• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 21 subscribers
  • Views 354 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.

Original question:

LMX2572: Does initialization require calibrating with R0 twice?

LMX2572: Fails to lock when system reset

Kazuki Itoh
Genius 9165 points
Part Number: LMX2572
Other Parts Discussed in Thread: LMX2582, LMX2594,

Hello team,

My customer found that the LMX2582 fails to lock when the system is reset by R0 register. This issue repetitively happens approximately one in ten systems.

This issue looks similar to what discussed on this thread. Could you please help me to address the issue?

Please find the internal link here for schematic and register configurations. (Please connect to VPN for download)

Details:

1. When powering on, the same register setting with system reset is used. After reset by R0, R125-R0 are set following the DS.

2. When power it on, the system succeeds to lock

3. When resetting the system, it fails to lock in 1/10 rate (in nominal temp.)

4. In lower temperature, lock failure rate increases

5. After failing the lock, it is very likely to lock when configuring R125-0 without R0 reset.

Questions:

1. In the previous thread, the user mentioned that this case depends on the IC temperature. Could you please elaborate on that?

2. In LMX2594 datasheet, it is mentioned that waiting sequence should be inserted before setting R0. 

Should this sequence be applicable to LMX2572 as well? (Meaning after setting R112-R0, and then have 10ms waiting and setting R0 again.)

Regards,

Itoh

  • 0
    TI__Mastermind 28960 points

    Hi Itoh-san,

    1. Writing R0 triggers the calibration routine. Sometimes, the state of the PLL is still settling while writing R0 the first time, which can invalidate the results of the calibration. Depending on the temperature, factors like the VCO gain and the LDO settling time will shift slightly. If the parameters are marginal compared to their nominal values when R0 is written the first time, the LMX2572 (and other devices following a similar architecture, including LMX2594) may calibrate with invalid coefficients, requiring a new calibration to be performed.
    2. The recommended initial power-up sequence for programming LMX2594 is also applicable for LMX2572, including the additional wait for 10ms followed by one additional write to R0 with FCAL_EN=1.

    Regards,

  • 0 in reply to Derek Payne
    TI__Genius 9165 points

    Hello Derek-san,

    Thank you so much. We are working closely with Noel to address this issue.

    One additional question from customer.

    In the system, two LMX2572 are used.

    Please find the internal link here for schematic. (Please connect to VPN for download).

    When the system reset, U2401 is likely to fail to lock, but U1200 locks well.

    The difference between UC2401 and U1200 is that U2401 OSCin is differential while U1200 is single-ended.

    Noel and we aligned that the lock failure should be caused by too large bias caps.

    So, my customer concerns that U1200 can fail to lock in some cases.

    When applying the system reset while power is on, do you recommend to apply reset, register setting R125->R0, wait 10ms, and R0 sequence to U1200 to make sure locking, as well as U2401?

    Regards,

    Itoh

     

  • 0 in reply to Kazuki Itoh
    TI__Guru* 81400 points

    Hi Itoh-san,

    After a RESET, OSCIN type is single-ended while the internal 50Ω termination is disabled. I am not sure whether this makes the difference. 

    Is it possible to measure the OSCIN signal while programming R6 to R0? (OSCIN type and internal termination registers are located in R5.)