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.

How to improve resolution of LDC1000 for measuring distance of cupper plane

Other Parts Discussed in Thread: LDC1000EVM, LDC1612, LDC1041

Hello everybody,

in the last few weeks I have been working with the LDC1000 EVM. I want to measure a distance between the planar coil printed on a PCB and an object (for first tests I used a cupper plane on both sides of a PCB).

My printed coil has a outer diameter of 16mm, 23 turns, width of traces and spacing are 0.15mm (6mil).

It's double-layer printed, with a inductance of about 8.85uH.

I want to start with the inductance measurement. For that purpose I use the frequency counter raw data.

Between the minimum and maximum distance (which are around 0.2mm (minimum) and 1mm (maximum)) the resolution of the frequency counter is around 60 units.

How can I improve the resolution to, e.g. 3-5 um? Should I use a coil with a smaller diameter?


Thanks in advance for any advice


Best regards

Stephan

  • Hello Stephan,

    Your steps look correct. What is the size of the target comparatiely to the size of the coil?

    Best Regards,
    Natallia Holubeva
  • Hi Natallia,

    the target's size is almost the same as the size of the coil.

  • Hello Stephan,

    Your design seems good. To improve the resolution, you could try decreasing the frequency of the sensor (by soldering a different cap for example). If you decrease the frequency, the % change you get by moving 3um will increase, so you should see a larger amount of codes changing.

    Best Regards,
    Natallia Holubeva
  • Hello Natallia,
    somewhere I read/saw that the higher the sensor frequency the higher the resolution. But with your advice I did improve the resolution the sensor. So the lower the frequency the higher the change in freqency by a movement of 3µm.

    Q1: So my next step is to increase the value of the coil and to find a suitable capacitor so that the sensor frequency is as close as possible to the 5kHz?
    Q2: How important is a acurate setting of the register "Sensor Frequency"? I noticed that the EVM-Board didn't start measuring at a sensor frequency below 100kHz.

    For my process I set Rp_max and Rp_min to the same value (e.g. 0.798kOhm) to deactivate the Rp measurement and to avoid artefacts in the L-measurement.
    Q3: Does this have any impact on my L-measurement?

    Best regards,
    Stephan
  • Hello Stephan,

    The register "Sensor Frequency" sets the watchdog timeout for the LDC1000. Physically, the sensor frequency is set by the values of L and C, and settling the above register tell the chip how to find out when the tank is no longer oscillating. Are you using the LDC1000EVM? Yes, setting Rp min to Rp max helps you avoiding artefacts in the L-measurement. Please note that the selection of Cfb (feedback capacitor) is extremely important.

    You might be interested in the following part: http://www.ti.com/product/LDC1612/description. Will be slower due to I2C, but there are things you can do with register settings to improve the resolution.

    Best Regards,

    Natallia Holubeva

  • Stephan,

    I would like to give you an example that will hopefully explain how the resolution increases with the decrease of frequency.

    LDC1000 has a reference clock that does not change. In order to detect the sensor frequency, we count the amount of clock cycles during a certain period of time. Please consider it as an example (numbers used may not be accurate). If you are operating at 5MHz, the number of clock cycles is 20. Thus, the resolution you can detect is 1/20 = 0.05. If you change the frequency to 5kHz and use the same reference clock and the same time interval, the number of clock cycles will be 20000. Therefore, the resolution you can detect is 1/20000 = 0.00005. This is how you improve the resolution. Please note you need to satisfy Rs constrain to keep Quality Factor high enough.

    Best Regards,
    Natallia Holubeva
  • Hi Natallia,

    thank you for your explanation. Your example was a great help.

    The new device LDC161x looks more suitable for my purpose, I was thinking about using more than one LDC1041. But now I will wait until the new EVM-boards are available to order.

  • Hello Stephan,

    I am always glad to help. Yes, LDC161x looks like a more suitable device. You could contact your distributor about the timeline, but I am hoping you will be able to order the EVM at the very end of this year or the beginning of the new year.

    Best Regards,
    Natallia Holubeva