• State
  • Locked Locked
  • Replies 7 replies
  • Subscribers 21 subscribers
  • Views 492 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.

LMX2595EVM: Occasional Non-Responsiveness of Frequency Synthesizer

Jonathan Chhang
Prodigy 30 points
Part Number: LMX2595EVM


Hello E2E Community,

I am using the LMX2595EVM to switch very quickly between frequencies, but occasionally I find that the synthesizer switches to the incorrect frequency in an indeterministic fashion. 90% of the time the synthesizer switches to the correct frequency, and 10% of the time it does not (and also the 10% of the time that it does not, LOCK_DETECT does not go high). I start with the recommended configuration sequence - program registers from R78 to R00. Then for each frequency I want to switch to, I program the following (I am using full assist VCO calibration mode - the force registers that force VCO_SEL, DAC_ISET, and CAPCTRL are already programmed in the configuration sequence).

0) Change PLL N

1) Change PLL numerator

2) Change PLL denominator

3) Change PFD_DELAY_SEL

4) Toggle VCO_2X_EN

5) Toggle OUTA_MUX

6) Change VCO_SEL

7) Change DAC_ISET

8) Change CAPCTRL

I suspect that this issue is caused by my use of full assist calibration, as the issue seems to disappear when I add a step (9) Set FCAL_EN which to my understanding does not utilize the full assist VCO calibration.

Is there anything I can do to avoid these occasional hiccups? Thank you very much!

  • 0
    TI__Guru** 110330 points

    Hi Jonathan,

    I will look into this later this week.

  • 0
    Prodigy 30 points

    Hi Noel,

    Thanks. I have some additional information.

    When the LMX does not respond (the 10% of the time it does not respond), it actually does not switch to an incorrect frequency, but rather, outputs nothing at all, at least nothing I can see on my spectogram. Furthermore, this problem seems to happen most often when switching to higher frequencies (like 18 GHz).

  • 0
    Prodigy 30 points

    Scratch that. The LMX does indeed just switch to the wrong frequency now and then in full assist mode, even though the bits I send to full assist calibrate are the same every time.

  • 0 in reply to Noel Fung
    Prodigy 30 points

    After investigating a little further, I have additional information. I find that disconnecting my 100 MHz reference makes my PLL lock to the incorrect frequency consistently.

  • 0 in reply to Jonathan Chhang
    TI__Guru** 110330 points

    Hi Jonathan,

    Something I need clarification:

    1. OUT_FORCE = 1 and OUT_MUTE = 0?

    2. Did you program R0 with FCAL_EN = 1 every time you switch the frequency?

    3. When failure happen, if you write the same registers once again, will that fix the issue?

    4. when failure happen, would you check the voltage on pin 3, pin 33, pin 27, pin 38 and pin 35? 

  • 0 in reply to Noel Fung
    Prodigy 30 points

    1. OUT_FORCE = 1 and OUT_MUTE = 0 indeed

    2. No, I did not

    3. That does fix the issue. The PLL no longer intermittently goes to the incorrect frequency. Now, I am running into a different issue. After switching between different frequencies for about 10 minutes, the PLL stops responding. The bitstream I am sending it is unchanging but the PLL decides to stop outputting any frequency after a while.

    4. This is difficult for me to do since I have the evaluation board and no test points to acquire such measurements. I can, however, use a magnifying glass and a pointy oscilloscope probe to get these measurements if they should prove helpful in diagnosing this problem.

  • 0 in reply to Jonathan Chhang
    Prodigy 30 points

    Noel,

    Sorry for the inconvenience. I have misjudged my previous statement on #3. The PLL acts and responds the same as I initially stated in the original problem statement.