Hello everyone, Thank you for signing up and backing this project!

We will cover the various other boards associated with the Vision SoM in this update, i.e., the Developer kit, Breakout board, and Artemis adapter.

Developer Board

The developer kit provides the following functionality:

• FTDI FT2232 based FPGA programming over USB supported by the open source icestorm and Lattice toolchains
  • Supports a high speed SPI interface to the module for the host interface
  • Independent UART interface to help debug and interface to the FPGA
• Additional microphone for stereo configuration when paired with the SoM microphone
• 3.3V level translated I2C allows connection over the QWIIC connector to various sensors
• Dual PMOD
• INA219B based current measurement at up to 500Hz to characterize the SoM

The power monitoring allows optimization of the FPGA code for power. eg. the plot below was obtained using the INA connected to an Arduino sketch and shows the power consumed by the Himax HM01B0 sensor at 10 frames per second.

Breakout Board

The SoM uses a high density connector to interface with the host PCB. This can be rather inconvenient when prototyping where a breadboard makes more sense. The SoM plugs into this adapter which exposes all pins in a breadboard friendly format.

Artemis Adapter

The Artemis adapter allows the SoM to communicate with the ultra low power Artemis Nano board which can form a complete system.

Note: This board is not being offered during this campaign, but the design files will be released if the project is funded.

Heres what the complete system looks like when assembled:

SoM Prototyping History

Here are some of the prototypes that led to the SoM as it is today…

SoM Proto 1:

This was the first iteration of the SoM! As you can see, its significantly smaller and has very few IO on the periphery. Only a single type of imager (Pixart, monochrome, global shutter) was supported.

There were some design defects as well: the Flash couldn’t be programmed due to a bug. Fortunately, there were some testpoints which we rigged up!

The module had 2 PDM microphones which we found consumed a lot of the FPGA resources. Besides, placing 2 microphones so close to each other provides no significant benefit when their signals are combined (beamforming) as shown by this theoretical plot.

SoM Proto 2

Himax sensor support was added along with the IMU. Besides adding a LOT more pins to the periphery, we removed one microphone and also changed the microphone to be I2S. Programming glue logic was added to enable programming the Flash and FPGA RAM for added flexibility.

Talk to us!

The project is not yet buttoned up! A lot of work remains in terms of software, documentation, marketing, training framework, hardware verification and so on. We invite you to collaborate using the Crowd Supply Discord channel and/or the tinyVision Discord channel.