• State TI Thinks Resolved
  • Locked Locked
  • Replies 15 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 1406 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.

FDC2214EVM: Measuer maximal capacitance of 100 pF and resolution better than 50 attofarad

Zohar Mann
Prodigy 70 points
Part Number: FDC2214EVM
Other Parts Discussed in Thread: FDC2214

I am using the TI FDC2214 EVM to measure capacitance. The maximal capacitance at the application is 10 pF and I need a resolution better than 50 attofarad ( 50*10^-18 ) , is this feasible , how should I configure the registers of the on EVM FDC2214 to meet the requirements ?

At current the reading is unstable, the jitter is +/- 0.05 pF ( I need 50 attofarad resolution ) in all four channels even for the on EVM sensors of channels 0 and 1 ( the PCB capacitors ).

I hooked ceramic capacitor of 470 pF to channel 2 and 3 ( to the provided on EVM pads ) and got wrong reading. As I understand it the EVM channels 2 and 3 pads are for differential sensor configuration. For my application I need the high resolution. For this I read it recommended to work in single ended mode. Is there any register should be changed.

When I hover my palm over the EVM the capacitance changes a bit.

  • 0
    TI__Mastermind 40295 points

    Zohar,

    Thank you for your inquiry and your interest in TI products.

    Do you have any requirements for the clock, Internal/external? 

    If external, what are the clock source specs?

  • 0 in reply to John Miller - Sensing
    TI__Mastermind 40295 points

    Zohar,

    Just so you know, many of us supporting these forums will be out for the next two weeks for the holidays.

    If you have any questions or concerns, please go ahead and submit them.

    Someone will be monitoring this forum during that time, but replies may be delayed.

  • 0 in reply to John Miller - Sensing
    Prodigy 70 points

    At current I do not have any constrains on the clock speed and source. The higher capacitance rate the better. At current I only have the evaluation board that is used for feasibility test. I browsed the registers initial values for the EVM and it is set to external clock. 

    Looking forward for your relay 

  • 0 in reply to Zohar Mann
    TI__Mastermind 18545 points

    Hello Zohar,

    Using a quality external clock gives the best stability in the clocking signal which helps get better resolution in the sensor measurement. Additionally, using a higher clock frequency (max 40MHz for the FDC2214) gives a better resolution as well. If you don't have any constraints already, I recommend using an external 40MHz clock that has good stability over temperature and time. The one on the EVM is pretty good if you want to use that one as a reference.

    Best Regards,

  • 0 in reply to Justin Beigel
    Prodigy 70 points

    Hello Justin,

    As I told you at current all my tests are done on the EVM ( will not move forward with design my own board till I convince using the EVM that the project can be implemented based on the FDC2214 ). Is the EVM clock good enough to meet my requirements ? Hovering my hand palm over the my sensor ( even 5 to 10 cm above the sensor ) changes a bit the reading , is that normal? Is the 50 attofarad resolution  feasible ? 

    Best regards

    Zohar

  • 0 in reply to Zohar Mann
    TI__Mastermind 18545 points

    Hello Zohar, 

    Hovering your hand over the sensor and seeing a bit of change is expected when using the EVM. The 50 attofarad resolution is a touch requirement and would be pushing the limits of the part. It would require very careful design and a low noise environment. Even so, I am not sure it is something that can be guaranteed. 

    Best Regards,

  • 0 in reply to Justin Beigel
    Prodigy 70 points

    Hello Justin,

    Is there a design at which hovering my hand will not change the measurement ? ( If I find the FDC2214 suitable for my project I'll design a PCB and electronics according recommendations ) . Given I'll design for low noise environment what is the best measurement  resolution   I may achieve ?Is it in the femtofarad/s resolution? Is 100 attofarad resolution achievable ? 

    As I told you I hooked a 470pF  to the EVM but get wrong measurements can you please help in solving this ?

    Best Regards

    Zohar 

  • 0 in reply to Zohar Mann
    TI__Mastermind 40295 points

    Zohar,

    According to the data sheet, the system noise floor for this device is 0.3fF (at 100sps), which is 6x larger than your 50aF required resolution.
    And this does not consider the additional noise effects due to the clock.

    It might be possible to improve the resolution (e.g. reduce the noise floor) by averaging the data with register settings and some extended averaging in your MCU.

    I am not sure if this will work, but will look at the feasibility of this approach and update the thread by Wednesday of next week.

    You mentioned your max capacitance is 10pF, and your resolution is 50aF.
    What is the minimum capacitance you need to measure?
    Is there a particular sensor excitation frequency you wish to use?

  • 0 in reply to John Miller - Sensing
    Prodigy 70 points

    John,

    Thank you for the prompt reply. I checked again and maximal capacitance is 1 pf  and the resolution I need is 0.1fF  but the lower the resolution the better. The minimal capacitance is  0.1fF.

    As I wrote in my previous email I hooked a ceramic capacitor of 470pF to the EVM but got a wrong reading. Can you please  refer to this ?

    Best regards

    Zohar

  • 0 in reply to Zohar Mann
    TI__Mastermind 18545 points

    Hello Zohar, 

    In regards to the 470pF capacitor, the device is better at determining the change in capacitance on a normal sensor but you might still be able to get a ceramic capacitor to work by adjusting the sensor drive to fix the oscillation amplitude. The sensor oscillation needs to be between 1.2-1.8V for best performance. I did some testing with ceramic capacitors and was able to get better results with smaller ceramic capacitors (in the nF range). 

    Best Regards,

  • 0 in reply to Justin Beigel
    Prodigy 70 points

    Hello Justin

    Did you check the nF ceramic capacitor using the EVM ? If yes can you provide more details regarding EVM configuration and capacitor range ? You mentioned that sensor oscillation needs to be between 1.2 to 1.8 V how do you set this ? Are there any application notes but the data sheet ? At least from my point of view the data sheet is not enough for get all the benefits from  the chip. Is there a better capacitance measurement  chip for my application ?

    Best Regards

    Zohar Mann

    CTO

  • 0 in reply to Zohar Mann
    TI__Mastermind 40295 points

    Zohar,

    Justin has tried some smaller ceramic caps and they do indeed work better.

    You can set the oscillation amplitude via the IDRIVE setting.

    An LDC app note gives some guidance on inductive sensor drive levels that can apply to the FDC part you are using: https://www.ti.com/lit/an/snoa950/snoa950.pdf 

     Please look it over and let me know if you have any questions.

  • 0 in reply to John Miller - Sensing
    Prodigy 70 points

    Hi,

    Can you please tell how to configure the EVM in order to measure ceramic capacitor 

    Thanks

  • 0 in reply to Zohar Mann
    TI__Mastermind 40295 points

    Zohar,

    I will check with Justin to see if he has a record of the EVM settings he used.

  • 0 in reply to John Miller - Sensing
    TI__Mastermind 40295 points

    Zohar,

    Justin did not keep a detailed record of his EVM settings for his experiments.
    We don't have any detailed advice we can offer, apart from the app notes we mentioned earlier.

    Regards,
    John