• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 20 subscribers
  • Views 958 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.

Original question:

LMX2594: LMX2594 Classic Linear FMCW Waveform

LMX2594: LMX2594 - Bandwidth for calibration-free ramping

Eric Pahlke
Prodigy 100 points
Part Number: LMX2594


My application requires a 62.5MHz ramp over 49uSec within the band 9.5-10.25 GHz.

Two previous threads have listed conflicting information about the maximum ramp size without requiring VCO recalibration:

Post 615157: "The dynamic tuning (using the varactor tuning only) for a given VCO/coarse tuning will not exceed 50-70 MHz."

Post 625236: "If you lock the device at 10 GHz, you can likely tune the VCO on the order of 130 MHz before it has to recalibrate."

Can I be reasonably certain that the LMX2594 will support a 62.5MHz ramp without requiring recalibration?  I was not able to find any info in the datasheet to help determine this range.

Can I extend the effective no-calibration-required range if I initially run the calibration at the center frequency of the ramp, rather than at the start frequency of the ramp?

Thanks!

  • 0
    TI__Mastermind 28021 points

    Eric,

    We have no guarantee for the calibration free range in the datasheet and don't test for this (at least directly).  Therefore, it might be hard to get a hard number, especially from an E2E post.  This being said, let me give you my casual opinion.

    Secondly, the VCO calibration range varies as a function of frequency.  For instance, at 15 GHz, you get much more range than at 7.5 GHz.  Also, around 12 GHz, you get a good amount of range.

    At room temperature, here is the ranges I got.  But this is zero margin, so you have to add for this.    Down is how far down, and up is how far up.  Number in MHz.  So to these numbers, I would reduce them.  So based on this, 62.5 is not a safe number for VCO1.  I would test whatever you want to do especially at hot as these numers will be lowest there.

    I do think that maybe you could take advantage of the down and up by calibratiing in the mid ramp. But make sure you have enough ramps for this.  Also be aware that RAMP_EN is in the same register as FCAL_EN.   

    Realize that these numbers I post below are just my casual opinion;  they in no way represent any sort of guarantee.

    Regards,
    Dean

    VCO Core Down Up
    VCO1 66 46
    VCO2 72 46
    VCO3 88 56
    VCO4 92 55
    VCO5 159 103
    VCO6 162 99
    VCO7 173 108

  • 0 in reply to Dean Banerjee
    Prodigy 100 points
    Dean,

    Thanks for this good information. It looks like using VCO3 (which covers a lot of my potential band) and calibrating mid-ramp, there is a fairly good chance this will work. However, is there another part which might be better suited for this application?

    Do you measure these ranges by adjusting the frequency and seeing where lock detect is lost?

    Also, can you clarify your comment "But make sure you have enough ramps for this."?

    Finally, with regards to your comment "Also be aware that RAMP_EN is in the same register as FCAL_EN. " - It sounds as if there is a potential issue I might run into because these are in the same register, can you clarify what the potential issue is?

    Thanks,
    Eric
  • 0 in reply to Dean Banerjee
    Prodigy 100 points
    (See follow up questions below)
  • 0 in reply to Eric Pahlke
    TI__Mastermind 28021 points
    For these measurements, what I did was first lock the VCO to frequency. Then increase the input reference frequency to the PLL that makes the VCO slam to the upper tuning range. Then I decrease the input frequency which makes the VCO slam to the lower tuning range.

    For the "But make sure you have enough ramps for this." means this. When you set RAMP_EN=1 the ramping starts. So you need one ramp to go from the center frequency to the lower frequency. Then another ramp can go from low frequency to high frequency. If you set RAMP1_RST on RAMP1 and RAMP1 _NEXT=RAMP1, I think this might work. But if you want to do anything more, like a triangle wave, you would need another ramp, ahd the part only has two.

    I'm thinking that when you enable RAMP_EN=1, you will be triggering FCAL_EN=1 and triggering the VCO calibration. So if you were thinking to lock the PLL in the middle of the range, but then start the ramp at the bottom, this would make the VCO calibrate back to the center of the range when you enable RAMP_EN=1 because you are triggering with FCAL_EN=1. The action of writing register R0 with FCAL_EN=1 triggers the ramping.

    Regards,
    Dean
  • 0 in reply to Dean Banerjee
    Prodigy 100 points
    Thank you. I'll do some experimenting with the eval board and see what happens.