Question about AD9549 output clk frequency


My customer use AD9549 and ask some question about AD9549 output clk frequency in their below test result.

They set AD9549 PLL output with S/R Devider x4, and then test below conditions.

1) If source is 25MHz, output of AD9549 is exact 100MHz. (25x4 = 100)

2) But if source is 49.952Mhz, output of AD9549 is 199.840MHz, not 199.808MHz. (49.953 x 4 = 199.808)

Q1) Is above result correct? ( Source input is a 10 bit digitized signal.)


Q2) Would you let me know whty this difference is occurred?

       Is this a PLL characteristic? conversion error? error from PPM? 

      or need to review schematic & configuration? device fault?

Please advise me.




  • HI,

    in order to determine the AD9549 configuration, they must have used the AD9549 evaluation software. Please send me the stp file they created.

    Then I'll respond to your questions


  • Dear Petre,

    Thank you for your kind concern.

    Please refer attached their STP file.




  • Hi,

    I took an AD9549 evaluation board, I provided REFA=10.24MHz (from an AD9545 evaluation board, HSTL, 15 mA driver) and SYSCLK=25MHz (from a Rohde & Schwartz SMA 100A, 2dBm), I launched the evaluation software and I loaded the stp program you sent me. I then measured the OUT frequency with a Agilent E5052B as 24.479998 MHz.

    So the output was exactly the desired frequency.

    You say: "If source is 25MHz, output of AD9549 is exact 100MHz. (25x4 = 100)"

    When you say source, I suppose you refer to REFA, so if I set the feedback divider to 4, I would get a 100MHz. I changed accordingly the setup and save it. I then powered down the board and closed the evaluation software. I then powered back the board (this time I provided 25MHz at REFA from the AD9545 evaluation board, HSTL, 15 mA driver), launched the evaluation software and loaded the setup. I measured 99.999988 MHz, which I believe is very good.

    I then repeated the procedure for REFA=49.952MHz. I obtained 199.807976MHz, which I consider very close to the targeted 199.8080000000010 MHz indicated by the evaluation software.

    I suppose the error you get if from the clock source which you say is digital on 10 bits. I recommend measuring the 49.952MHz clock to see if the frequency is the one desired.

    I attach the stp files I used for REFA=25MHz and REFA=49.952MHz.