04-Nov-2019

Attending

  • @Kristofer PISTER

  • @Titan Yuan

  • @Jove Yuan

  • @Meera Lester

  • @Alex Moreno

  • @Lydia Lee

  • @Thomas Watteyne [remote]

  • @Tengfei Chang [remote]

Action Items

@Tengfei Chang complete the frequency sweep test from OpenMote to SCuM on other channels

@Tengfei Chang design frequency scan algorithms

Minutes

  • @Titan Yuan

    • SCuM BLE TX/RX

    • There are still some random bit errors when checking for 32-bit target value from BLE RX

    • However, interrupts (STARTVAL_ISR and RAWCHIPS_32_ISR) are being triggered correctly (verified by turning the Analog Discovery on and off)

    • If I flip an error in the input, interrupts (STARTVAL_ISR and RAWCHIPS_32_ISR) are no longer triggered, so bit errors might be a hardware/timing issue independent of whether the 32-bit target value has been seen or not

    • Note: after enabling STARTVAL_ISR, it always triggers once after optical calibration even if 32-bit target value has not been seen

  • @Jove Yuan

    • Update on the programmer

    • Calculated the power delivered to optical receiver of each LED

    • Tested optical receiver with another Teensy logging OPTICAL_DATA from SCuM, Teensy (the one used for logging) reads wrong values

    • @Kristofer PISTER Just use the logic analyzer instead of using Teensy to test

    • @Thomas Watteyne Recommended logic analyzer saleae

  • @Tengfei Chang and @Thomas Watteyne

    • update on frequency sweep experiments

      • OpenMote picks up alternate channel transmissions, explains some earlier results

      • non-linearity of coarse.mid.fine calibratino explains earlier results

      • new results on OpenMote->SCuM calibration

    • all code submitted as PRs, together with tests to run tests on Jenkins at each future PR

    • next week:

      • understand whether we can infer RX settings from TX settings reliably

      • understand whether we can infer “freq+N” settings from “freq“ settings

      • design quickscan algorithm

      • come up with design for cold-boot calibration, options:

        • “Cal box“ with 16x (or less) OpenMotes which ACK

        • “Cal box“ with some OpenMotes which send beacons periodically (fast)

        • just listen to a 6TiSCH network

  • @Meera Lester

    • ADC rate-spot testing

  • @Alex Moreno

    • update on SCuM and zappyv2

    • show the result of VDDSCuM I vs V under 2 suns

    • video shows SCuM is working with solar powered