• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 98 subscribers
  • Views 259 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.

Accuracy of the CLK module. OmapL138, DSP/BIOS

Sergey Pot
Prodigy 20 points
Other Parts Discussed in Thread: OMAPL138

Hello!

I'm working  with DSP core of OMAPL138 chip. I am using DSP/BIOS 5.41 and CCS3.3.

Omap is used to read data from external receiver via McBSP port and process it. I/m using SIO to work with McBSP. Sample rate of of receive is 48kSps. Also I have internal timer with tick rate is 72kHz. Timer is configured via CLK module of BIOS. Reference clock is 14.4MHz and PRD value of timer is 200.

I measured number of times I've got an interrupt from receiver per 1 second of the internal timer. The expected value is 48000 interrupts. But received value is nearly 51000.

I don't understand what's going on...

Please help!

  • 0
    Prodigy 120 points

    Hello,

    The timers work fine and, if you use an external crystal oscillator they are accurate. So regarding your problem, check your PLL settings and timers settings. Maybe, the bios module "think that omap" works at 300MHz and in fact it works at 456MHz (that 's happened to me with another module... )

  • 0 in reply to ludovic garcin
    Prodigy 20 points

    Hello! Thanks for reply!

    You are right, the timers work fine. I have done some experiments. I purposed that the reason of problem is in dropping interrupts from timer.

    I checked register EVTFLAG3 (0x0180000C) for event 96 - drop of interrupts  for different situations.

    In first case all my functions were enabled (read data from receiver,timer, board configuration).In the second case I disabled all my functions that uses DSP/BIOS API. Only timer and zero-data (because receiver wasn't configured) processing were working. In first case there were dropped interrupts and in second case - not.

    The conclusion I can made that API functions "break" interrupts.