• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 82 subscribers
  • Views 773 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.

External clock error

Alvaro Martinez
Prodigy 110 points

Dear Sirs,

I am programming a F28335 using a self-design board and using the RAM. I have just connected and I found some problems related to the system clock.

In the design, I am connecting a 100MHz clock to the pin xclkin. X1 is grounded and X2 is disconnected (as described in the manuals). The CLK measured in the input pin fulfills the electrical requirements (see the picture attached).

Since the debug is stopped in the the main() (at the very beginning), the register SysCtrlRegs.PLLSTS.bit.MCLKSTS is equal to 1 and there is no way to change it (using MCLKCLR). I have checked also that the DSP is working in limp mode (the CLKOUT is arround 500kHz). I notice this error because the program was stuck in the SYSCTRL__InitPll function. 

I have no idea how can it be in limp mode if the XCLKIN is correct (at least in the pin). Am I missing something regarding configuration? 

As this error happens before I can configure the system CLK, I have no idea if there is a mistake somewhere, or if the DSP is damaged. 

Any help will be welcome. 

Thank you in advance!

Alvaro

  • 0
    Guru 20525 points

    Hello Alvaro!

    Check please your settings for:  DSP28_PLLCR & DSP28_DIVSEL. Perhaps there is issue like this http://e2e.ti.com/support/microcontrollers/c2000/f/171/p/279762/976406.aspx#976406

    Regards,

    Igor

  • 0 in reply to Igor Gorbachev
    Prodigy 110 points

    Hello Igor!

    I'm afraid the problem is not related to these settings. Indeed, the Init_PLL funtion checks the SysCtrlRegs.PLLSTS.bit.MCLKSTS register before configuring the clock. 

    void SYSCTRL__InitPll(Uint16 lVal, Uint16 lDivSel)

{
//Check PLL
if (SysCtrlRegs.PLLSTS.bit.MCLKSTS != 0){
// There is an error
// Replace this line with a call to an appropriate
// SystemShutdown(); function.
asm(" ESTOP0");
}
//continue configuring the PLL/CLOCK...
    }

    In any case, I use DSP28_PLLCR = 2 and DSP28_DIVSEL=2 so that I get at the end 100MHz. 

    Regards,

    Alvaro

  • 0 in reply to Alvaro Martinez
    Guru 20525 points

    Hi!

    Yes. Settings are right. If I understand you did use scheme 3-10. Maybe it makes sense to try 3-11? And did you try to return to using internal crystal? Does DSP work at this case correctly?

    Regards,

    Igor

  • 0 in reply to Igor Gorbachev
    Prodigy 110 points

    Hello Igor,

    Thank you for your interest! Yes, I am using the scheme 3.10. I'm afraid it is not "easy" (I don't even know if it is possible) to modify the board and convert it into scheme 3.11.

    I would like to add that this very same design works perfectly in other boards, so this error is particular of this specific board. I don't know if the PLL or the VCO could have been damaged somehow and if there is any test I could do to verify this.  

    Regards,

    Alvaro

  • 0 in reply to Alvaro Martinez
    Guru 20525 points

    Hello Alvaro!

    Alvaro Martinez said:

    I would like to add that this very same design works perfectly in other boards, so this error is particular of this specific board.

    If these boards are identical and its have the same firmware then there are two variants 1) A device (DSP) of second board  is damaged  OR 2) The second board has some defect (for example bad soldering etc.).

    Alvaro Martinez said:

    I'm afraid it is not "easy" (I don't even know if it is possible) to modify the board and convert it into scheme 3.11.

    Well, you certainly know more clearly. But it seems there is no problem (only other pin and other voltage level of 100 MHz frequency ). But maybe there is some principal constructive problem... 

    Alvaro Martinez said:

    I don't know if the PLL or the VCO could have been damaged somehow and if there is any test I could do to verify this.  

    I do not think that some testing can help if the device is damaged. But in my opinion it make sense to try the work from internal crystal at least.

    Regards,

    Igor

  • 0 in reply to Igor Gorbachev
    TI__Guru* 95315 points

    Alvaro,

    If the same design works perfectly in other boards, it could be a damaged device. By "other boards" , I presume you are referring to other identical boards. Not boards of a different design that uses the same processor.  If feasible, please replace device on the board to see if problem goes away. You could also try a lower valued oscillator.

    Just out of curiosity, why are you not using a lower-valued clock input and take advantage of the PLL to step up the clock frequency to 100 MHz? 

    Hareesh

  • 0 in reply to Hareesh Janakiraman
    Prodigy 110 points

    Hello Hareesh,

    Yes, when I say "in other boars" I mean identical boards (same design). Another specialists is checking the board and it seems the device is partially burnt. I will try to replace the device although it is a bit risky :D. 

    I have no idea why they chose this clock. I didn't make the design myself!

    I think we can close this topic. Thank you all!