• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 22 subscribers
  • Views 340 views
  • Users 0 members are here
Support feedback
Options
Options

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.

LMX2594PSEVM: Struggle to get a constant phase angle after SYNC-Pulse

Thomas Roehl
Prodigy 10 points
Part Number: LMX2594PSEVM

Hi,

I am currently working on the LMX2594PSEVM board and am having problems synchronizing the two PLL's externally.

For testing purpose I use TICS to control the PLL's. To messure the phasedifference between the two Signals I use a VNA (www.rohde-schwarz.com/.../1EZ82).

The two PLL's are both locked and working fine. The settings in TICS are as follows:

From the datasheet, I read that these settings fall under SYNC Category 3 for this use case.

But I get a strange behavior, when I send the SYNC pulse multible times (Setup- and Hold-time are above 3ns... so basically okay). Based on my current understanding, the phase relationship of the two output signals should not change when I send the pulse multiple times. However, that is exactly what is happening. In the current settings, there are two angles that (seemingly) change randomly. For example, like this:

1. Pulse: 43°

2. Pulse: 43°

3. Pulse: 36°

4. Pulse: 43°

5. Pulse: 36°

6. Pulse: 36°

...

For my use case, I require a fixed phase relationship of the PLLs after the SYNC.

Did I do something wrong with the TICS settings or did I misunderstand the whole sync feature?


Many thanks

Thomas

  • 0
    TI__Guru* 97760 points

    Hi Thomas,

    I assume you have uncheck INPIN_IGNORE bit. 

    What is MASH_RST_COUNT value? You may try to double it to see is there is improvement.

    Could you also provide a screenshot of your OSCin and SYNC signal?

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Thanks for your reply, but unfortunately increase the MASH_RST_COUNT value doesn't help.

    In the actual configuration I am using only the eval board. The SYNC-Pulse is also set over TICS.

    I know sometimes the setup- and holdtime is under the min, but this doesn't affect my problem. 

    Even when the Timing is 10 times good (checked with scope) , the output phase switch between two values randomly.

    My complete Registermap is in the 2500MHz_V2 textfile.

    2500MHz_V2.txt
    Fullscreen
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    R112 0x700000
    R111 0x6F0000
    R110 0x6E0000
    R109 0x6D0000
    R108 0x6C0000
    R107 0x6B0000
    R106 0x6A0000
    R105 0x690021
    R104 0x680000
    R103 0x670000
    R102 0x660000
    R101 0x650011
    R100 0x640000
    R99 0x630000
    R98 0x620000
    R97 0x610888
    R96 0x600000
    R95 0x5F0000
    R94 0x5E0000
    XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

  • 0 in reply to Noel Fung
    Prodigy 10 points

    Addendum: My settings work fine for other frequencies. Only the dividers and Fref were changed:

    FOUT: 2500, Fref: 64, N: 39, NUM: 1, DEN: 16, CD: 4 (NOT WORKING)

    FOUT: 6000, Fref: 64, N: 46, NUM: 7, DEN: 8, CD: 2 (WORKING)

    FOUT: 200, Fref: 32, N: 50, NUM: 0, DEN: 1, CD: 48 (WORKING)

  • +1 in reply to Thomas Roehl
    TI__Mastermind 27911 points

    Thomas,

    The phase SYNC will synchronize the dividers consistently.  However, you are seeing about 6 degrees of variation, which works out to about 7 ps at 2.5 GHz output.  Note that when you calibrate, the calibration state could be different.

    I suspect that if you read back rb_VCO_CAPCTRL from the 36 degree and 43 degree state you might find that you get two different values.  We have considerable overlap in these capacitor codes to ensure this.

    On page 22 under the section "7.3.14 Fine Adjustments for Phase Adjust and Phase SYNC", it mentions that there is about 10 ps variation due to this.

    If you want to eliminate this variation, you need to use full assist calibration.

  • 0 in reply to Dean Banerjee
    Prodigy 10 points

    Thank you Dean,


    You are right with your assumption of the cause. Unfortunately I can't find any detailed info for the full assist mode. How do you determine VCO_CAPCTRL and VCO_DACISET ? Is there also a formula for it like for VCO_CAPCTRL_STRT and VCO_DACISET_STRT in partial assist calibration?


    I have also tried the partial assist mode. This has also already solved the problem (at least with my current settings).

  • 0 in reply to Thomas Roehl
    TI__Guru* 97760 points

    Hi Thomas,

    Before you can "apply" full assist. you need to build a look up table by doing VCO calibration for all the frequency of interest. After each calibration, you can read back VCO parameters, i.e., VCO_CAPCTRL, VCO_DACISET, VCO_SEL. 

    When you apply full assist, enable all _FORCE bits, use the LUT to get the required VCO parameters for the frequency you are going to lock.

    For details, please read https://www.ti.com/lit/pdf/snaa336.