DS14928 · DatasheetArm® Cortex®-M33 32-bit MCU with FPU, 144 MHz, 593 CoreMark®, up to 512‑Kbyte dual‑bank flash memory, 128-Kbyte RAM
Download .pdfOpen in the full document

Peripheral interconnect matrix

Peripheral interconnect matrix

Several peripherals have direct connections between them. These connections allow autonomous communication between them and support the saving of CPU resources (thus power supply consumption). In addition, these hardware connections allow fast and predictable latency.

Depending on the peripherals, these interconnections can operate in Run and Sleep modes.

Reset and clock controller (RCC)

The clock controller distributes the clocks coming from the different oscillators to the core and to the peripherals. It also manages the clock gating for low‑power modes and ensures the clock robustness.

It features:
  • Clock prescaler: in order to get the best trade‑off between speed and current consumption, the clock frequency to the CPU and peripherals can be adjusted by a programmable prescaler.
  • Clock security system: clock sources can be changed safely on the fly in Run mode through a configuration register.
  • Clock management: in order to reduce the power consumption, the clock controller can stop the clock to the core, individual peripherals, or memory.
  • System clock source: four different clock sources can be used to drive the master clock SYSCLK:
    • 4 to 50 MHz high-speed external crystal or ceramic resonator (HSE). The HSE can also be configured in bypass mode for an external clock.
    • 144 MHz or 48 MHz on high-speed internal RC oscillator (HSI), trimmable by software
    • 144 MHz programmable-speed internal oscillator (PSI)
  • Auxiliary clock source: two ultra‑low‑power clock sources that can be used to drive the real-time clock:
    • 32.768 kHz low-speed external crystal (LSE), supporting four drive capability modes. The LSE can also be configured in bypass mode for an external clock.
    • 32 kHz low-speed internal RC (LSI), also used to drive the independent watchdog.
  • Peripheral clock sources: several peripherals have their own independent clock whatever the system clock. Two dividers, each with a large scale of configurable division factors, can generate independent clocks for the ADCs, USARTx, UARTx, SPIx, I2Cx, I3C1, and FDCAN (available only in STM32C552x devices).
  • Startup clock: after reset, the microcontroller restarts by default with an internal 144 MHz clock (HSI). The prescaler ratio and clock source can be changed by the application program as soon as the code execution starts.
  • Clock security system (CSS): this feature can be enabled by software. If an HSE clock failure occurs, the master clock automatically switches to HSI and a software interrupt is generated if enabled. LSE failure can also be detected and generates an interrupt.
  • Clock-out capability:
    • MCO (microcontroller clock output): it outputs one of the internal clocks for external use by the application.
    • LSCO (low-speed clock output): it outputs LSI or LSE in all low-power modes.

Several prescalers allow AHB and APB frequencies configuration. The maximum frequency of the AHB and the APB clock domains is 144 MHz.

Clock recovery system (CRS)

The devices embed a special block that allows automatic trimming of the internal 144 MHz oscillator to guarantee its optimal accuracy over the whole device‑operational range. This automatic trimming is based on the external synchronization signal. This signal is either derived from USB_SOF signalization, from an LSE oscillator, from an external signal on the CRS_SYNC pin or generated by user software. For faster lock-in during startup, automatic‑trimming and manual‑trimming action can be combined.

General-purpose inputs/outputs (GPIOs)

Each of the GPIO pins can be configured by software as output (push‑pull or open‑drain), as input (with or without pull‑up or pull‑down) or as peripheral alternate function. Most of the GPIO pins are shared with digital or analog alternate functions.

After reset, all GPIOs are in Analog mode to reduce power consumption.

The I/Os alternate function configuration can be locked if needed following a specific sequence in order to avoid spurious writing to the I/Os registers.

Multi-AHB bus matrix

A 32-bit multi-AHB bus matrix interconnects all the masters (CPU, LPDMA1, LPDMA2) and the slave peripherals (flash memory, SRAMs, AHB, and APB). It also ensures a seamless and efficient operation even when several high-speed peripherals work simultaneously.

Low-power direct memory access controller (LPDMA)

The low-power direct memory access (LPDMA) controller is a bus master and system peripheral. The LPDMA is used to perform programmable data transfers between memory‑mapped peripherals and/or memories via linked‑lists, under the control of an off‑loaded CPU.

The LPDMA main features are:

  • Single bidirectional AHB master
  • Memory-mapped data transfers from a source to a destination:
    • Peripheral to memory
    • Memory to peripheral
    • Memory to memory
    • Peripheral to peripheral
  • Transfers arbitration based on a 4-grade programmed priority at the channel level:
    • One high-priority traffic class for time-sensitive channels (queue 3)
    • Three low-priority traffic classes with a weighted round‑robin allocation for non‑time‑sensitive channels (queues 0, 1, 2)
  • Per channel event generation on any of the following events: transfer complete, half transfer complete, data transfer error, user setting error, link transfer error, completed suspension, and trigger overrun
  • 12 concurrent LPDMA channels (8‑channel LPDMA1, 4‑channel LPDMA2):
    • Intrachannel LPDMA transfers chaining via programmable linked‑list into memory, supporting two execution modes: run‑to‑completion and link step mode
    • Intrachannel and interchannel LPDMA transfers chaining via programmable LPDMA input triggers connection to LPDMA task completion events
  • Per linked-list item within a channel:
    • Separately programmed source and destination transfers
    • Programmable data handling between source and destination: byte‑based padding or truncation, sign extension, and left/right realignment
    • Programmable number of data bytes to be transferred from the source, defining the block level
    • Linear source and destination addressing: either fixed or contiguously incremented addressing, programmed at a block level, between successive single transfers
    • Programmable LPDMA request and trigger selection
    • Programmable LPDMA half-transfer and transfer-complete events generation
    • Pointer to the next linked-list item and its data structure in memory, with automatic update of the LPDMA linked-list control registers
  • Debug:
    • Channel suspend and resume support
    • Channel status reporting and event flags
  • Privileged/unprivileged support:
    • Support for privileged and unprivileged LPDMA transfers, independently at the channel level
    • Privileged-aware AHB slave port
Table 2. LPDMA1 channels implementation and usage
Channel x Hardware parametersFeatures
dma_fifo_size[x]dma_addressing[x]
x = 0 to 700
Channel x (x = 0 to 7) is implemented with:
  • no FIFO. Only a single source transfer cell is internally registered.
  • fixed/contiguously incremented addressing
Table 3. LPDMA2 channels implementation and usage
Channel x Hardware parametersFeatures
dma_fifo_size[x]dma_addressing[x]
x = 0 to 300
Channel x (x = 0 to 3) is implemented with:
  • no FIFO. Only a single source transfer cell is internally registered.
  • fixed/contiguously incremented addressing
Table 4. LPDMA1 and LPDMA2 autonomous mode and wake-up in low-power modes
FeatureLow-power modes
Wake-upLPDMA1/2 in Sleep mode

Nested vectored interrupt controller (NVIC)

  • 81 maskable interrupt channels (not including the 16 Cortex®‑M33 with FPU interrupt lines)
  • 16 programmable priority levels (4 bits of interrupt priority used)
  • Low-latency exception and interrupt handling
  • Power management control
  • Implementation of system control registers

The NVIC and the processor core interface are closely coupled, enabling low‑latency interrupt processing and efficient processing of late‑arriving interrupts. All interrupts, including the core exceptions, are managed by the NVIC.

Extended interrupt and event controller (EXTI)

The extended interrupts and event controller (EXTI) manages the individual CPU and system wake‑up through configurable and direct event inputs. It provides wake‑up requests to the power control and generates an interrupt request to the CPU NVIC and events to the CPU event input. For the CPU, an additional event generation block (EVG) is needed to generate the CPU event signal.

The EXTI wake-up requests allow the system to wake up from Stop modes.

The interrupt request and event request generation can also be used in Run modes.

The EXTI also includes the EXTI mux I/O port selection.

The EXTI main features are the following:

  • 35 input events are supported.
  • All event inputs allow the possibility to wake up the system.
  • Events that do not have an associated wake‑up flag in the peripheral have a flag in the EXTI and generate an interrupt to the CPU from the EXTI.
  • Events can be used to generate a CPU wake‑up event.

The asynchronous event inputs are classified into two groups:

  • Configurable events (signals from I/Os or peripherals able to generate a pulse), with the following features:
    • Selectable active trigger edge
    • Interrupt pending status register bits independent for the rising and falling edge
    • Individual interrupt and event generation mask, used for conditioning the CPU wake‑up, interrupt, and event generation
    • Software trigger possibility
    • EXTI I/O port selection
  • Direct events (interrupt and wake-up sources from peripherals having an associated flag which requires to be cleared in the peripheral), with the following features:
    • Fixed rising edge active trigger
    • No interrupt pending status register bit in the EXTI (the interrupt pending status flag is provided by the peripheral generating the event)
    • Individual interrupt and event generation mask, used to condition the CPU wake‑up and event generation
    • No software trigger possibility