2 Replies Latest reply on Jun 29, 2010 10:45 AM by robert.seczkowski

    Capsense PRS switch frequency

    robert.seczkowski

      "CapSense® Sigma-Delta Data Sheet" says:

         

      Spread-spectrum,
      average is FIMO/4,
      peak is FIMO/2

         

      What is the center frequency. Is it FIMO?

         

      I don't think so. I bet rather FIMO/4(switching phase alignment). So the freq spec is 0 to FIMO/2. Is it true?

         

      I don't think so. The highest freq range for PRS is a clock speed.(PRS is not clock doubler)

         

      What should then be the right answer?

         

      I think the right answer is below:

         

      Spread-spectrum,
      average is FIMO/4/4,
      peak is FIMO/2/2

         

      Regards

         

      Robert

        • 1. Re: Capsense PRS switch frequency
          kristopher.young

          If you think about a pseudorandom sequence (PRS) signal clocked at FIMO, the highest frequency you can get from it is one FIMO clock cycle high followed by one FIMO clock cycle low (or vice versa). The frequency of this pattern is FIMO/2. You can't get a higher frequency than this from the PRS. You will occasionally see this high frequency pattern from the sequence, however you will also find two, three, or more consecutive FIMO clock cycles where the output stays the same. If you do a statistical analysis on the output transitions, the average frequency is FIMO/4.

             

          To convince yourself even further, check out this page that simulates a linear feedback shift register (LFSR) using Java. You can use 8 or 16 as the value of N (number of register stages). It will give you the pseudorandom bit stream for the output. Look at this stream and think of generating one of these bits per FIMO clock. This will give you a good idea of the frequency content in a PRS.

             

          Kris

          • 2. Re: Capsense PRS switch frequency
            robert.seczkowski

            I meant capsense interference  pattern in terms of capsense Cs and Cmod switching.

               

            Capsense dataseet says, that switching occurs at frequency 4 times less than FIMO.

               

            You've said, that PRS is FIMO/2. So it looks, that PRS frequency is two times higher, than expected by Capsense.

               

            for example if I have local inteference of 1024KHz, should I set PRS to 0.8*(2* 1024) or 0.8*(4*1024).

               

            Regards

               

            Robert