36.3.3.1 LDS - Load Data from Data Space Using Direct Addressing

The LDS instruction is used to load data from the system bus into the PHY layer shift register for serial readout. The LDS instruction is based on direct addressing, and the address must be given as an operand to the instruction for the data transfer to start. The maximum supported size for the address and data is 32 bits. The LDS instruction supports repeated memory access when combined with the REPEAT instruction.

After issuing the LDS instruction, the number of desired address bytes, as indicated by the Size A field followed by the output data size selected by the Size B field, must be transmitted. The output data is issued after the specified Guard Time (GT). When combined with the REPEAT instruction, the address must be sent in for each iteration of the repeat, meaning after each time the output data sampling is done. There is no automatic address increment when using REPEAT with LDS, as it uses a direct addressing protocol.

When the instruction is decoded and the address byte(s) are received as dictated by the decoded instruction, the DL layer will synchronize all required information to the ACC layer. This will handle the bus request and synchronize data buffered from the bus back to the DL layer, which will create a synchronization delay that must be taken into consideration upon receiving the data from the UPDI.