SLAAED9 November   2023 TAA5412-Q1 , TAC5311-Q1 , TAC5312-Q1 , TAC5411-Q1 , TAC5412-Q1

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1 Introduction
  5. 2 Diagnostic Monitoring Architecture
  6. 3 Monitored Faults
    1. 3.1 Microphone Faults
      1. 3.1.1 Inputs Shorted to Ground
      2. 3.1.2 Inputs Shorted to MICBIAS
      3. 3.1.3 Input Open Circuit
      4. 3.1.4 Input Pins Shorted Together
      5. 3.1.5 Input Overvoltage Detection
      6. 3.1.6 Inputs Shorted to VBAT
    2. 3.2 Line Out Faults
      1. 3.2.1 Output Overcurrent
      2. 3.2.2 Virtual Ground
    3. 3.3 Other Faults
      1. 3.3.1 MICBIAS Overvoltage
        1. 3.3.1.1 DIAG_CFG11 Register (page = 0x01, address = 0x51) [Reset = 0x40]
      2. 3.3.2 MICBIAS Overcurrent
      3. 3.3.3 MICBIAS Load Current
        1. 3.3.3.1 DIAG_CFG6 Register (page = 0x01, address = 0x4C) [Reset = 0xA2]
        2. 3.3.3.2 DIAG_CFG7 Register
      4. 3.3.4 Overtemperature Fault
      5. 3.3.5 Supply Back Pumping
  7. 4 Enabling Diagnostics and Programming Thresholds
    1. 4.1 DIAG_CFG0 Register (page = 0x01, Address = 0x46) [Reset = 0x00]
    2. 4.2 DIAG_CFG1 Register (page = 0x01, Address = 0x47) [Reset = 0x37]
    3. 4.3 DIAG_CFG2 Register (page = 0x01, Address = 0x48) [Reset = 0x87]
  8. 5 Fault Diagnostic Setup Procedure
  9. 6 Fault Reporting
    1. 6.1 Live Registers
      1. 6.1.1 CHx_LIVE Register (page = 0x01, address = 0x3D) [Reset = 0b]
      2. 6.1.2 CH1_LIVE Register (page = 0x01, address = 0x3E) [Reset = 0h]
      3. 6.1.3 INT_LIVE0 Register (page = 0x01, address = 0x3C) [Reset = 00]
      4. 6.1.4 INT_LIVE1 Register (page = 0x00, address = 0x42) [reset = 0x00]
      5. 6.1.5 INT_LIVE2 Register (page = 0x00, address = 0x43) [reset = 0x00]
    2. 6.2 Latched Registers
      1. 6.2.1 Clearing Latched Registers
    3. 6.3 Fault Filtering and Response Time
      1. 6.3.1 Debounce
      2. 6.3.2 Scan Rate
        1. 6.3.2.1 DIAG_CFG4 Register (page = 0x01, address = 0x4A) [reset = 0xB8]
      3. 6.3.3 Moving Average
        1. 6.3.3.1 DIAG_CFG5 Register (page = 0x01, address = 0x4B) [reset = 0h]
  10. 7 Responding to a Fault
    1. 7.1 INT_CFG Register (page = 0x00, address = 0x42) [reset = 0b]
      1. 7.1.1 DIAG_CFG10 Register (page = 0x01, address = 0x50) [Reset = 0x88]
    2. 7.2 Manual Recovery Sequence
    3. 7.3 Recommended Fault Register Read Sequence
  11. 8 Using PurePath Console
    1. 8.1 Advanced Tab
    2. 8.2 Diagnostics Walk-through
      1. 8.2.1 Diagnostics Configuration
      2. 8.2.2 Debounce Configuration
      3. 8.2.3 Latched Fault Status
  12. 9 Diagnostic Monitoring Registers
    1. 9.1 Voltage Measurements
    2. 9.2 MICBIAS Load Current
    3. 9.3 Internal Die Temperature
  13. 10Summary
  14. 11References

6.3.3 Moving Average

For more robust detection, the fault monitoring can be based on the moving average of the diagnostic signal. With this setting enabled, each new sample is averaged with the previous set of samples and if a fault occurs, the latched registers does not latch until the average of the samples crosses the programmed threshold. This setup acts as a simple FIR filter and avoids triggering faults from transient events in the system. If debounce and moving average are used together, then the latched register does not latch until the moving average of consecutive samples crosses the programmed threshold and remains there for the programmed debounce count. The moving average can be set to weigh new and old data equally, or skewed to give old data 0.75 weight and new data a 0.25 weight to further improve transient immunity. The moving average setting is common to all input channels, but can be enabled or disabled independently for the MICBIAS load current and overtemperature faults to achieve faster response times for these faults. Moving average settings are contained in the DIAG_CFG5 register (Page 1, address 0x4B).