Use GPTs in real-time capable applications

Azure Sphere supports using general purpose timers (GPTs) in real-time capable applications.

Each real-time core on the MT3620 supports five GPTs. Timers GPT0, GPT1, and GPT3 are interrupt-based. These timers count down from an initial value and assert an interrupt when the count reaches 0. Timers GPT2 and GPT4 are free-running timers. These timers count up from an initial value.

Two modes are defined for interrupt-based timers:

• One-shot mode: The times stops when it counts down to 0.
• Auto-repeat mode: The timer restarts after it counts down to 0.

GPT control registers for each real-time core are mapped to addresses 0x2103_0000 through 0x2103_FFFF in the core's private memory.

For more information about using the real-time core GPTs in applications, see Azure Sphere MT3620 M4 API Reference Manual.

Note

The Azure Sphere OS does not reset peripherals on start-up. Your applications should ensure that peripherals are in a known-good state on start-up or after restart.

The GPT sample applications

The following sample applications demonstrate how to use GPTs on MT3620 real-time cores:

• The GPT_RTApp_MT3620_BareMetal sample uses GPT1 and GPT3.
• The MT3620_RTApp_FreeRTOS_GPT sample uses GPT0 and GPT3.