• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 716 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LMX2592: Using a DDS as reference frequency for the LMX2592

gerhard Hofbauer
Prodigy 30 points
Part Number: LMX2592

Hi,

I would like to know if it is possible to use a DDS wth for example 50 MHz to 90 MHz as reference frequency to generate 5 GHz - 9 GHz.

While sweeping the DDS per Software command step by step, is it possible to operate LMX2592 without calibration? Just by knowing the VCO core and switching accordingly? I there also a switching VCO tuning capacitors involved in the calibration process? What would be the lowest switching delay method to implement such a sweep?

with kind regards,

Gerhard.

  • 0
    TI__Expert 7855 points
    Hi Gerhard,

    That is a large range of input frequency change. It is possible to figure out each corresponding VCO and capacitor code and manually set for each reference frequency but that is a more tedious process. That is benefit of the internal calibration process that allows you to configure all those frequencies 5-9GHz by programming and having the internal algorithm find the right VCO and capacitor codes. Are you able to implement register settings instead of using the sweep in reference frequency?

    Regards,

    Brian Wang
  • 0 in reply to Brian Wang
    Prodigy 30 points

    Hi Brian,

    As I understand it, it is possible to reach a resolution in the Hz range with the Sigma-Delta Modulator.

    What if, for example, I have got 800 points equally or logarithmically spaced in the 5 - 9 GHz range,

    what would be the average switching time from point to point, how many recalibrations would there be and how long would the recal time be.

    What would be the total sweep time?

    In comparison I could use a 5-9 GHz VCO (which we have a discrete design for) with an integer PLL say N = x100  and 50 - 90 MHz

    swept reference frequency. (If it would be a DDS, we wouldn't care about the spurious signals in the loop bandwidth). Switching speed is then very fast, when using a loop bandwidth of 500 kHz or more.

    Can the LMX2592 compete with such a solution, can it do the same job in matters of speed?

    This is for a T&M application where a frequency is set, a measurement done, next frequency point set a.s.o.

    with kind regards,

    Gerhard.

  • 0 in reply to gerhard Hofbauer
    TI__Expert 7855 points
    Hi Gerhard,

    Sorry took a while with the Thanksgiving break. I your method with the 5-9GHz VCO now and just tuning the reference frequency. The way with the LMX2592 is different but at the advantage that each calibration basically tunes and finds the capacitor and inductor bank and sets you right in the middle of the desired output frequency. In comparison with a single 5-9GHz VCO, you may have different phase noise characterisitics at top middle and bottom end of the frequency range.

    About the switching speed from one frequency to another, there is:
    1. calibration time (finding the capacitor and inductor setting, which can take in the hundreds of microseconds, but can be assisted and reduced down to the tens of microseconds).
    2. analog settling time (which happens upon completion of the calibration portion, and this is the normal settling speed that is related to the loop bandwidth as you talked about).


    Hope this helps!