• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 20 subscribers
  • Views 594 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.

LMX2572EVM: Readback parameters not according to datasheet

Izik
Expert 1545 points
Part Number: LMX2572EVM

Hello

1. I have lock the device LMX2572EVM using autocal (no assist) and readback the values of VCO core, CAP and DACI

The selected capacitors are not according the table 136 on the datasheet. 

For example for output frequency of 4830MHz I was expecting to use VCO 4 and cap value between 136 to 25, but readback from the EVB shows VCO = 3 CAP=8 and DACIset=67

All parameters are out of range. Please advise. 

2. There's a mismatch between the EVB and the datasheet on register R20 bit 14: according to the datasheet is should be '0'. But reading from the EVM yields '1'

3. The part has a default value of DACIset = x"80. Why did you choose such value which is illegal? 

Thankslmx2572 readback parameters.xlsx

  • 0
    TI__Guru* 85330 points

    Hi Izik,

    In reality, each VCO core has some overlapping regions with its adjacent VCO core. 

    For example, the max. output frequency of VCO3 will not hard-stop at 4650MHz. Similar, the min. output freq of VCO4 could be smaller than 4650MHz. Frequency overlapping is required to ensure continuous operation without "holes".

    In this case, it is not surprise to me that auto calibration selected VCO3 instead of VCO4. The calibration algorithm will search for the right VCO starting from VCO1. If it found VCO3 is useful, it will stop here. Because of part to part variation, it is possible that some units will select VCO3 instead of VCO4. 

    FYI, the higher the capcode value, the lower the VCO freq. Your readback data told us that VCO3 is selected but a very small capcode is needed in order to tune it to a higher freq. range.

    The POR value of R20[14] is 0, but we suggest make it equals 1. Please check again with the latest datasheet. TICS Pro has already made this bit equals 1, your readback data is correct. 

    Unless you have VCO_DACISET_FORCE = 1, the value in VCO_DACISET is not used.

  • 0 in reply to Noel Fung
    Expert 1545 points

    Hello Noel

    Thank you for the fast and detailed answer. 

    1. I thought that table 136 specifies firm values which cannot be exceeded. Please advise. 

    2. Can you please advise what are the overlapping VCO frequency bands between each VCO core?

    3. I have measured locking Vtune on the EVM and the value usually is 1.3-1.4V. However on my design when I lock I get 2-2.2 Vtune on lock. Can you please explain why? I use active loop filter. 

    4, What is the Vtune voltage range? Does the Vtune port draw current? 

    Thanks again 

    Izik

  • 0 in reply to Izik
    TI__Guru* 85330 points

    Hi Izik,

    The boundaries of each VCO will overlap with its adjacent VCO by approx. 100MHz. This number will vary over PVT. It is therefore not possible to specify the absolute max. and min. VCO frequency. 

    If you do auto calibration, you don't have to worry which VCO is being selected. As long as the configuration is set right, calibration algorithm will pick the right VCO for you. 

    In assist modes, you have the freedom to pick the VCO you want.

    Is your active filter an inverted amplifier? Our VCO uses negative polarity Vtune. That is, the higher the Vtune, the lower is the frequency. In this case, if VCO3 is selected, then a smaller Vtune is requirement in order to make the VCO frequency higher. However, if you use inverted amplifier to build the loop filter, then a higher Vtune is actually required to make the VCO frequency higher. 

    With passive loop filter, the max. Vtune is less than 2.5V.

    CPout is the output of charge pump, it will either source or sink current. 

  • 0 in reply to Noel Fung
    Expert 1545 points

    Hi Noel

    You said that with passive loop the vtune max is 2.5V

    Should be any difference when using an active loop?  It is the same VCO.

    We lock typically on 2V. Is that ok, or on the edge of the VCO capacitor BW?

    Each VCO is divided to ~190 caps.   But the VCO constant is typically 60MHz/V. Is that means that roughly each capacitor has a 150MHz BW?

    Thanks

  • 0 in reply to Izik
    TI__Guru* 85330 points

    Hi Izik,

    The max. vtune for the active loop filter should also be 2.5V, because the same VCO is being locked. 

    What is the bias voltage of the +Vin pin of your amplifier?

    The Kvco is not a constant. Let's assume each VCO core covers 600MHz range, then each cap code will change approx. 3MHz. Again this is not a constant number.

  • 0 in reply to Noel Fung
    Expert 1545 points

    Hi Noel

    Can we continue this discussion on email? 

    Please send me your mail

    Thanks

  • 0 in reply to Izik
    TI__Guru* 85330 points

    Hi Izik,

    Do you know Nir?

  • 0 in reply to Noel Fung
    TI__Guru* 85330 points

    sorry, click the Reply button too fast.

    If you know NIr, you can connect to me via him.

  • 0 in reply to Noel Fung
    TI__Guru* 85330 points

    Let's continue thru email.

    Closing this post right now.