• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 21 subscribers
  • Views 130 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.

LMX2615-SP: Full-Assist calibration using a look-up table. Are CAPCTRL values temperature dependent?

Katelyn Mak
Prodigy 10 points
Part Number: LMX2615-SP

Tool/software:

Hi,

 

I am using the LMX2615-sp to change frequencies at ~100 Hz as part of a PID controller.

 

I have noticed that changing PLL frequencies using the calibration sequence (FCAL_EN = 1) causes large, undesirable spurs across a MHz to GHz range around the target frequency only occurs during calibration. Once the calibration is done, there are no longer any of these "calibration spurs", only fractional spurs remain.

My understanding is that these calibration spurs are expected due to the PLL switching VCO-cores and CAPCTRL values until lock is attained.

Using a full-assist calibration (i.e. setting the VCO_SEL, VCO_CAPCTRL, VCO_DACISET values by readback without setting FCAL_EN = 1) does not induce these calibration spurs.

I would like to store the VCO_CAPCTRL values that are read back from no/partial-assist calibrations in a look-up table so that the in normal operation, the calibration-sequence never needs to be run. I read that the VCO_CAPCTRL and VCO_SEL values change board-to-board and that the VCO_DACISET values are temperature dependent. However, I want to confirm whether the VCO_CAPCTRL values are temperature dependent as well. Ideally, I would like to set-and-forget the VCO_CAPCTRL values in the look-up table.

 

My questions are:

    1. Can I reliably program frequencies using just a look-up table of CAPCTRL and VCO_SEL values captured from a single sweep? Or do I need to conduct multiple sweeps across a temperature

        range?

    2. What is the minimum step-size I need to sweep to construct the look-up table of CAPCTRL values?

Thank you

  • 0
    TI__Guru** 107610 points

    Hi Katelyn,

    Since VCO characteristic changes with temperature, all of these parameters (capctrl, daciset, vco_sel) are temperature dependent. 

    Think about no-assist calibration, unless we need to change the frequency after it is locked, we usually would not re-calibration the device even though temperature has change. Our device will remain locked over the temperature range. In other words, the same VCO parameters are valid for any temperature. 

    Same characteristic applies to full assist. We can reuse the same VCO parameters that were obtained at a certain temperature for any temperature. Of course, for optimum performance, it is always good to have several LUT for different temperature range. These VCO parameters vary over parts, so each part must has it own LUT. 

     For the whole 7.5GHz VCO range, 500 points is sufficient for the LUT.

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Thank you!