) sized… I worry it.
Bacause i bought Nintendo 3DS LL than normal version lol.

my dev kit arduboy back face and size is beautiful. very amazing.
i want keeping. but my word is not strong.

A 128 by 128 pixel display with 8 bit color would still require 16K of RAM if you wanted to buffer it in system memory. For double buffering you would need 32K, which wouldn’t be doable with a processor that only had 32K total.


External ram may be part of the equation too, but it really depends on the ability to page the buffer to the display. i.e. putting it on the same serial bus and switching the chip selects so when it reads out, it sends it to the screen directly from the ram.

Color, honestly, if we want to keep it the size of a credit card, won’t happen. It doesn’t fit and the power consumption is too high for our small battery.

Id love to get gray scale with the custom oled but we probably need to order a million units to get that… Still plugging away trying to get in with the manufacturer.

So it seems the arm-0 with SD card and potentially some ram could be a nice option, same screen.


If you’re going to stick with the same screen then the RAM in the processor will likely be sufficient.

1 Like

grey scale on a custom oled with a DMG resolution in the same form factor will sell a lot so keep plugging away on that deal!

I’m a bit against the choice of Pico-8. Not that a nice retro style console with Lua support wouldn’t be nice, but as far as I’ve understood Pico-8 is not Open Source Software and that is what Arduboy is about. Open software and hardware.

Anyway, I’d like to see some low power ARM, (at least) grayscale screen and possibility of loading multiple games on the next generation Arduboy. Oh, and also some sort of communications between the consoles would be great. I already have an idea of a game that could really use multiplayer (the concept is not new one, though).

I think that the use of (Micro)SD cards as cartridges is a bit bad idea as those cards are way oversized for most (all?) games that such console might have and especially MicroSD cards are easy to lose (they are very tiny). So it’s more like allowing the console to hold more games and maybe also allow the user to easily copy games to the console by just dropping them to the card than to swap the cards all the time (like cartridges).

1 Like