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: IMproving lock time for the device

Part Number: LMX2594


I’ve been using the pllatinumsim sw and I have a reference design where I set up Fosc=80Mhz, OSC_2x enabled, Fpd =160Mhz, 15mA=Icp & loop Bw ~440kHz.

 

The lock time requirement is <=10us!

 

Currently the reference design tells me that I should be able to achieve 10.7us of analog lock time (without CAL).

 

I’ve been working on the synth fitted on a radio prototype and I am talking through serial port.

The first SW we developed was using a “partial assist” CAL, but I’ve measured on the oscilloscope a lock time of ~ 200us stepping from 10750Mhz to 9375Mhz.

 

I decided to implement the “full assist” CAL but I am still far away from the lock time requirement.

Before running FCAL, I set up the registers on VCO_SEL, VCO_DACISET, VCO_CAPCTRL, PLL_N, PLL_DEN, PLL_NUM and the best lock time I got so far is now 40us.

 

As far as I understand the FCAL executes a calibration with two algorithms. One that ramps Vtune and an internal amplitude algorithm to optimize the phase noise.

At this stage I am not sure if this is a limit from the FCAL process or I can achieve better.

 

  • Ajayt,

    With full assist, you should be closer to 3 us and limited by the write speed to the device.

    The calibration runs a calibration that:
    1. Finds the select VCO core (VCO_SEL)
    2. Finds the correct VCO band (VCO_CAPCTRL)
    3. Finds the correct VCO amplitude (VCO_DACISET)

    So full assist is specifically writing VCO_SEL, VCO_CAPCTRL, and VCO_DACISET and that's it. There really is no calibration as you are manually forcing the values.

    But if you run any sort of calibration, even if you put the exact correct values as starting ones, it will take time.

    Regards,
    Dean