37.4.2.6 Stop Condition Timing

A Stop condition occurs when SDA transitions from an Active state to an Idle state while SCL is Idle. Host hardware will issue a Stop condition when it has completed its current transmission and is ready to release control of the bus. A Stop condition is also issued after an error condition occurs, such as a bus time-out or when a NACK condition is detected on the bus. User software may also generate a Stop condition by setting the Stop (P) bit.
After the ACK/NACK sequence of the final byte of the transmitted/received packet, hardware pulls SCL low for half of an SCL period (TSCL/2) (see the figure below). After the half SCL period, hardware releases SCL and then samples SCL to ensure it is in an Idle state (SCL = 1). Host hardware then waits the duration of the Stop condition setup time (TSU:STO) and releases SDA, setting the Stop Condition Interrupt Flag (PCIF). If the Stop Condition Interrupt Enable (PCIE) bit is also set, the generic I2CxIF is also set.
Important: At least one SCL low period must appear before a Stop condition is valid. If the SDA line transitions low and then high again while SCL is high, the Stop condition is ignored, and a Start condition will be detected by the receiver.
Figure 37-37. Stop Condition Timing
Important:
  1. At least one SCL low period must appear before a Stop is valid.
  2. See the device data sheet electrical specifications for Stop condition setup and hold times.