• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 11 subscribers
  • Views 650 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.

DAC38RF83: tuning and baseband operation

Scott Greeley1
Intellectual 585 points
Part Number: DAC38RF83


Hi.     Two questions on the DAC38RF83:  

1) Can you provide a more detailed procedure for tuning the internal PLL/VCO?   Figure 167 in the revC datasheet indicates that it needs to be tuned, but doesn't say exactly how to do this.   What readings from the Page 0, reg 6 are acceptable, and if they're not acceptable, how should we adjust the PLL_VCO value of Page 4, reg 0x33?    Note: we're using the DAC38RFxx GUI, but only for 'simulation' purposes to determine register settings.   We're using our own software for configuring the DAC.

2) We want to operate one of the multi-DUC paths using baseband data (no mixing).   Can you confirm that to do this, all we need to is to disable the mixer and NCO (using the MULTIDUC_CFG2 reg)?    Then there will be no mixing, and the interpolated I & Q data will be summed and output to the DAC.    If we set the Q data to zero, this should output baseband signal based on the interpolated I data from 0 to Fbw Hz.   (where Fbw ~= 0.8 * Fs/(2* interp factor) )

thanks!

Scott

  • 0
    TI__Intellectual 1400 points

    Hello Scott,

    Your questions have been acknowledged and we are currently working through solutions to provide you.

    Best Regards,

    Kelvin

  • 0
    TI__Intellectual 1400 points

    Hello Scott,

    Please refer to the document attached regarding your questions on the DAC38RF83 

     E2E_Response_Scott_DAC38RF82EVM.pdf

    Best Regards,

    Kelvin

  • 0
    Intellectual 585 points

    Hi Kelvin,   

    Sorry, I should have said that we are NOT able to use the DAC38RFxx EVM GUI with our custom hardware.    (we are using it for simulation only purposes, to assist in determining the DAC38RF83 register settings).

    The block diagram in the datasheet (Figure 167) says to increment/decrement the VCO Tune value (page 4, reg 0x33, bits[14:8]), when the LFVOLT reading is out of range.     However, it doesn't say how much to increment/decrement, or which direction to do so?    For example, do we increment when LFVOLT is under range, and by how much?

    Thanks for your assistance, and detailed response to our questions!

    Scott

  • 0 in reply to Scott Greeley1
    TI__Intellectual 1400 points

    Hello Scott,

    Ideally the value of LFVOLT should be kept between 3-5 where if under range the value should be incremented by 1 until the value reaches the optimal range and the PLL is locked. Similarly, if it is over this range the value should be decremented by 1 until this optimal range is met. 

    I have included the images below to describe the relationship of the LFVOLT value with respect to the PLL so that in your custom designs you can have more flexibility in how you wish to adjust the LFVOLT values as it relates the PLL system. 

    In Figure 74, we examine that in addition to manually adjusting the VCO Tune value through manually adjusting the registers, (page 4, reg 0x33, bits[14:8]), the VCO tune value is manipulated by the PFD and charge pump. This subsystem will be further examined below. 

    As shown above the PFD and charge pump combination work in tandem to produce the LFVOLT value in question. The multiple dependencies that affect this subsystem are what are adjusted to reach an optimal LFVOLT value between 3-5 when using the "PLL AUTO TUNE" function as previously demonstrated in the GUI. Given that you will not be using the GUI I am making note of this relationship so that you have will have a broader understanding as to the factors that play into the LFVOLT values should they need to be referenced in your designs.

    Furthermore, I am also adding this table above from the datasheet to describe how to manipulate the charge pump if needed above. Once again, LFVOLT can be manually incremented and decremented by directly manipulating the register values however I wish to explain how such value is being developed in the first place as the absence of the GUI configurations leads to further options that you can explore in how you wish to adjust either the LFVOLT value itself or the factors that feed into its' creation. 

  • +1 in reply to Kelvin Ayala
    Intellectual 585 points

    Thanks Kelvin!    We're almost ready to test our new custom hardware, and this will be very useful.

  • 0 in reply to Kelvin Ayala
    TI__Intellectual 1400 points

    Hello Scott,

    I am adding some clarifications and distinctions to this thread in the case that it should be referenced in the future. To clarify the LP filter is better described as being maintained by the PFD and charge pump rather than being manipulated. This is caused through the negative feedback in the system. Overall, there are 2 approaches in altering the PLL VCO: by directly altering the PLL loop voltage register values and/or adjusting the charge pump that will consequently change the loop's behavior by changing the loop's bandwidth.