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.

LDC1000-Q1: Non-monotonic Response Mitigation

Part Number: LDC1000-Q1

I noticed in "SNOA931 - LDC1612/LDC1614 Linear Position Sensing" that a racetrack design is suggested for linear position sensing. Figure 6 states that past the innermost turn (highest field density) the response becomes non-monotonic. The suggested design uses a two layer PCB with 28 turns per layer. Am I understanding correctly that if the design was 14 turns per layer on four layers, or even seven turns per layer on eight layers that the usable sensor length would be increased (non-monotonic point moved closer to 100 mm)? How would reducing the number of turns per layer affect the output codes?

I have a design that currently uses two LDC1000-Q1 parts. I am already investigating the LDC1001 as a valid drop-in replacement. If there is a compelling reason to migrate beyond the LDC1001 capabilities then I am open to the idea. At this point I am focusing solely on an optimized coil design.

  • Hello,

    If you reduce the number of turns per layer, you will reduce the sensing resolution, because the same sensing distance will need to be covered with half as many turns. Any change in the number of turns per layer or the number of layers will change the coil's inductance, which will result in an offset in the output codes. If fewer turns per layer are used, the output codes will likely have a shallower slope because the response to the target will be weaker. 

    The LDC1612 is a good choice for this application (over the LDC1000 or LDC1001). It is our highest resolution inductive sensing device, it is lower power than the LDC1000 (both supply voltage and supply current), it supports a 5x higher reference clock than the LDC100x (which significantly increases the resolution), and it is also automotive Q-100 qualified.

    Best Regards,

  • Kristin,

    A reduction in the total inductance can be tolerated to a point with my current design. Moving the non-monotonic point is my highest priority. I am not able to use the ratio of two coils in part of the operating region to cancel out many factors such as temperature, application tolerances, etc. These require addition control schemes to correct, and they are proving to be quite complex. 

    My assumption is that reducing the turns will necessarily shift the non-monotonic point. Is this correct? 

    What is the minimum trace width and spacing that TI recommends?

    Are there services available to assist me to optimizing a coil design for my application?

    Thanks,

    Joseph

  • Kristen,

    I replied a couple of days ago, but just realized that the tag "TI thinks this issue is resolved" was on the first post. I wanted to check in to see if you saw my reply and additional questions. Thanks for your help so far.

    -Joseph

  • Hi Joseph,

    My apologies for the delayed reply. Yes, you are correct that reducing the number of turns will shift the non-monotonic point somewhat. If you can tolerate a reduction in resolution, this may be acceptable for your application.

    We recommend 4mil trace width/spacing for the coil, though we recommend wider traces for the connection between the coil, the capacitor, and INxA/INxB.

    Unfortunately we don't have any tools for optimizing stretched coil design. Our tools are designed for traditional spiral coils. We also don't have any contractors we can recommend to assist with the design.

    Best Regards,

  • Kristen,

    No worries. I appreciate your help so far.

    I am creating a set of designs that utilize the LDC161x line to see if there is some way to work around the non-monotonic response, either by two, more efficient coils, or four coils in some way to be able to safely ignore the non-monotonic regions and still be able to utilize a ratio between at least two channels.

    I would like to get your opinion on a variation of the racetrack design presented in TI's literature. I attempted to place the highest field density point as I understand it to the the edge of the PCB. What problems might you expect from a design such as below? 

    Thanks,

    Joseph

  • Hi Joseph,

    This coil design will likely be very low inductance, because the current flow and the magnetic fields from every other turn will oppose each other. The inductance may be too low for the LDC to effectively drive. I only recommend using a spiral coil. 

    Best Regards,

  • Kristen,

    That makes perfect sense. Thank you for all of your help. 

    Thanks,

    Joseph