3 Replies Latest reply on Aug 14, 2018 11:02 AM by kfr

    PSoC4 and PSoC4 BLE IMO accuracy With API-called calibration

    kfr

      In the datasheet for PSoC4 and PSoC4 BLE we spec the IMO accuracy at +/-2%.  In the notes sections it states: "With API-called calibration"

      http://www.cypress.com/file/416486/download Table 47.

       

      1) What does this statement mean?

      2) Does it mean that there is an API required to maintain this accuracy that the user is responsible to call?

      3) Does Creator automatically take care of calling this API?

      4) Does a function need to be called for any scenarios where we would need to re-calibrate? (switching frequencies for example).

       

      I assume that these devices are factory trimmed but the statement seems to imply the user is responsible for calling a function to keep this accuracy in at least some scenarios.

       

      Regards,

      Kevin

        • 1. Re: PSoC4 and PSoC4 BLE IMO accuracy With API-called calibration
          rzzh

          Hi Kevin,

           

          1) What does this statement mean?

          - It means, with API-called calibration, IMO accuracy error could be lower than +/-2%, and depend on trimming clock accuracy, IMO accuracy error could be as lower as possible.

          2) Does it mean that there is an API required to maintain this accuracy that the user is responsible to call?

          - To maintain +/-2% accuracy doesn't require API called by users.

          3) Does Creator automatically take care of calling this API?

          - It depends, I think. For ce97601, it uses PSoC4200. If users would like to get a more accurate IMO to get a lower UART baud-rate error or other more accurate clock requirement situation, they can use API-called to calibrate IMO. There is an AN about how to calibrate IMO. http://www.cypress.com/documentation/code-examples/ce97601-improving-accuracy-psoc-4-internal-main-oscillator

          But for PSoC4100S or PSoC4000S, users can select IMO trim with WCO or not trimmed. If trim with WCO selected, it doesn't require to take care of API called to calibrate 48MHz IMO.

          4) Does a function need to be called for any scenarios where we would need to re-calibrate? (switching frequencies for example).

          - Yes. After IMO changed, it is better to re-calibrate it.

          Kindly correct me if I made mistakes.

           

          Thanks,

          Ryan

          • 2. Re: PSoC4 and PSoC4 BLE IMO accuracy With API-called calibration
            kfr

            Ryan,

             

            1) What does this statement mean?

            - It means, with API-called calibration, IMO accuracy error could be lower than +/-2%, and depend on trimming clock accuracy, IMO accuracy error could be as lower as possible.

             

            If the accuracy could be BETTER with the API called calibration, I don't think this message is clear.  The spec says +/- 2% "with API called calibration" which implies the API needs to be called to get +/- 2%.  If this is not the case, there should be a asterik note on +/-2% and a note that says accuracy can be IMPROVED beyond +/- 2% with API called calibration.

             

            Regards,

            Kevin

            • 3. Re: PSoC4 and PSoC4 BLE IMO accuracy With API-called calibration
              kfr

              Ryan,

               

              It looks like the BLE and module datasheets may be outdated, as most recent PSoC4 datasheets remove the API reference and instead add the note that the IMO changes above 85C

               

              ±3% if TA > 85 °C and IMO frequency < 24 MHz.

               

              So +/-2% is guaranteed across voltage and process up to 85C.  For extended temp devices they can see up to +/-3% in certain conditions. 

               

              http://www.cypress.com/file/139956/download

               

              Regards,

              Kevin

              1 of 1 people found this helpful