• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 3 answers
  • Subscribers 30 subscribers
  • Views 270 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.

Original question:

FDC2114: how does the FDC2114 measures

FDC2114: resonance frequency measurement

Marc Friedheim
Prodigy 110 points
Part Number: FDC2114
Other Parts Discussed in Thread: FDC2214,

Hello, 

I've connected the FDC2214 in the following way, with the measurement capacitor in parallel and tested various caps measuring the resonance frequency with an oscilloscope between GND and IN0A or IN0B. I'm also able to read out the chip through I2C over a range from 70pF to about 200nF.

So far the measurements have been succesful but only up to capacitors around 8nF. Attached capacitors from 80nF and above only show disturbance on the oscilloscope measurement.

I'm using a RTB2002 scope with probes with a capacitance of 12pF and 10MOhm.

Its a bit confusing to me that it does work for the lower caps/higher frequencies, but not for the higher caps/lower frequencies. 

Can someone explain this to me?

  • 0
    TI__Mastermind 35685 points

    Marc,

    Are you using our EVM or your own PCB?

    If it is our EVM, are you using it as-is, or have you substituted any components besides the capacitors?

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 110 points

    John, I'm using a own board with except the processor similar hardware to the EVM. L~1mH and C~18pF

  • 0 in reply to Marc Friedheim
    TI__Mastermind 35685 points

    Marc,

    What is meant by "...capacitors from 80nF and above only show disturbance on the oscilloscope measurement."?

    What is the goal for your nominal sensor frequency w/o a nearby target?

    Are you able to measure the impedance behavior of your inductor over frequency?
    This could be important because the inductor could be self-resonant over the frequencies of interest.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 110 points

    John,

    I mean I can see the sweep of the resonance circuit up to about 8nF, see image below. But for large capacitors like the 80nF and above I'm not able to measure anything with the scope. However through the bus using the microchip I can readout its frequency at 80nF and above up to about 200nF. However I want to know why I can't measure it using my scope...

    Which measurement device would you recommand to measure impedance behavior over frequency? 
    Also giving the fact the software can measure it, do you think this is the problem?

    Regards, Marc

  • +1 in reply to Marc Friedheim
    TI__Mastermind 35685 points

    Marc,

    One technique we use is to put a leaded 1k resistor between the o-scope probe tip and the test point.
    It helps reduce the probe's parasitic C loading on the test point.

    The fact that you can't measure with a larger C that should give an fsensor > 10kHz  sounds suspicious.
    One of the sensor's pins should look like the waveform you showed, while the other pin should look the same, except it is time-shifted by a half-cycle.
    If you can measure the two sensor pins differentially it should look like a full sinusoid centered at 0V.
    If you don't see a waveform like that, its possible the device reporting is unreliable.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 110 points

    Thanks John, adding this resistor solved the issue I had. 

  • 0 in reply to Marc Friedheim
    TI__Mastermind 35685 points

    Great Marc!

    Thank you for letting me know.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 110 points

    Hi John, 

    Thank you for your help!

    I do have a related question. In the data-sheet I see for maximum capacitance (250nF) a 1mH inductor is recommanded to go to a 10kHz resonant frequency.
    For our application we would require 10kHz or preferably 5kHz resonant frequency at lower capacitances (e.g. 8nF). Would this be possible somehow?

  • 0 in reply to Marc Friedheim
    TI__Mastermind 35685 points

    Marc,

    It sounds like you will need to use larger inductors to achieve the requirements.

    Please be mindful of the self-resonant behavior of larger-valued inductors.
    For frequencies greater than the inductor's self-resonant frequency, it will act like a capacitor.
    The result is the intended capacitor/inductor pair will not resonate.

    The rule-of-thumb is the LC resonant frequency be least 25% less than the inductor's self-resonant frequency.

    We don't have any data to offer one way or the other on the 5kHz sensor frequency.

    Regards,
    John

  • 0 in reply to John Miller - Sensing
    Prodigy 110 points

    Hi John,

    Thanks for the reply. This helps me.

    If I understand you correctly, in case we increase the 1mH inductor to a 2mH inductor and keep the 18nF caps we should make sure:

    the frequency of the LC-oscillator f=2πLC should not exceed the self resonance frequency of the inductor f = 1 / (2π × √(L × C)).

    Is there any other or further hardware limitation from the chips side? I mean max inductor size e.g.? Or is it just the capacitance limit of ~250nF?

  • 0 in reply to Marc Friedheim
    TI__Mastermind 35685 points

    Marc,

    The frequency of the LC oscillator should be less than 75% - 80% of the inductor's self-resonant frequency.

    We don't mention anything about the LC sensor's equivalent parallel resistance (Rp)  in the FDC2114 data sheet, but that may be a consideration if you are using larger inductors.

    The reason: The FDC2xx series is very similar to our family of inductive sensors.
    The FDC211x devices are similar to the LDC131x inductive sensors, and the FDC221x devices are very much like our LDC161x devices.
    The LDC data sheets suggest limits for Rp values, mostly because the inductors for those parts are based coils implemented with PCB traces.
    The Rp spec places limits on how narrow you can make the coil's PCB traces for a given coil.
    The sensor Rp range is 1k to 100k, with a typical absolute minimum of 250.
    An LC with a low Rp will be lossier (e.g. lower Q) than an LC with a larger Rp (higher Q).

    This was not included in the FDC2xxx data sheets because the sensors for those devices are usually larger copper PCB traces or solid 2-dimensional shapes, where narrow traces (with higher losses and correspondingly lower LC sensor Rp's) - are unlikely.

    But with larger discrete inductors with a lot of windings, and/or a high permeability core, a lossy LC tank with lower Rp may be something to consider. 

    regards,
    John