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.

  • TI Thinks Resolved

LMK61E07: I2C communication interfered with clock output

Prodigy 90 points

Replies: 11

Views: 963

Part Number: LMK61E07

Hi there,

I've noticed that the phase noise of the LMK61E07 is not as good as it was when I2C operations taken place.

I've tried periodically sending commands to change the LMK61E07 frequency or sending command just to read the product ID, or sending non-I2C PRN data to the I2C ports on the chip.

All the above three significantly worsen the phase noise of the LMK61E07 at the sideband related to the I2C port's communication.

Please see the following figure attached.

Thanks.

Regards,

Wade

  • Hi Weida,
    Can you share your LMK61E07 schematic with power supply ? Just want to check if there are other causes.
    I can expect there is some spurs related with I2C clock, but do not expect there is such worse phase noise degraded.

    Regards,
    Shawn

    ____________________________________________________________________________________

    To design your own Clock Tree solution, visit WEBENCH Clock Architect ​​​​

    More information Clock and Timing System products: http://www.ti.com/clock-and-timing/overview.html

  • Hi, please check the attachment for what we've seen in our EVM when we do continuous write on I2C bus. As Shawn recommends, please share the schematics for us to review. Thanks a lot.

    LMK61xx I2Cwrite phase noise.pptx

  • In reply to Madhu Balasubramanian:

    Thanks both!

    I've been using silicon labs evaluation board, Si5xxUC-EVB.

    www.silabs.com/.../ug298-si5xxuc-evb-ug.pdf

    The microcontroller has been disabled.

    Three wires, SDA, SCL and GND are connected from Xilinx KC705.

    Is that just because the evm is from a wrong company...?

    There doesn't seem to have significant defects and I've added quite a few decouple caps as well.

    Thanks a lot.

  • In reply to Weida Zhang:

    Even without I2C active, the phase noise doesn't look as clean as what we measure. Did you do hand solder on the Si Labs board? Did the reflow profile follow what's given in JEDEC-STD-020?
  • In reply to Madhu Balasubramanian:

    We didn't use reflow. It was hand soldered.

    Do you mean this may due to mechanical stretching/tension?

    We can certainly buy another eval board and chip to do a proper reflow if the soldering is likely the cause to the fault.

    Thanks for your feedback!

    Regards,

    Weida

  • In reply to Weida Zhang:

    I suspect this could be due to mechanical tension from the solder profile exceeding recommended reflow temperatures. Please do try machine solder with right profile and let us know the results.

  • In reply to Madhu Balasubramanian:

    Thank you for your insight! We will try it and post the result here.
    Regards
  • In reply to Weida Zhang:

    Hi guys,

    We have bought the LMK61E2EVM and tested it.

    It appears the default frequency is as the datasheet specified, however, when set to a different frequency (125MHz is the frequency we wanted to use it) the performance is much worse.

    Please see the attached figure.

    I tried to alter the internal PLL setting, however, the Ko (VCO gain) seems not mentioned anywhere.

    Also, I found that when the EVM is powered with ext 5V, and let the on board LDO reg to provide the 3.3V, the noise is much worse than using an external 3.3V to power the chip directly.

    Any suggestions?

    Thanks.

    Regards,

    Weida

  • In reply to Weida Zhang:

    Is this issue still open? If so, can you please share your LMK61E2 programming config file (.tcs) that you used for your testing? Looking at the yellow trace, it appears the registers settings may not have been correctly programmed.

    In the LMK61E2 programming tool, did you enter 125 (MHz), select "LVPECL" output type, click "Generate Configuration", write all registers, then measure the output clock phase noise (preferably using a balun for differential-to-single-ended conversion).

    Alan
  • In reply to Alan O:

    Please find the updated result attached. I've tried to upload the TCS file, however the website avoids me from doing so. The registers values are provided following.

    R0 0x0010
    R1 0x010B
    R2 0x0233
    R8 0x08B0
    R9 0x0901
    R16 0x1000
    R17 0x1180
    R21 0x1502
    R22 0x1600
    R23 0x172A
    R25 0x1900
    R26 0x1A34
    R27 0x1B20
    R28 0x1C00
    R29 0x1D01
    R30 0x1E3F
    R31 0x1FFF
    R32 0x20FF
    R33 0x2103
    R34 0x2224
    R35 0x2327
    R36 0x2424
    R37 0x2502
    R38 0x2600
    R39 0x2707
    R47 0x2F00
    R48 0x3000
    R49 0x3110
    R50 0x3200
    R51 0x3300
    R52 0x3400
    R53 0x3500
    R56 0x3800
    R72 0x4802

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.