Loading ...
Sorry, an error occurred while loading the content.

287Re: [TMS470_ARM] Analog Watchdog Questions

Expand Messages
  • JohnStosh
    Apr 26, 2010
      One issue that might be causing your symptoms:If you have a lot of RAM memory to initialize (or zero), then the support code may take a long time to execute.  This means the analog half of the watchdog may cause a reset before the first line of your code gets a chance to execute.  To handle this issue, there's a callback routine.  This callback routine is called from the IAR support code before it sets up memory.  It is named __pre_****_init or something like that.  The guys on the IAR support phone line are very helpful.
      I'm sure there are other issues that can cause your symptoms.
      --- On Sat, 4/24/10, Scott Coleman <dvarapala@...> wrote:

      From: Scott Coleman <dvarapala@...>
      Subject: [TMS470_ARM] Analog Watchdog Questions
      To: TMS470_ARM@yahoogroups.com
      Date: Saturday, April 24, 2010, 7:25 PM


      I have an Olimex TMS470_P256 development board and I'm trying to use the Analog Watchdog feature of the TMS470. I'm using IAR EWARM 5.41, and my program is configured to boot from flash. My main() has nothing in it but an infinite for loop which kicks the watchdog and toggles the LED so I can tell it's running; if I keep the watchdog disable jumper in place until main() is called and the LED is flashing, I can remove the AWD disable jumper and the code continues to run fine. However, if I reset the board with the AWD disable jumper removed, the code never makes it to main() and the LED never flashes. If I then install the AWD jumper, the board boots fully and the loop in main() runs correctly.

      Clearly the watchdog is timing out and resetting the micro somewhere in the startup code before main() is called. To those of you who are using the analog watchdog in your designs: did you run into this issue, and how did you handle it?
    • Show all 3 messages in this topic