• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 85 subscribers
  • Views 215 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TMS320F28377D-EP: Errata - PLL: May Not Lock on the First Lock Attempt

Laurent Z.
Prodigy 215 points
Part Number: TMS320F28377D-EP

Hello,

According to the errata, there is an issue with the PLL not being properly lock.

The PLLSTS[LOCKS] bit is set, but the PLL does not produce a clock.

TI recommends a minimum of five lock sequences in succession when the PLL is configured the first time after a power up.

A lock sequence means disabling the PLL, starting the PLL locking, and waiting for the LOCKS bit to set. After the final sequence,
the clock source is switched to use the PLL output as normal.

 I have few questions regarding this for loop though:

  • Why five attempts? Has this number shown the best success rate?
  • Why not 10, 15, 20 attempts? What is the magic number?
  • What is the duration of the locking sequence? 

My understanding is that if the 5 attempts were not successful, the watchdog will eventually reset the device once the PLL is transitioned to be used as the SYSCLK because no clock would be present to execute instructions to reset the WD counter.

Laurent

  • 0
    Prodigy 215 points

    My guess is that the function ServiceDog(), should be uncommented only if the Watchdog was enabled, prior to setting the PLL, in the application software. Is that correct?

  • 0
    TI__Intellectual 1410 points

    Hi Laurent,

    In "set up the system PLL" section in the System clock setup part in TRM specifies- 

    Lock the PLL five times. This number can be increased depending on application requirements.

    A higher number of lock attempts helps to ensure a successful PLL start.

    I will get more information on the 5 attempts.

    Thanks

    Aswin

  • 0 in reply to Aswin P
    Prodigy 215 points

    Thank you Aswin,

    Do you have any update regarding my questions? I did see before in the TRM the statement you mentioned above, but my question is why 5? I do understand that the more attempts the better but I also don't want to waste too much time doing this.

    • Why five attempts? Has this number shown the best success rate?
    • Why not 10, 15, 20 attempts? What is the magic number?
    • What is the duration of the locking sequence?

    Thank you

    Laurent

  • 0 in reply to Laurent Z.
    TI__Intellectual 1410 points

    Hi Laurent,

    I will reach out to another expert to get help in answering the question.

    Thanks

    Aswin

  • 0 in reply to Aswin P
    TI__Genius 11511 points

    Hi Laurent, 

    Laurent Z. said:
    • Why five attempts? Has this number shown the best success rate?
    • Why not 10, 15, 20 attempts? What is the magic number?

    Based on our characterization, five attempts is sufficient and has had an exceptional success rate in the field.  

    Laurent Z. said:
    What is the duration of the locking sequence?

    Please see the PLL lock time specification in the commercial/automotive datasheet.  The spec is the same as the -EP datasheet but I noticed the footnote in the -EP datasheet only mentions two lock sequences so I have filed a literature bug to update the footnote to match.

    The below is the time for locking once so you'll need to account for how ever many attempts your decide to use.

    Regards, Joe

  • 0 in reply to Joe Prushing
    Prodigy 215 points

    Thank you very much for this information.

    However, just to make sure, is it really the time for one locking sequence? The footnote mention the errata workaround which recommend 5 times.

    My understanding per the footnote is that this times (50us + cycles) is the execution of the full InitSysPll() with 5 locking sequence.

    Or is it 5(pll locking sequence) x 50us + 2500*tc?

  • 0 in reply to Laurent Z.
    TI__Genius 11511 points

    Hi Laurent, let me look into this a bit more.  Others concur with my conclusion but I will do some more research and get back to you early next week.

  • +1 in reply to Joe Prushing
    TI__Genius 11511 points

    Hi Laurent, based on my investigation I'd say the 50us is one-time and the 2500*tc is in the loop so in the case of 5x, the time would be 50us + 5*2500*tc