• State
  • Locked Locked
  • Replies 6 replies
  • Subscribers 83 subscribers
  • Views 1930 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.

TMS320F28335: No clock present - "limp mode" clock on startup.

Sergey Feingold
Prodigy 220 points
Part Number: TMS320F28335

I'm having an issue on a board I am testing. 

I have a scope connected to the XCLKOUT pin of the F28335. When I power the board on, the signal shown is a square wave of ~3.7MHz, which changes to a square wave of ~1.9 MHz about every second. I believe the change is due to the external reset supervisor which expects a watchdog reset every so often. When I hold the device in reset only the ~1.9MHz clock is present.

I'm not sure why this is. Do I need to change any configuration settings for the oscillator to start? I've attached some pictures of the clock schematic and board layout. I believe this is all set up correctly but...no clock!

The capacitors are directly beneath the crystal on the bottom side. 

This is the crystal being used: 

Any advice?

  • 0
    TI__Genius 17370 points
    Hello,
    I am writing to let you know that a C2000 team member has been assigned to this post and should be answering shortly.

    Regards
    Baskaran
  • 0 in reply to Baskaran Chidambaram
    TI__Expert 8105 points
    Sergey,

    You stated that when you hold the device in reset you see the a 1.9MHz signal at the XCLKOUT pin. This gives me a clue that some other circuit on the board is driving this signal because during reset all MCU pins are forced to input mode. Could you check what else might be driving the XCLKOUT signal on your board?

    Regards,
    Peter
  • 0 in reply to Peter Galicki
    Prodigy 220 points

    I don't know what else it could be...XCLKOUT is only connected to a test point and nothing else:

  • 0 in reply to Sergey Feingold
    Prodigy 220 points

    I am also curious about this from the hardware design guide, as it seems to contradict what was mentioned earlier:

    So then it seems that there SHOULD be a signal present on XCLKOUT even during reset, but if it reflect SYSCLKOUT/4 that would imply SYSCLKOUT is ~7.2MHz which doesn't seem to make sense. Is SYSCLKOUT a direct reflection of the crystal oscillator frequency?

  • 0 in reply to Sergey Feingold
    Prodigy 220 points
    One other thing worth noting, I proved the crystal oscillator directly and it does indeed show a 30MHz sine wave. I did this on 3 different boards (with 2 different crystal capacitor values) with the same result, so I don't think it's the probe capacitance jump starting the clock or anything. Since the crystal is running and connected to the MCU - what could be happening?
  • 0 in reply to Sergey Feingold
    TI__Expert 8105 points
    Sergey,

    You are correct about XCLKOUT not being affected by XRSn on this device. XCLKOUT becomes input during XRSn on our newer devices that have it multiplexed with GPIOs.

    I think for the next step in this debug could you try running a known-good software example that you can download with the Control Suite. This example is called timed_led_blink. Please load it using the emulator and run it while observing the XCLKOUT pin.

    If that works correctly, the next step would be to take a look at your software that is being executed at startup (for example is it running out of Flash, and is it actually programmed into Flash when trying to execute)

    Regards,
    Peter