4 Replies Latest reply on Jul 18, 2019 3:12 AM by LinglingG_46

    PSoC5LP CapSense "Individual difference of RAW count value"

    MaMi_1205306

      Hi,

       

      We are considering products using CapSense of PSoC5LP.

      Therefore, We created 20 prototype boards..


      In order to tuning, RAW count initial value (not touched) was confirmed for all substrates.

      It was found that the raw count values differ if the substrates are different even with the same sensor.

       

      For example, 13bit resolution,

      Board #1, Sensor 1 : 7500 RAW count,

      Board #2, Sensor 1 : 7300 RAW count (Min value),

      Board #3, Sensor 1 : 7400 RAW count,

      Board #4, Sensor 1 : 7700 RAW count (Max value),

      and so on.

       

      Do you know the factors that cause the initial value of RAW count to differ depending on the board?

       

      (a) Cmod capacitor accuracy
      (b) Variation of sensor capacity per board
      (c) PSoC CapSense component Accuracy

      I think the above factors can be considered.
      Which one is particularly influential?

       

      In this case, which RAW count should be used for tuning?

      Which max value of RAW count or min value of Raw cont or average of all?

       

      Regards,

        • 1. Re: PSoC5LP CapSense "Individual difference of RAW count value"
          LinglingG_46

          As my understanding, " (b) Variation of sensor capacity per board " is the most likely factor.

           

          I think "7500,7300,7400,7700" these parameters are all two high for the CapSense system. It is recommend the customer to set the raw counts 80% of the full scale. More information you can refer to the blow linker file Page 67:https://www.cypress.com/file/45441/download

          In addition,  you need to add the sensor overlay to do the test.

           

           

          • 2. Re: PSoC5LP CapSense "Individual difference of RAW count value"
            MaMi_1205306

            Hi,

             

            Thank you for your reply.

             

            You say that "RAW count value is too high",

            That RAW count value (about 7300 to 7700) is the value when the sensor is not pressed.


            However the image below shows the RAW count value when the sensor is not pressed is around 6300 in materials you sent to me.

            Please tell us the reason why RAW count value is 6300 is good but 7500 is bad.

             

             


            And I think that "set the raw counts 80% of the full scale" is the value when press the sensor.

            Is that recognition correct?

             

            Regards,

             

            • 3. Re: PSoC5LP CapSense "Individual difference of RAW count value"
              MaMi_1205306

              Please answer the question.

              The concept of tuning parameters for the accuracy (different initial raw values) of sensor capacitance by the board.
              In other words, does it mean that Cp(min) to Cp (max) also needs to satisfy "set the raw counts 80% of the full scale"?

               

              • 4. Re: PSoC5LP CapSense "Individual difference of RAW count value"
                LinglingG_46

                Hi,

                 

                Please tell us the reason why RAW count value is 6300 is good but 7500 is bad.

                 

                As my understanding, if the raw count value is 7500, it means that occupies 7500/8096=92% of the full scale.1: That's easy to spill over.

                2: The rawcount  performance worse in higher than 90% (Empirical results).

                3: Our design guide give us the reference values 80% of the full scale.

                 

                In addition, the capsesnse guide documentation gives us the reference that the Cp should be in the range 5pf-45pf in smartsense mode.

                 

                In other words, does it mean that Cp(min) to Cp (max) also needs to satisfy "set the raw counts 80% of the full scale"?

                The Cp is only one of the variables, we should change other parameters to let the count value meet our requirement.

                 

                Hope it can be helpful for you.