• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 21 subscribers
  • Views 359 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.

Original question:

LMX2594: possible mismatch in the datasheet?

LMX2594: valid range for the DACISET

Viktor Vojta
Intellectual 315 points
Part Number: LMX2594

I am utilizing LMX2594 in full assist mode, only fixed set or frequencies is used, thus look-up table was created first with calibration values.

  1. I want to ask regarding the valid range of DACISET especially for VCO7. I read-back the value 217 at room temperature and f=14.99GHz, but the valid range is 323...244. Is it OK? Or shall I respect the range in the datasheet and load the value 244? I calibrated it with ACAL_CMP_DELAY 10 and 32.
  2. In the full-assist I set bit FCAL=0, the Lock Detect works OK. But there is a E2E-post from 22 Oct 2018, where FCAL=1 is recommended in the full assist mode. May I expect strange behavior for FCAL=0 even in the full assist mode?

  • 0
    TI__Mastermind 33580 points

    Hi Viktor,

    1. The limits for ACoreMin and ACoreMax are set to allow any device to lock, regardless of process and temperature variations. If you are doing partial assist, and creating register values that will be used by many devices without confirmation of calibration, I recommend observing the ACoreMax limit. If you are doing full-assist with a limited number of devices, or a best-fit calibration for each device, the true value given by the calibration can be used instead. If you use full assist with settings outside of ACoreMin and ACoreMax, they may not be applicable for every device, especially for frequencies at the boundary of the VCO range. Partial assist will still be able to settle to the 217 value on devices where it is the best fit, it will just take slightly longer to calibrate.
    2. Are you referring to this post? It is recommended to leave FCAL_EN=1 because the lock detect mechanism always looks for VCOCal. When R0 is programmed with FCAL_EN=1, even if there are no steps required to complete the frequency calibration, the VCOCal flag will be set, permitting the lock detect to register a valid lock. If you set FCAL_EN=0, the PLL may not properly indicate lock.

    Regards,

  • 0 in reply to Derek Payne
    Intellectual 315 points

    Hello Derek,

    thank you for great explanation.

    1. I will use no-assist calibration for each LMX-device (e.g. always by instrument booting-up) and during normal operation full assist with look-up table.
    2. Yes, this is the post. OK, I will change my setting acc. you recommendation, FCAL_EN=1.

    Best Regards, V. Vojta