ARM Dev Board With USB Uploading

[George and Bogdan] wrote in to tell us about a cool Kickstarter they’ve been working on. It’s called the MatchboxARM, and like other tiny-yet-powerful ARM dev boards floating around, this one features a very fast and capable processor and more than enough pins for just about any project. One interesting feature of this board, however, makes it stand out from the pack: it has a USB mass storage-based bootloader, meaning uploading new code is as easy as a drag and drop.

This isn’t the first dev board we’ve seen to sport this feature: the Stellaris Launchpad has had this for a while and even the lowly ATtiny85, in the form of a Digispark has a mass storage-based bootloader. The MatchboxARM, though, brings this together with a very powerful ARM microcontroller with enough I/Os, ADCs, PWM pins, and I2C and SPI ports for the most complicated projects.

42 thoughts on “ARM Dev Board With USB Uploading

  1. I was wondering why the exact type of Cortex M3 is never mentioned on the Kickstarter page. The only STM32 component that fits the description is the STM32F103C8T7.

    While this seems like a nice board, the cost of the components should be about 6€ in quantities of 100 (e.g. at Farnell). Selling these things for about UKP 15 (about 17€) seems a bit outrageous, even when considering 20% VAT in the UK, especially compared to the price of the STM32F4 Discovery boards – these are physically bigger, but can also be found for about 17€ and have a significantly more powerful CPU.

      1. >>>
        PCBs, assembly and design time are all completely free and should not be added to the BOM cost?

        Of course not!
        Your supposed to go homeless and starve to death,
        not earn a proper living from the work you do.
        How dare anyone even think to make a profit and succeed?
        (It’s to bad there isn’t a special open source, sarcasm font design.)

        Yeah, I seen that attitude to the point of a near death experience.
        And that was from people claiming a religious belief in Jesus Christ.

  2. I’m always concerned when I see USB (although mainly Mass Storage) based bootloaders/devices. Security would be one of my worries, at least with other programming protocols (SPI, JTAG) you usually need programmers, cables etc. With this, before you know it, someone has changed your firmware in a few seconds with out anyone realising.

    1. It’s a dev board, you wouldn’t put it into a product that you’re mass producing. When you’re talking high-volume, paying for a USB connection your end users won’t use wouldn’t make sense.

    1. Simply clicking on the kickstarter tag pretty much reveals if this was policy prior to the sale, it was policy not followed. Hopefully such blanket policies aren’t nor ever will be part of the Hackaday SOP.

      1. I hope they /do/ enforce that policy. “articles” that fails to even mention the kind of processor on the board, leading to a kickstarter that give you an expensive board (hold on, 15 quid for a stm32 breakout??) that is not open hardware/source have nothing to do on HaD. It’s another “arduidiot bandwagon” kickstarter.

  3. The mbed development board has both a fast ARM chip and USB mass storage based programming… And a bunch more features AND HackADay has reported on it before!

    Still, its a neat bit of kit!

  4. The NXP’s Cortex-M3 LPC1343 microcontroller has only 33-pin, 32 KB Flash and 8 KB RAM.
    The MatchboxARM microcontroller STM32F103C8T6 has 72 MHz, 48 pins, 64 KB Flash, 20 KB RAM.
    The Mbed is a great platform (I have one :)), only that the LPC1768 comes in 100 pin’s package. Not so easy to solder one.

    Developing the MatchboxARM using the STM32F103C8T6, had started having in mind one clear goal: after the user will learn how to correctly use all the peripherals of the microcontroller used on MatchboxARM, be able with minimum economic investment to design and build his own project board at home, using the STM32F103C8T6 microcontroller or his big brother CBT6 which has 128 KB Flash.
    That way the projects be easyer to debug before sending them to the PCB manufacturers!

    We encourage the creative tinkering, and with our little development board, try to offer a fast path and affordable way bring life to your projects.

    Thank you all!
    Bogdan from MatchboxARM team.

    1. If this is a comparison to the USBUG then you have some things incorrect.
      The LPC1343 comes in 33pin QFN or 48pin QFP. The USBUG uses a 48pin footprint. Though the title says LPC1343 board on the DP article, it is pin compatible with anything in that line, so you can have a USBUG with a LPC1347 72Mhz 64K ROM 12K RAM. Or you can go Cortex-M0 with pin compatible LPC11XX variants.
      To do some USB testing I’ve actually taken the PCB files for the USBUG and adapted it to the 33pin QFN making it smaller and adapted the USB connector to a Male connector so I can just plug it into a computer thumb drive style.

  5. I don’t understand why this needs a kick starter. Seems like something one of us would develop because we were bored. It’s not like there are any real manufacturing/sourcing/marketing cost that need to be covered- at least to the extent that some niche product like this would require. Heck, a diy-solder stencil and a toaster oven would be enough to manufacture 90% of this thing- you could even put 100 of them on a tray at a time.

    1. You are welcome to suggest us future improvements.
      We plan, if we reach our founding goal, to make a second version with W5200 chip on the board, so the IoT using Xively be easy peasy.
      You can contact us on facebook. Search MatchboxARM. (a great development board)
      You can tell us also about some of your great projects and maybe we can find a common ground.

    2. Well the hardware is nothing special. I designed one just last week as I had some extra room in board going to iteadstudio. This project only makes sense if they are going to put together some easy to use development enviroment. Something like Arduino for STM32.

  6. Developing my drivers and libraries onto a new platform is tedious – like starting from scratch I’d rather use something that has tons of hardware shields, oodles of code, and is quick and dirty to get to prototype.
    These new boards seem like the new new thing, but they’re a blank slate and interesting for a little while, but not quite useful.

  7. Let’s be serious! It is a great project. It’s so tiny and nice. This board it’s a must in my workspace. The STM32 is trendy these days. Check the afrodevices or the Crius IMU.

  8. In reply to asldjflkjf.
    Criticize this project say much about you.
    We are open mind people. I like diversity. Every one must have the oportunity to show the world his work. And this my friend it is something that deserve respect! That’s why today we have linux, windows, mac and so many other.

      1. It’s not an arduino compatible with anything. It’s not because you plonk a MCU on a board with pin headers that it becomes an “arduino”, It’s a frigging breakout board with a stm32 on it, and, the other major difference here, is that I didn’t make a kickstarter trying to profit from it.

  9. It is nothing special. There are tons of AN on USB bootloaders. I have extended one for a client to be able to change a data “file” in an external memory and it wasn’t a big deal. Apart from that, when developing, it is better to have JTAG since debugging is frequently needed (except if someone has the clearest and most focused mind in the world).

  10. I’m surprised the article doesn’t mention arduino at all.

    Also, 15-20 GBP with shipping isn’t that much for a dev board, that’s pretty much the production cost for low volume boards both if you use hand assembly or low volume pick and place.

    The current version might sell better since it looks like an arduino nano but for a breadboard dev board i like their SIL prototype better than the DIL version shown here.

  11. Every one seems to have overlooked the Freescale Freedom series of boards: Currently two different M0+ versions (at under £10 each) and an M4 version (at just over £10). They have Arduino compatible headers, tricolour LED, capacitive touch slider, accelerometer, etc. etc.

    It has the option of having USB Drag/Drop to program, or using a full “openSDA” programmer/debugger (which CAN be used to debug/program off-board with alternative firmware) with a firmware change which, funnily enough, is done by dragging and dropping firmware onto the debugger in bootloader mode.

    The “official” dev environment (codewarrior) is a version of eclipse with proprietary extensions, but this gives you component level drag and drop. There are also easy methods of developing with other IDEs, including online and mBed.

  12. GBP 15 is about $23 – That is how much a the OM13000,598 costs (LPCXpresso Board with a LPC1769 120MHz ARM Cortex-M3 MCU with 512 kB flash and 64 kB SRAM).
    This seems to give much more bang for the bucks since you apart from the fact that you’re getting a more powerful mcu you also get a detachable USB/JTAG programmer/debugger as well.

  13. The MatcboxARM it isn’t only a breakout board. It comes with the USB bootloader which help users programing the micorcontroller flash in a easy manner.
    We also provide examples for all the peripherals of the micro.
    And this if one key feature of MatchboxARM.

    There are tons of USB bootloaders, but we can find one for STM32F103CxT6 until now.
    The JTAG pins are available on MatchboxARM.

    We are a team of two young people that need a place to start and for people that started from zero, they know how difficult is.
    The backers are not ripped off!
    The backers trust in that our project may have a future and are trying to help us.

    The KL25Z it is a nice board but hey, behind that board it’s a huge company, the FREESCALE SEMICONDUCTOR. Than’t where the $12.95 came from.

    Thank you all!
    All the comments, good or bad it’s an important feedback for us, because knowing this, we know better what to improve in the future!
    Bogdan from MatchboxARM team.

  14. Here’s my feedback that nobody asked for:
    In order for a dev board to be successful it has to have either one or more of the following qualities:
    – dev backing with a lot of examples and tutorials
    – very low price
    – publicity (media budget, interesting stuff done with it, celebrities using it, …)

    Arduino has 1 and 3, Launchpad and other Energia boards have 2 and maybe 1 as well.
    It’s cool to have a powerful uC but for >90% of the projects it’s underutilized anyway. Stellarpad has an even more powerful CPU at less money.

    I’m pretty confident the project’s goal will be met but nothing to brag about. It’s better to focus on bringing a cheaper dev board into Europe where pretty much anything has to be imported. Devs will follow.

    I have amassed quite a few dev boards and bare chips but lately (1.5yrs) I have only been using Arduino clones or Energia-compatible boards. You just connect the nRF, DDS or LCD module, copy&paste someone’s code and it just works, no fiddling around.

  15. I think the 40-pledge should offering two boards, instead of one board with “direct contact” to the development team. You need to have contact with the developers in any case, it’s a new product. Otherwise very tempted to get one. Good luck, Matchbox ARM!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.