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.

LMX2592EVM: can't lock when VCO switched to doubler mode

Genius 5355 points

Part Number: LMX2592EVM
Other Parts Discussed in Thread: LMX2594, LMX2592, LMK61E2

Hi Support,

Encountered can't lock when VCO switched to doubler mode.

Do we have to do anything on the EVB to support VCOx2?

Please help on the recommendation on increasing the dynamic range of the input signal, currently phase noise degrades a lot when change in Reference level.

Thanks.

  • Hi Ikon,

    Are you using the software TICS PRO? If you enter in the target frequency in the output frequency box, for example 9GHz, the software will automatically change all the settings you need. I suspect some setting was not updated right since it looks like you can lock with normal VCO mode (which means the device should operate fine) and just something wrong in doubler mode. Perhaps VCO2X_EN=1 was turned on and PLL_N_PRE was not adjusted?

    What levels are you using for the reference?
  • Hi Brian,

    Confirmed that VCO2X_EN=1 and N divider =4.

    Do we need to change the loop filter values since the tuning sensitivity is reversed when switched to doubler mode?

    Please provide us the some default Register settings for VCO X2 frequency for confirmation.

    Currently we use 50MHz 0.7V P-P signal.

    In addition, for LMX2594, is it recommended to use 100MHz as reference? Any issues?

    Thanks.

  • Ikon,

    Yes, it's probably easier to use an example. See attached is a TICS Pro file to load and try (Fosc=50MHz and Fout=9GHz). Yes, essentially with doubler mode the tuning sensitivity doubles, since for every same vtune change, output frequency changes by 2 times of the fundamental VCO.

    For the LMX2594, you can use any input reference frequency within the electrical specs, the recommendations for higher frequency reference is due to that you theoretically get a benefit of lowered PLL flat noise.

    Regards,

    Brian Wang

  • Hi Brian,

    Please advise attachment.

    Thanks.

  • Hi Brian,

    Is the example tested for doubler mode? cos its not working.

    Thanks.

  • ikon,

    Yes, I see a clean 9GHz signal. Maybe it is software related. Do you have a recent version download of TICS Pro? What does your spectrum look like after programming? It is completely off from 9GHz?

    Regards,

    Brian Wang

  • Hi Brian,

    Is there any programming sequence to set VCO Doppler Mode to x2?

    Reason is that when we load Doppler Mode x2 to the device after powered up, the required frequency did not locked.

    But when we set the Doppler Mode x1 after powered up and then set Doppler Mode to x2, the required frequency could lock.

    Please advise. Attached the screenshots.

  • Hi Brian,

    In addition... on page 37 of the datasheet (http://www.ti.com/lit/ds/symlink/lmx2592.pdf)... What is the clock frequency based on for 8.2.1.3 Application Curve?

    Thanks.

  • Ikon,

    That is a 100MHz input frequency using a very low phase noise Wenzel Crystal because that is the best source we had. Then the input doubler is enabled to get PFD=200MHz. The same curve should be attained with low noise 200MHz input source and bypassing the input paths for PFD=200MHz.

    Regards,

    Brian Wang
  • Ikon,

    I see you use the TICS Pro from your screen shot normally. I do "USB communications" --> "Write all register" (or CTRL +L shortcut) and you see it writes everything from highest register to lowest. As long as the lowest register is written last it is OK. I'm wondering, you said you had 50MHz input but I saw in your screen you set for 100MHz. When you locked to X1 mode, is the output signal a clean/strong signal >+5dBm or so? Another thing is if you can write a RESET=1 --> RESET=0 (this does a soft reset), which clears all the registers, I usually do this if get a failed lock, maybe some registers get stuck in wrong state.

    Regards,

    Brian Wang
  • Hi Brian,
    On the USB communication... we just click on load registers. So not exactly sure what the different from what you mentioned. Please enlighten.
  • Hi Brian,

    In addition, need your advise how to generate 200MHz clock from 50MHz clock using TI clock generator, since phase noise is better with 200MHz clock for LMX2592 chip.

    Is there any clock generating chip from TI which not sensitive to vibration on reference?

     

    Thanks.

  • Ikon,

    Have you tried the RESET toggle?
  • Ikon,

    The LMX2592 has a multiplier in the input path that will work for this, 50MHz input --> MULT=4 --> PFD=200MHz.

    For clock generating, is this just one output reference for the LMX2592? Take a look at the LMK61__ family of products (for example the LMK61E2). It is a programmable oscillator, and there are fixed frequency variants too. This one has mechanical shock and vibration qualifications (MIL-STD-202, Method 213, and MIL-STD-202, Method 204). Let us know if this is the kind of product you are looking for. Thanks!