• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 3 answers
  • Subscribers 21 subscribers
  • Views 523 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.

LMX2594: Influence of reference phase noise on total noise

user5959213
Intellectual 865 points
Part Number: LMX2594

hi all,

  If i know my reference phase noise,how i estimate my pll output phase noise within loop bandwidth ?My pll is lmx2594.

From dean's pll book, i know that there are many blocks affect the pll noise,such as input buffer,mulipilier,R divider,PFD and charge pump.It's too complicated to calculate  the output phase noise because i have to add all this noise together . Is there a simple estimation formula ?

 thanks.

  • 0
    Intellectual 865 points

    hi all,

     From lmx2594 EVM instructions page 17,i see that there is a note said that a noise source 10dB down from the pll noise will contribute to raise the noise by 0.4dB, and how this calculate it?

  • 0 in reply to user5959213
    TI__Guru** 108410 points

    Hi There,

    Please use PLL Sim to do the simulation. You can upload the phase noise of your ref clock to this tool to make the simulation more accurate. 

  • 0 in reply to Noel Fung
    Intellectual 865 points

    hi Noel,

     Yes,I have use pll sim do the simulation and I can see all the parts influence the output phase noise from the plot.here I want to quantify every parts contribution.do you have theoretical derivation to get the plot?

    Thanks

  • 0 in reply to user5959213
    TI__Mastermind 28011 points

    Hi,

    PLLatinum Sim is based on my PLL book and was actually developed at the same time as the 5th edition was being.  The derivations are in the PLL book.

    From the comment of 10 dB down contributing 0.4 dB.

    Noise_PLL = x dB

    Noise_OSC = x-10 dB

    Noise_Total = 10*log(  10^(Noise_PLL/10)+ 10^(Noise_OSC/10)   )

    = NOISE_PLL + 10*log( 1 + 10^(-10/10) ) = NOISE_PLL+0.41

    Regards,

    Dean

  • 0 in reply to Dean Banerjee
    Intellectual 865 points

    hi dean and Noel,

     Thank you very much! I have down your book and print it for a better study.

  • 0 in reply to Dean Banerjee
    Intellectual 865 points

    Hi Dean,

         I have a problem for the calculation.The value i calculate  is not agree with the software(pllatinum sim) calculated.i don't know where I went wrong.My calculation process is as follows:

         I use the following formula to calculate pll_noise_total(offset)=10*log(10^(pll_noise_flat/10)+10^(pll_noise_1/f(offset)).

    while pll_noise_flat=Pll_FOM+20*log(Fvco/Fpfd)+10*log(Fpfd/1Hz) and pll_noise_1/f(offset)=PLL_Flicker_Norm+20*log(fVCO/1GHz）-10*log(offset/10Khz)

    here pll is lmx2594,so Pll_FOM=-236dBc/Hz,PLL_Flicker_Norm=-129dBc/Hz.Then i can get pll_noise_total(offset) value with the help of the formula as follow:100hz:-87.41dBc/Hz;1k:-95.098dBc/Hz;10k:-98.357dBc/Hz,100k:-98.871dBc/Hz

        And the Oscin(368.64M) parameters is as follow:100hz:-93dBc/Hz;1k:-119dBc/HzdBc/Hz;10k:-133;100k:-138dBc/Hz.

      The vco output is 12000Mhz,with oscin scaled(20*log(fpd/oscin))  i can get the oscin phase noise as follow:100hz:-62.748dBc/Hz;  1k:-88.748dBc/Hz; 10k:-102.748dBc/Hz; 100k:-107.748dBc/Hz.

       My loop filter BW is 200kHz,so i think the phase noise within the loop is decided by the pll_noise_total and the oscin noise.so i add them to get the 100Hz to 100k offset phase noise.but the value i calculate  is not agree with the software(pll sim) calculated. The software calculated is much better than i calculate.I don't know what's wrong.Is there anything wrong with my calculation?

    Thank you very much.

  • 0 in reply to user5959213
    TI__Guru** 108410 points

    Hi there,

    The loop filter effect has to be taken into account as well. Suggest read AN-1879 for details. http://www.ti.com/lit/an/snaa062a/snaa062a.pdf

  • 0 in reply to Noel Fung
    Intellectual 865 points

    Hi Noel,

    I checked my calculation carefully against the software calculated,I found the calculated value within 10K is in good agreement with the software simulation value, and there is a discrepancy between 100k and later.And i found that  the influence of VCO at 100K i have not taken into account.The noise of VCO and PLL at 12G is equivalent.So there is another problem. From the perspective of software, the noise of VCO at 100k is -113.53657dbc/hz,i didn't know how to get this?From the datasheet,the noise of VCO at 12G is around -103dBc/Hz.I have doubts about this offset at present.

    Thanks.