• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 21 subscribers
  • Views 565 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.

CDCE62002: CDCE62002

Chris Cotton1
Prodigy 80 points
Part Number: CDCE62002


Hi,

I am using CDCE62002.  I am attempting to change the loop BW for a board that is already in production. My goal is to select a BW that filters jitter from a noisy reference clock (>1ps rms) but also be able to track some modulation (~1khz) on the input reference clock without causing cycle slip or similar loss of lock issues.

My current BW is 20kHz (icp=3ma, PFD = 31.25Mhz, VCO=2GHz)

My fixed constraints are:

Input clock: LVDS 250Mhz

Output clock: LVDS 250Mhz

C1 33nf
C2 1uf

R2 39

Goal BW is between 80kHz to 100kHz. When I try using the EVM software, I have been unable to find a result that meets this desired bandwidth and satisfies Gamma Optimization and T3/T1 ratio. Can you tell me if the CDCE62002 is capable of producing this BW given the fixed constraints?

Thanks,

Chris

  • 0
    TI__Mastermind 28560 points
    Hello Chris,

    Typically increasing charge pump current, minimizing loop filter components, and reducing total PLL N will help you design for wider loop bandwidth....

    By changing the LFRCSEL to 0x0f (R3 = 5k and C3 = 38 pF with 3.75 mA charge pump), I was able to increase to 24.5 kHz loop bandwidth.

    Unfortunately the maximum phase detector frequency is 40 MHz, otherwise you would have been able to set PLLR=6 and re-program PLLN=24 (assuming prescaler = 2) which would result in the same 2000 MHz VCO frequency but with a smaller N which helps to increase loop bandwidth... I was able to get 32 kHz LBW with LFRCSEL = 0x0f. But again, this is out of spec operation so I don't recommend it.

    73,
    Timothy
  • 0 in reply to Timothy T
    Prodigy 80 points

    Thank you for trying. I was able to come up with one solution that results in 114kHz BW. The issue is that the Gamma Optimization is slightly out of range as well as the T3/T1 ratio but I find that this ratio tends to be reported incorrectly by EVM version 1.2.1 which is what I am using. I would attach the EVM software file for you to import but I don't see a place for attachments here. Soof course it is just a text file so you can evaluate by creating your own .ini file with these settings

    The jist is INPUT:=250MHz, OUTPUT:=250MHz, PFD =10.417MHz, VCO=2000MHz

    REGISTERS
    0 52260550
    1 878B6171
    2 00000002

    PORTS
    0 DD
    1 FF
    2 DF
    3 F9

    INPUTS
    PRI 250
    AUX 25

    EXTERNAL COMPONENTS
    C4 1000n
    R4 39
    C5 33n

    Thanks,

    Chris