• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 83 subscribers
  • Views 661 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Linux/LAUNCHXL-F28069M:

Petric Zeljko
Prodigy 90 points
Part Number: LAUNCHXL-F28069M
Other Parts Discussed in Thread: LM629, , C2000WARE

Tool/software: Linux

                             Hi there!

             As the stone age HW/SW R&D, some 15 yrs ago I designed and built my beta DDC around LM629, using LS7083 for encoder, souped up 8051 for comms and axis supervision.  Hellish to tune  the PID, but  terrific accuracy, dynamic range, repeatability and reliability.   Even built sort of separate encoder logger, to graph the horrors of my deeds                                                    Now, aging  60's, jumped  the ARM bandwagon charmed by its 32 bit counters, encoder inputs and great ARM toolchain with libopencm3 good enough to carry on with lean and mean bare metal coding.  Picked the wrong tree obviously, since the way it handles I/O etc, turns me back to C2000 for its  PID stuff, etc.

         Goal:    Two small DC servos with 1000 CPR encoders served separately but simultaneously  by  LAUNCHXL-F28069M.

1) Can I develop lean and mean firmware in Linux environment, using your PID code chunks, without HAL, shooting straight to registers ?  ( your competitors blackbox HAL bloatware  is not deterrent by its waste of memory, but, the  more  code underneath, more chance for crazy spins and terribly waste of time in debugging ).

2) Can single F28069M bandwidth beat two LM629 in parallel, regarding above mentioned criteria , with faster loops etc.? 

3) If not on Linux, can I can use your IDE on Win 7, but retaining bare metal approach where possible ( no HAL,  RTOS etc ) ?

Thanks,

Petric

  • 0
    TI__Expert 7650 points
    Hi Petric,
    Thanks for your interest in C2000 MCUs for servo control!
    1. Our Code Composer Studio IDE is supported on Windows, Linux and Mac. C2000Ware installers also support all three platforms.
    2. My suggestion would be for you to consider the F28004x as an alternative to the F2806x. You will find that it has more performance, more features and as good or better pricing.
    3. For current loop development, I would also recommend that you check out our Fast Current Loop technology. Check out this blog post and the links included therein:
    e2e.ti.com/.../fast-current-loop-performance-better-than-we-thought-and-measurable-in-your-own-lab
    4. As to your points about forgoing a HAL and "shooting straight to registers" you may want to investigate the bitfield abstractions that are available in C2000Ware and also compare them to the HALs/driverlib that are also supported there.
  • 0
    Prodigy 90 points
    I've got more than adequate answers and the valuable suggestion from Mr. Fortman and ordered the LAUNCHXL-F28049 board from TI, same day he replied. Once it arrives to Southeast Europe, and get the downloaded software installed, there are thousand pages of documentation to browse through, in attempt to get my specific application running ASP. Any further hypothetical questions make no sense in embedded world, and if stumbled on specific issue, I won't hesitate to ask again. Incredibly fast response from qualified person with straight answers, is something your competitor could not yield, despite wider community and knowledge base on general purpose cortex F4 and F7 MCU's. In the meantime, I'm getting familiar with dedicated motor control SW/HW blocks related to F28049 motion control, and breadboarding own 3F power stage with sensor amplifiers . In short, no idling time and big thanks for the head start.
    Kind regards,

    Petric