• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 32 subscribers
  • Views 915 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.

FDC1004: Designing a sensor

Rory MacKean66
Prodigy 10 points
Part Number: FDC1004

We're using the FDC1004, and have been happy with the performance of the sensor during our prototyping.  Our primary sensors are rectangular, approximately 100mm x 40mm.  They follow the surface of a cylinder with a 200mm diameter, so they need to be able to be curved.  As we look at methods for manufacturing the sensors in production, I'd be curious if there are any guidelines on the designs of the sensors themselves.  

In particular:

-- Does the separation distance between the sensor and the shield have any effect?  We've observed that sensors we make using a traditional PCB (with ~1.6mm separation) perform better than ones we made using flexible PCB process (with ~.1mm separation), but that could be an artifact of our prototyping processes.  (In this case, worse performance is characterized by more "crosstalk", where we see an increase in channel 2 counts when an object is placed in front of the channel 1 sensor)

-- Should the sensor / shield be solid? Or would a hatched pattern be better in some way?

-- Are there guidelines on the material between the sensor and shield?  Obviously we don't want something conductive, but what are good and bad materials?

-- What is the effect of the area of the sensor?  Are there upper constraints we should have in mind for the sensor size?  For example, if we can cover our sensing area with a zig-zag shape to get a 50% reduction in sensor size, does that impact performance one way or the other?

Thanks!

Rory

  • 0
    TI__Genius 11340 points

    Hi Rory,

    1.  Typically, a reduction in distance between the shield and the sensor leads to reduced sensitivity and range but more effective at blocking interference. One reason that could explain crosstalk in your particular case is that because the PCB is flexible, the CH2 could be facing in a way that is easier to pick up objects placed in front of CH1. We have a technical document about active shielding (Capacitive Sensing: Ins and Outs of Active Shielding) that you could check out.

    2. I would recommend the sensor/shield be solid.

    3. We don’t have a guideline, but FR4 is a good material to put in between the sensor and the shield.

    4. The sensitivity and range increase as sensor size increases, but also higher chance for the sensor to be affected by noise. You would need to see that for your specific application, the sensitivity is still acceptable even with a reduction in sensor size.

    Let me know if you have additional questions!

    Best,

    Jiashow 

  • 0 in reply to Jiashow Ho
    Prodigy 10 points

    Jiashow,

    We’re still working on the sensor module Rory described.  The result that most confuses us is this:

    When the sensor to shield capacitance is low, say 200 pF or less, the sensor behaves as expected.  But when the capacitance is high (1800 pF) the shield seems to become ineffective at isolating the sensor.  In this case bringing a 1 cm square grounded patch near the shield causes a significant change in the measured capacitance to be reported .  The spec states that the shield to ground capacitance should be no more than 400 pF (we don’t violate this constraint) but mentions no limit on the shield to sensor capacitance.  

    Our preferred solution is to build a sensor module from flex PCB.  But the sensor/shield separation of only about 0.1 mm leads to high capacitance and non-isolation.  We tested a hand-built sensor of the same size and shape with a 1.0 mm sensor/shield separation.  This gives us lower capacitance and the expected isolation but implementing this would complicate manufacturing.

    Is there a limit to the allowed sensor/shield capacitance or have we fooled ourselves in some way?  All tests were made using the evaluation board.

    Joe

  • 0 in reply to Joseph Jones
    TI__Genius 11340 points

    Hi Joe,

    You shouldn't see any sensor-to-shield capacitance if the shield and sensor are in phase. It is likely that your shield and sensor are out of phase so you see a voltage difference thus capacitance between them. A capacitance of 1800pF would swamp out the sensor even with the CAP DAC connected, which supports up to 100pF.

    Could you double check the phase of your sensor and shield as well as the capacitance reading and CAP DAC setting? It is likely that you are saturating your input.

    Best,
    Jiashow