• State Verified Answer
  • Replies 3 replies
  • Subscribers 151 subscribers
  • Views 128 views
  • Users 0 members are here
Options
Related

ad9552 frequent out is not right

johnjf @ go
on Dec 22, 2020

dear ADI

  SOS

     Use AD9552 ,

      set 10M clock REF input

       A0-A2  000

       Y0-Y5   000100    

     Use the DATACODE  1031 , use SPI control ,the output is 62.5MHz   ,it is ok 

 but  use the DATACODE  1912 , use SPI control , same SW, the output is 69MHz   ,it is not right 

why ? what the the difference between the datacode? how to get the right 62.5Mhz output frequent using 1912 datacode ?

the circuit as below 

Help to check the reason 

thanks a lot 

  • +1
    •  Analog Employees 
    on Dec 22, 2020 4:18 PM 2 months ago

    Hi,

    You state Y0 to Y5 = 000 100, yet you talk about a 62.5MHz output, which may mean Y5 to Y0 = 000 100.

    Then, when you say a chip with a date cade on the package 1031 works, you talk about SPI control. So you preset the chip with the A2 to A0 and Y5 to Y0 pins, and then you change some register setting. But you do not send me the register setting, so I cannot comment on it.

    The data sheet rev E, page 13 states that if you change the VCO frequency band using the SPI interface, you need to force a VCO calibration. If you do not do this, the chip would not generate an exact 62.5MHz. So do you change the VCO band using the SPI?

    Regarding the schematic, these are my notes. I'll point if a choice was bad.

    - pins XTAL have capacitors to ground. But the data sheet states at page 8 that if you do not use a crystal resonator, these pins should be left floating.

    - pin REF does not show its schematic, so I cannot comment on how the 10 MHz reference clock is provided to the chip

    - OUTSEL is connected to ground, which means the outputs are LVDS

    - RESET pin: I do not understand why the pull up capacitor C18 on it. Then you seem to drive it from a controller. 

    -LOCKED pin: did this pin indicate LOCKED status for the chip  1912? I suppose it indicated LOCKED status on chip 1031.

    - OUTx pins: you do not show their schematic. I hope they use one of the schematics in figures 17 or 18, page 12 from the data sheet.

    My bottom line is that the schematic has the problem at XTAL pins. Just take out the capacitors. Make sure the RESET pin is kept low by the controller. Then make sure you calibrate the VCO if you  change its band using the SPI.  

    Petre

  • 0 on Dec 23, 2020 1:44 AM 2 months ago

    dear Petre 

    thanks a lot .

    list my register setting as below ,help to check bugs 

    addr <=X"19"   data<= X"C1" 

    addr <=X"18"   data<= X"36" 

    addr <=X"17"   data<= X"00" 

    addr <=X"16"   data<= X"00" 

    addr <=X"15"   data<= X"00" 

    addr <=X"14"   data<= X"08" 

    addr <=X"13"   data<= X"30" 

    addr <=X"12"   data<= X"75" 

    addr <=X"11"   data<= X"BA" 

    addr <=X"33"   data<= X"00" 

    addr <=X"05"   data<= X"01" 

    addr <=X"0E"   data<= X"04" 

    addr <=X"05"   data<= X"01" 

    addr <=X"0E"   data<= X"84" 

    addr <=X"05"   data<= X"01" 

  • 0
    •  Analog Employees 
    on Dec 23, 2020 2:17 PM 2 months ago in reply to johnjf @ go

    Hi,

    The registers seem to be set correctly and the VCO calibration seems correct. But I did not try it on an evaluation board.

    I found only two things, but they seem not consequential to me:

    - Not clear to me why register 0x19 has a value of 0xC1 when the data sheet says bits 6:0 are unused.

    - Not clear why the registers 0x14 and 0x13 have non zero value (that is Modulus=480,000) while the registers 0x15 and 0x14 are at 0 (that is Fractional part of SDM equal to 0). 

    Did the LOCKED pin: indicate LOCKED status for the chip  1912 and for 1031? 

    Did you try this program on a 3rd chip?

    Did you read back the registers to see they have the value you programmed?

    Petre