• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 20 subscribers
  • Views 1219 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.

CDCE913 Jitter Problem

Rob Froyd
Prodigy 20 points

Other Parts Discussed in Thread: CDCE913

I am having problems with Jitter on a CDCE913 Clock PLL.

On a video Decoder System, our SD-SDI Timing Jitter is > 0.2 UI.  The video clock is being generatoed by a CDCE913.  If I turn the VCXO via a register, Timing Jitter < 0.2UI.

If I use a precision voltage source into the VCTR input, my Timing Jitter is too high.

If I use the EValuation Board to produce the Video Clock, and I operate the Evaluation Boards PLL, VCTR input with a precsion voltage source, my Timing Jitter < 0.2 UI.

It would appear that something is wrong with our "Videe Decoder's" circuit design, and/or layout of the CDCE913.

What should I look for to fix this problem?

Can the Crystal selection cause a Jitter Problem?

Noise on the Voltage Source?

It would seem, that it is some sort of noise into the VCTR input.

Sincerely,

Rob Froyd

 

  • 0
    TI__Expert 5725 points

    It might be a signal integrity issue.

     

    How is the jitter being measured? Are you measuring the jitter at the output of the CDCE913 when is driving the video decoder circuit with and without Vctr? How does the clock output looks like? Would it be possible to get some screenshots of the CDCE913 output? What is the setup of the device? If you are using the evaluation board TI Clock Pro, would it be possible for you to post the setup used in .hex format? (File>Save .Hex)

     

    What is the ripple amplitude in the Vctr pin? Would it be possible to get an screenshot of this signal too?

    Are you using the CDCE913 as a VCXO or just as a crystal oscillator? If the CDCE913 is being used as oscillator, then Vctr can be left open (as it has an internal pull up resistor).

     

    When using the CDCE913 as a clock generator, then the input crystal stage should be set to XTAL.( (by setting register 01xh bits 3:2 to 00)  When used as a VCXO then the input stage selected should be VCXO (by setting register 01xh bits 3:2 to 01)

     

    The crystal selection would not cause a problem in the jitter, as well as, but care must be taken when selecting the load capacitance. If the CDCE913 is used as a VCXO In that case the XTAL selection should be made targeting a pulling range, paying special attention to the parameters Cs, Cm, Lm and Cl. Usually they are not provided by the crystal manufacturer unless specially requested. You can read more in the following post:

     

     

    (http://e2e.ti.com/support/clocks/f/48/p/56234/200872.aspx#200872).

    There are some applications notes on the topic on selecting a XTAL for the CDCE913 working as a VCXO :

     

    http://focus.ti.com/general/docs/techdocsabstract.tsp?abstractName=scaa088

    http://www.ti.com/litv/pdf/scaa085

  • 0 in reply to mapefer3
    Prodigy 20 points

    1) We are using the CDCE913 as a VCXO.

    2) We understand setting register 1 for XTAL or VCXO.

    3) Our Jitter Performace is fine if we set the CDCE913 to XTAL mode.  This expirement, we believe points to a problem at the CDCE913.

    4) With the Crystal we have selected, and then pulling the VCTR manually, we have a pull range from +30 PPM to -62 PPM.  (26.998320 MHz to 27.000820 MHz). 

    5) I will review the scaa088.pdf app note.

    6) We measure the Timing Jitter at the output of our Video Decoder Box, at the SD-SDI output.  We use a Tetronix WFM 7120, that is built for the purpose.  The video system has many parts, but we have isolated the start of the problem at the CDCE913.

    7) I thought the output clock looked pretty good.  We can investigate getting a screen shot. 

    8) Device set up.  We have a DSP prgram that programs and reads the registers.  Here are the numbers, from register 0 on up.  We have 3 outputs 27 MHz, 74.175824 MHz, and 27 Mhz.  The 3rd output is the one that finally reaches the SD-SDI Video Decoder Box Output.  Numbers below:

    0x81,0x05,0x34,0x01,0x02,0x80,0x40,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
    0x00,0x00,0x00,0x00,0x4D,0x02,0x03,0x00,0xEA,0x66,0xE2,0x07,0x00,0x40,0x02,0x08,
    0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,

    9) We have Jitter problems with a straight DC voltage (0.9 Volts) at the VCTR pin on our board.  So in theory there isn't a control ripple, but possible a noise ripple.  We can try and quantify that with a measurement.

    10) My gut feeling is we have a power supply problem, or a noise problem at the VCTR input. 

  • 0 in reply to Rob Froyd
    TI__Expert 5725 points

    Hi Rob,

    I am looking forward your screenshots from point 7 and 9,