What would be the suggested logic for RIO communication on my HI 3000 series unit?

What would be the suggested logic for communication using discrete and block transfer? If I use discrete, what data and in what order can I access it? If it is by block transfers what would be the configuration and length of them so as the order of the data showed in the data table in which I will read?

NOTE:  RIO is now obsolete in the HI 3000 series units.  If you already have RIO, it will continue to work, however, you can no longer purchase the RIO option in the HI 3000 series units.

This answer applies to the HI 3000 series units. These units have a baud rate, rack number, starting quarter number. They also have a setting for the rack size. This allows the user to determine if these units will be a 1/4, 1/2, 3/4, or Full rack. Rack size will have a determination in the starting quarter. A full rack must start at quarter 1, 3/4 rack can start at quarter 1 or 2, 1/2 rack at quarter 1, 2, or 3, 1/4 rack can start at any quarter. The rack size will also determine the size of the I/O (discrete) table. NOTE: BYTE 0 OF THE I/O TABLE IS RESERVED AND CANNOT BE USED. 1/4 rack is 3 bytes (bytes 1 to 3) of discrete I/O, 1/2 rack is 7 bytes (bytes 1 to 7), 3/4 rack is 11 bytes (bytes 1 to 11), Full rack is 15 bytes (bytes 1 to 15). All sizes have a block transfer of up to 24 words.

The RIO tables in the HI 3010, HI 3030 and the HI 3300 are 64 words long. The first byte (byte 0) is reserved. Starting at byte 1 is the discrete I/O table for the rack size configured. The block transfer area in the RIO tables starts at the end of the discrete data area. Ex: for 1/4 rack device, the discrete I/O would be bytes 1, 2, 3, and the block transfer area would start at byte 4 (word 2); if the unit was configured as a full rack, the discrete data area would be bytes 1 to 15 and the block transfer area would start at byte 16 (word 8). The remainder of the RIO table in the unit can used as scratch pad area.

The data in each of the sections of the RIO table would be user mapped data. This can be any data that can be mapped. The only limitation of the data would be the size of the RIO table being used.

Example:
An HI 3030 is configured as a 1/2 rack device and wanted the net and gross weight from 2 scales. As a 1/2 rack device, we would have 7 bytes available as discrete output. We can then map the net weight from scale one into the RIO output table word 1 and the net weight from scale two into the RIO output table word 2. (These would be mapped as short integer values as float values (4 bytes each) would not fit into the output table in 7 bytes.) This would give us the net weights in the discrete input table of the plc/slc in words 1 and 2.

We could map the gross weight for scale 1 as a float value into the HI 3030 RIO table word 4 and the gross weight for scale 2 as float value into RIO table word 6. Now when we do a block read we would receive the gross weights in the first 4 words of the block transfer data..