• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 316 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.

FDC2114: how does the FDC2114 measures

Marc Friedheim
Prodigy 110 points
Part Number: FDC2114
Other Parts Discussed in Thread: FDC1004

Hello, I'm trying to use the FDC2114 for a research project and am able to measure various capacitances.

I'm a little puzzled since most of the data sheet images are drawn as single ended configurations (with sensor connected one way to GND)

However this is not the way I can measure appropiate values of capacitors. This only works in differential setup where Cmeas is between InA and InB.

Also this is the only succesfull way I'm able to apply the mentioned formula, which also refers to differential configuration.

 

Question A: what are the settings to use this chip in single ended configuration and which formula can be used to measure?

For research we try to measure and understand the resonant circuit driver in differential mode.

The image I get on the scope when measuring between GND and A on channel 1 and GND and B on channel 2 looks like this:

Question B: I would have expected a complete sinusoidal signal, can someone explain how this circuit measuring exactly.

Hope someone can explain the questions.

  • 0
    TI__Mastermind 35635 points

    Marc,

    As you oscope signal show, the single-ended input signals are half-sinusoids referenced to ground.
    Considered differentially, IN0A - IN0B is a full sinusoid centered about ground.

    The equation for Csensor is valid for calculating the differential sensor capacitance in parallel with the inductor.
    The diagram above the equation for the differential capacitance shows how a single parallel capacitance can be expressed as two virtual single-ended capacitances in series with the common point connected to ground.
    If the total value of the sensor's parallel capacitance is Cp, then the values of the two virtual single-ended capacitances would be 2Cp. 

    This idea can be used to account for a single-ended sensor capacitance.
    In the upper diagram you posted, assume the single-ended sensor capacitance is Cse, and the lumped capacitance Cp can be expressed as two series capacitances of 2Cp with their shared point connected to ground. 
     ------->        

    You can calculate the equivalent capacitance of the three-capacitor network with the usual equation for series capacitance.
    When I do that, I get (no guarantees!) Ctot = (Cp + Cse/2)/(1 + Cse/(4C)).

    You can then substitute the expression in the usual equation for the LC tank resonant frequency, fo = 1/(2*π*√(LCtot)), and solve for Cse.

    regards,
    John

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

    Hi John, thanks for the reply.

    Additionally, can you advise on the cables to use between board and capacitor to measure? Cable length is between 0,5 - 2m. I've tried several cables like twisted, coaxial, and shielded/grounded. The last one was the only solution that performed at all. But probably you have some explanation why twisted/coax didn't work out or additional advices here?

    Thanks.

  • 0 in reply to Marc Friedheim
    TI__Mastermind 35635 points

    Marc,

    As a guess, I'd say the parasitic capacitance of the cables, along with resistance losses, are the main culprits.
    Twisted pairs are infamous for parasitics, and even lower-loss shielded coax cables will have parasitic capacitances, which will interact with the sensor capacitances.

    One potential alternative is the FDC1004. It is a capacitive sensor that doesn't use a resonant circuit and it has active shield drivers that could be helpful for an application that can use shielded coax as an interconnect.
    For your application, you could try connecting the sensor and FDC1004 via shielded coax, but instead of grounding the coax shield, you could connect it to the FDC1004 active shield drive. That would minimize parasitics from the cables because the shield driver is forcing the cable shield to be at the same potential as the cable's signal-carrying conductor, thereby minimizing capacitive losses, and shielding the cable from EMI.

    More collateral on the FDC1004 is available at the E2E FDC1004 Frequently Asked Questions in the form of links to app notes, blogs and reference designs.

    Regards,
    John