• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 178 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.

LMX2595: switching time

SM
Expert 1730 points
Part Number: LMX2595
Other Parts Discussed in Thread: PLLATINUMSIM-SW,

Hello,

    For an application we need to have very fast switching and settling times. 

May I ask you to please provide these numbers for the complete range (up to 20GHz). There is a mention of 20us VCO calibration speed, is there a connection to the switching time. 

What is the mechanism to update the switching frequency with a deterministic latency? 

Also, the datasheet doesn't speak much about the performance above 15GHz, could you please provide the phase noise spec. 

2) In order to generate signals above 20GHz, can a frequency doubler be used? 

3) What are the PFD spur levels? Also, near the carrier are there any other kind of spurious? 

4) Is a frequency resolution of 1KHz possible? 

5) How is output frequency accuracy calculated from the reference frequency stability/accuracy? Is an accuracy of 10KHz possible? 

6) For the phase noise plots provided in the datasheet, I assume they are for a certain loop bandwidth? How is the switching/settling time calculated based on this?

Thank you for your help in understanding these points,

  • 0
    TI__Genius 11900 points

    Hello SM,

    The phase noise performance for LMX2595 can be simulated on PLLATINUMSIM-SW. Including the performance in the 15-20 GHz range. Please let me know if you run into any issues and I can help you resolve them. Additionally we actually do have plots for 15 GHz - 20 GHz outputs, please refer to Figures 10 -14.

    The VCO calibration depends on a number of factors that are described in this application note: https://www.ti.com/lit/an/snaa336/snaa336.pdf

    Additionally the lock time can also be simulated on PLLATINUMSIM-SW.

    The LMX2595 also has a mode called Full assist that shortens the VCO calibration procedure to only include the analog lock time.

    SM said:
    What is the mechanism to update the switching frequency with a deterministic latency?

    If you are looking for deterministic lock time, this would depend case to case. The time it takes to switch frequencies depends on the loop filter, bandwidth and phase margin, and consequently your PLL settings as well as input frequency and start and stop frequency. Here's a video describing the the transient response and lock time: https://training.ti.com/ti-precision-labs-clocks-and-timing-pll-transient-response 

    SM said:
    2) In order to generate signals above 20GHz, can a frequency doubler be used? 

    Unfortunately we do not have a high frequency doubler available at TI right now. The team is aware of this need. If the doubler input is compatible with the output of the LMX2595 then go ahead.

    SM said:
    3) What are the PFD spur levels? Also, near the carrier are there any other kind of spurious? 

    Spurs levels can also be calculated on the PLLATINUMSIM-SW tool. The location and magnitude of the spurs will depend on the particular settings of the PLL. Typically a fractional PLL will produce more spurs than an integer PLL.

    SM said:
    4) Is a frequency resolution of 1KHz possible? 

    Yes. The N divider denominator of the PLL is a 32-bit field and provides plenty of precision.

    SM said:
    5) How is output frequency accuracy calculated from the reference frequency stability/accuracy? Is an accuracy of 10KHz possible? 

    The output will follow the input. If the input is accurate and stable, the output will be to. For example if the input drifts by 1 ppm then the output will also drift 1 ppm.

    SM said:
    6) For the phase noise plots provided in the datasheet, I assume they are for a certain loop bandwidth? How is the switching/settling time calculated based on this?

    The phase noise plots on the datasheet are based on the EVM loop filter. The loop bandwidth is around 300-400 kHz and will depend on the specific frequency. We did not measure the lock time for this, but based on a simulation on the tool, the lock time in Auto-calibration mode is around 80 us. In full assist mode this can be shortened to around 20-25 us, while keeping the default loop filter.

    Thanks,

    Vibhu