Hello again, from Jigsaw Renaissance. I promised you a look at our initial bill of materials, so here it is.
Our device has a few core functionalities which we need to work on. These include:
Audio input and processing
Audio output
Interaction through motion (e.g. shaking the device, dancing)
Interaction through explicit input (e.g. buttons)
Communication between devices
To simplify development and hopefully save enough time to get the core features working by the end of April, we have decided to use the Olimex STM32-103STKSTM32-103STK development board. We will modify these with STM32F103RGT6STM32F103RGT6 microcontrollers. This implementation of the Cortex M3 has 96K of RAM and 1MB of flash in a dual-bank configuration, which will be very helpful for some of our planned features. A PCB specific to our device will be designed and fabricated later in the project. As the software is by far the largest and most challenging component of this project, we want to be able to start work on that as soon as possible.
While there is already an accelerometer on the Olimex dev board, we will be working with the Freescale MMA8452QTMMA8452QT as this part is likely more suitable to our final design in terms of cost and features. We will be fabricating a simple external board for this part as well as for some other items such as LEDs and switches. This will help during initial prototyping (i.e. until we design the main PCB).
Our current estimated budget for electronics is as follows. This will leave us with a fairly decent margin for unanticipated issues, casework, additional prototypes, and so forth.
$218 for 2x Olimex boards
$44 for 5x micros
$18 for 3x radio chips
$36 for 3x accelerometers
$100 (?) for 3x blank boards
$50 or so for remaining parts (switches, LEDs, speakers...)
There are a few key issues which we’re going to face, but the main challenge here is time. A month is really not much to work with for an electronics project of any complexity—and our planned device is not going to be very simple. There are several aspects which need to be working simultaneously in order for the user experience to come together. This involves several fairly complex software tasks in a very resource-constrained environment, even with the STM32F103RGT6’s comparatively ample feature set. Furthermore, we really need to get an initial version into the hands of users and iterate based on how they receive it and interact with it.
Given the constraints of the contest, our planned solution here is to initially stick to a few core features and get them working in as simple of a version as can possibly do the job. This is, of course, what you’d want to do for a reasonably large project anyway. It does leave us a little uncertain as to what the version we demonstrate for the contest will include. But regardless, the contest has provided us with the means to bootstrap the project.
We fully expect that further development will be needed after the contest. We can get some version of the device working in a month, but there’s no way we can adequately iterate it into a form which adequately serves students’ needs. Our hope for the GGHC is to create a version which will get educators interested in working with us to build the version we need to build in order to meet the device's educational goals.
Stay tuned.