1
\$\begingroup\$

I have a project where i need to run a python script on a microcontroller. The python script is basically a GUI based script using pyQt and the internal ram of a microcontroller isn't enough. After surfing the internet for hours I came to a conclusion that I will need a powerful microcontroller and an external SDRAM to support Linux and then run the script.

I don't know if that's the best way to run a python script, but if it is then how do i interface the Allwinner A13(or any cortex A microcontroller) with a huge SDRAM(about 200MBytes).

If you could guide me the right direction or suggest anything, it will be deeply appreciated.

\$\endgroup\$
5
  • 6
    \$\begingroup\$ ::boggle:: pyQt on a microcontroller? I think you would do better to eyeball a RaspberryPi. It'll have everything you need, both software and hardware. \$\endgroup\$
    – JRE
    Commented Mar 8, 2016 at 16:28
  • \$\begingroup\$ Are you trying to design a custom PCB? If you don't have much experience with PCB layout or ciruitry design a better option would be as @JRE suggested. \$\endgroup\$
    – R. Johnson
    Commented Mar 8, 2016 at 16:31
  • 1
    \$\begingroup\$ Your project plan amounts to: Build a system around an ARM, and make sure it has enough RAM for your task and for a complete Linux system. Create a graphics subsystem and the needed graphics drivers. Port the linux kernel and all needed userland components (pyQt and Python included) to your new hardware platform. Does that not sound a bit ambitious? \$\endgroup\$
    – JRE
    Commented Mar 8, 2016 at 16:32
  • \$\begingroup\$ I have some experience with PCB and circuit design, I have worked on Atmega8 and other 8 bit microcontrollers and also on STM32 microcontrollers. \$\endgroup\$ Commented Mar 8, 2016 at 16:33
  • 4
    \$\begingroup\$ With that experience, if you start looking into the details of a remotely modern embedded Linux boards, you'll realize it is not something you want to design yourself unless you are going into quantity production or your requirement is fundamentally too unique to fulfill by creating an adapter/daughter/host board for an existing module. \$\endgroup\$ Commented Mar 8, 2016 at 16:41

1 Answer 1

3
\$\begingroup\$

You could get a Raspberry Pi. That would simplify a lot of things that you want to achieve.

\$\endgroup\$
7
  • \$\begingroup\$ I want to learn building it myself thats why I dont want to go for raspberry pi. \$\endgroup\$ Commented Mar 8, 2016 at 16:35
  • \$\begingroup\$ You could study the schematics of an existing Linux SBC, list out the components required and the function they perform, select components that perform similar functions and use them to solder your own board. Then search for a Linux distribution that supports the controller that you have selected and write the drivers for your hardware. \$\endgroup\$ Commented Mar 8, 2016 at 16:37
  • 2
    \$\begingroup\$ Buying an off the shelf embedded Linux board is definitely the right choice for anything needing one where an existing offering is at all plausible and the needed quantity is below the thousands. However, the raspberry pi line is often a terrible choice for anything beyond a learning/hobby prototype - the SD card dependence makes the system very fragile in hardware and state, and the lack of power management features can often be an issue. \$\endgroup\$ Commented Mar 8, 2016 at 16:39
  • \$\begingroup\$ Can You suggest an open source SBC? except raspberry pi. \$\endgroup\$ Commented Mar 8, 2016 at 16:39
  • 1
    \$\begingroup\$ The BeagleBone Black is open source, and it is possible to get all of the parts for it. However, if you buy the parts you will probably pay twice the cost of buying a Beaglebone Black that is guaranteed to work. \$\endgroup\$
    – Steve G
    Commented Mar 8, 2016 at 16:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.