Freescale iMX31 project

Support projects for the iMX31 work included a Telit phone module, a Philips MiFare NFC module, instrumenting parts of the TrollTech Qtopia phone stack and cross-compiling the complete Linux/ Qtopia phone stack for an ARM target. The work was done from May to September, 2008. A separate Atmel SAM7 ARM project was done for the same company in October, 2008.

On arriving in Adelaide, Australia, in 2008, we were fortunate to contract on the iMX31 ARM for a company developing a smart phone. The development platform was a Freescale evaluation board with a Hitachi high-resolution screen. The target ran Linux and the Qt phone application. The development was cross compiled on a x86-64 PC. The kernel, cross compilers and Qt toolchains were kindly installed by one of their developers.

Our input was as a contractor joining the project just before government funding was cancelled (with the change of Australian government in 2008). How different the local electronics industry would have been if the ideas that were being discussed then were fully funded. Fortunately, we were kept on while venture capital options were explored. We picked up numerous smaller support projects, and although we did compile the kernel and Qt stack, we did not work on the iMX31 as there was only one board. Freescale, like other silicon vendors had pre-announced a device with Ethernet, in addition to all the other good things on the iMX31. The iMX31 was a rather expensive board to merely purchase for a contract. As a spin-off, we would be able to leverage the development for an industrial terminal on hardware that was being locally produced, which was at the schematic capture point ready for layout.

The auxilliary interfaces and projects, like a Telit phone module and Philips MiFare NFC were fairly short but interesting. Testing was initially against our x86-64 Linux box, then ported by the main developer onto the iMX31. There were discussions in buying an Abatron JTAG probe, as it claimed to be able to handle the Linux kernel debugging after the MMU and TLB were enabled.

In 2011, we searched for the smart phone but assume the development was purchased by a company who could put in more resources, as unfortunately, in the smart phone and tablet space, competitors are Apple, Samsung, Blackberry and all the Androids. We will not be documenting the NFC or Telit work that was part of the contracting period, but if you have testing work for debugging a stack, perhaps we can assist.

We maintain an interest in the Freescale, Xilinx and Texas Instruments announcements on the new ARM devices. Although we would like to develop an industrial terminal that these devices make possible, we would need partners able to finance the development. For now we think that the Apple iPad could provide the user interface with a less capable ARM core for “bare metal” real-time work. This is a lower risk option as the time scales for tablet introduction will be a year maximum, and not much software gets done by small teams in a year when simultaneously trying to debug hardware.