Modbus data in a device is addressed as follows:

Note that data addressing (at least in decimal and hex formats) is zero-based, where the first instance of a data item, for example coil 1, is addressed as item number 0. As another example, holding register 108 is addressed as 107 decimal or 006B hex.

However, it is common for a vendor to list a device’s data items using a 5-digit Modbus address, for example, holding registers starting with 40001, as shown for a meter in Table 1.

In the Table 1 example, each table row represents a 16-bit holding register. The holding register with Modbus address 40011 (phase B power), can be alternately addressed as a holding register (with an “implied 4nnnn value”) having a decimal address of “10” (“0010”), or a hex address of “A” (“000A”).

Integer-stored data often requires additional math operations. NiagaraAX Modbus proxy points provide this capability “built-in,” via the proxy extension’s Conversion slot. For example, selecting “Linear” in Conversion provides entries for scale and offset, to process the raw Modbus value to a finished value. Typically, the vendor ‘s documentation for a Modbus device includes any “scaling” values needed for any such data items.

Within any particular data group (coils, inputs, input registers, holding registers), it is typical for a Modbus device to use consecutive addresses, particularly for related data. For example, in the Table 1 example, holding registers 40001-40015 are used consecutively for integer data, where each is a separate, integer, “data point”.

In this example device, register 40259 begins a consecutive series of holding registers used to access “floating point” data values—note, however, that an address “gap” exists in this case. The address gap (while not necessary), was probably implemented by the device vendor for clarity. Note also that floating-point data values (being 32-bit based) require the use of two consecutive registers for each data point. See Numerical data types for related information.

Modbus messaging supports device queries for data using both a starting address and range, which is dependent on data items being consecutively addressed. This allows for message efficiency when retrieving multiple data points, as it can be handled in one message response.

The address range for data in any data group (coils, inputs, input registers, holding registers) received in a query must be implemented by the receiving device—otherwise, it will generate an exception response. For example, a read request of holding registers 40003—40015 to the device represented by Table 1 receives a normal response (data values), while a similar request to registers 40003-40017 results in an “illegal data address” response (as holding registers 40016 and 17 are not implemented). See Exception Responses for more information.

A NiagaraAX Modbus integration makes use of consecutively addressed data in two different ways:

Generally speaking, when configuring a Niagara Modbus point for a data address, you should choose “Modbus” addressing from the available “Address Format” types (Figure 3). This lets you enter target data addresses directly from the device’s documentation, without having to “subtract 40001” for example, or perform other mental math.

Figure 3. Modbus address format recommended when adding proxy points

Also, for read-only client points, using the Modbus address format also frees you from setting the register type property (“Reg Type”), as it automatically sets this property going by the leading numeral of the full Modbus address (3 for input registers, and 4 for holding registers), as shown in Figure 4.

Figure 4. Reg Type property of Modbus

When entering a Modbus formatted address for a coil, “leading zeros” are ignored—for example the Modbus address “00109” is the same as entering Modbus “109”. Again, note that unlike Decimal or Hex address formats (“zero-based formats”), the Modbus address format is “one-based”, meaning that a coil addressed as Modbus 109 has a “Decimal address” of 108, and a “Hex address” of 6D.