meta data for this page
This is an old revision of the document!
lamaPLC: Simatic and Modbus
Introduction
Certainly, I am aware that numerous descriptions of Modbus can be found online and in technical literature. As the oldest and most widely used industrial communication method, it serves as the backbone of industrial connectivity. While newer, more sophisticated communication protocols have emerged, Modbus remains prevalent. In fact, you might even encounter it on the first intergalactic spacecraft.
Although this communication method is widespread and often underestimated, it can lead to unexpected issues during commissioning, usually more negatively than positively. With over 25 years of experience in automation programming, primarily with Simatic systems, I'm sharing my observations. While the following is somewhat subjective, I hope many readers will have an „aha” or facepalm moment, helping them resolve certain problems.
Modbus Fundamentals
Origin and basics of Modbus
Modbus originated in 1979 and was created by Modicon (now part of Schneider Electric). During this period, industrial automation moved from relay-based systems to digital logic. As the pioneer of the first Programmable Logic Controller (PLC) a decade earlier, Modicon developed Modbus to facilitate communication among these controllers and with external devices via serial lines. The protocol features a straightforward query-response model, in which a “master” (client) initiates communication with one or more “slaves” (servers) to transfer data.
The protocol's emergence as a worldwide industry standard was fueled by several key factors:
- Open and Royalty-Free: Since its inception, Modicon has made the protocol available as an open standard, enabling any manufacturer to implement it without licensing costs.
- Technical Simplicity: Its minimal processing requirements and simple message format facilitated its adoption by hundreds of vendors for applications ranging from sensors to motor controllers.
- Adaptability: Initially designed for serial interfaces such as RS-232 and RS-485, the protocol has evolved to meet industry needs. In 1999, Modbus TCP was introduced, allowing the original protocol to operate over modern Ethernet and TCP/IP networks.
In 2004, Schneider Electric officially transferred the rights to the Modbus Organization, an independent nonprofit that continues to manage and promote it as a public domain standard. Today, it is often called the “grandfather of industrial networking” due to its continued widespread use in both legacy factories and modern IoT systems.
Core application areas of Modbus
Industrial Automation & Manufacturing
- Control Systems: Connecting Programmable Logic Controllers (PLCs) with sensors, actuators, inverters, and motors to automate assembly lines.
- Data Acquisition: Using SCADA (Supervisory Control and Data Acquisition) systems to monitor real-time production data, such as oven temperatures, vibration levels, and pressure.
- Legacy Integration: Retrofitting older machines to communicate with modern control systems through Modbus-to-IoT gateways.
Smart Buildings & Facility Management
- HVAC Control: Managing heating, ventilation, and air conditioning systems based on occupancy and environmental conditions.
- BMS Integration: Centralizing data from lighting, security, and elevator systems for improved energy efficiency.
- Smart Metering: Connecting Modbus-enabled smart meters to monitor electricity, water, and gas usage across residential or commercial complexes.
Energy Management & Renewables
- Solar & Wind: Monitoring Photovoltaic (PV) inverters, trackers, and batteries to optimize energy generation and storage.
- Electric Vehicles (EV): Integrating charging infrastructure with building energy systems to manage load and prevent grid strain.
- Smart Grids: Enabling real-time communication between grid management systems and remote sensors at substations.
Water & Wastewater Management
- Process Monitoring: Automating chemical dosing units, monitoring pump station status (pressure, flow rate), and checking water quality (pH, conductivity).
- Infrastructure Safety: Detecting sudden pressure changes to identify pipeline leaks or bursts immediately.
Modbus Client and Server
In the Modbus protocol, the terms Client and Server specify the roles of devices during communication. These terms are the current, official replacements for the older “Master/Slave” terminology.
Modbus Client (formerly Master)
The Client is the active device that initiates all communication transactions. It's sending questions.
- Action: It sends a “Request” to a specific device to read or write data to it.
- Behavior: It should wait for a response or a timeout before issuing the next command.
- Typical Devices: SCADA systems, HMI panels, or a primary PLC.
Modbus Server (formerly Slave)
A server is a passive device that responds to client requests.
- Action: It waits for an incoming message, processes the request (for example, looks up a sensor value), and sends back a “Response”.
- Behavior: It never initiates a conversation; it only responds when spoken to.
- Typical Devices: Sensors such as temperature and humidity sensors, motor drives like VFDs, power meters, or I/O blocks.
Modbus RTU and TCP, and Hybrid
The core difference is the transport layer: Modbus RTU is designed for a physical wire (Serial), while Modbus TCP is designed for a network (Ethernet). Before Ethernet's advent, only the RTU (Remote Terminal Unit) protocol was available, primarily using RS-232 or RS-485. The rise of Ethernet greatly expanded options, enabling communication over the faster, more versatile TCP/IP protocol. Today, these two methods serve as the Modbus transport layer.
Although quite different, they can sometimes be integrated, such as in multimaster (hybrid) RTU scenarios. In this scenario, a signal converter needs to be integrated into the TCP network. The converter has its own IP address to receive Modbus/TCP telegrams. It's crucial that the client is aware of the Modbus/RTU network behind the converter, which uses traditional RTU addressing. Often, clients cannot manage both RTU addressing and TCP simultaneously, so verifying their ability to do so is essential. The Siemens TIA Portal is suitable for this task, and I will offer an example of its use later.
| Feature | Modbus RTU | Modbus TCP |
|---|---|---|
| Media | Serial (RS-485/232) | Ethernet / Wi-Fi |
| Error Check | CRC (at the end) | TCP/IP Checksum (built-in) |
| Topology | Daisy-chain | Star (Switch-based) |
| Speed | Typically 9600 or 115200 baud | 10/100/1000 Mbps |
| Master/Client | Only one Master | Multi-Master |
| Port | Serial COM Port | TCP Port 502 |
| Segment distance | 1200 meters | 100 meters between switches |
| Addressing | Master ID: 0, slave ID: 1 to 247 | IP Address (like: 192.168.178.123) |
| Wiring | 3-core cable with shielding or 4-core cable | Ethernet cables (RJ45) and network switches |
Modbus/RTU
Comparison of RS-232 and RS-485
The Modbus/RTU transmission options depend on the physical layer: RS-232 or RS-485. RS-232 is quite uncommon because it only supports point-to-point connections, meaning one Client and one Server. In contrast, RS-485 is a more flexible option; its technical specifications are outlined in the table below.
| RS-232 | RS-485 | |
|---|---|---|
| Operating mode | asynchronous transmission | synchronous transmission |
| Number of drives and receivers per line | 1 drive 1 receiver (point-to-point) | 32 stations per segment |
| Data transfer method | half-duplex, full-duplex | half duplex |
| Data transmission | p2p | multipoint |
| Max. cable length | 15 m | 1200 m |
| Max. data transfer 12 m 1200 m | 20 kbps (1 kbps) | 35 Mbps 100 kbps |
| Receiver input resistance | 3..7 kΩ | ≧ 12 kΩ |
| Drive Load- Impedance | 3..7 kΩ | 54 Ω |
| Receiver “dead band” | ±3 V | ±200 mV |
| Receiver voltage level | ±15 V | -7..+12 V |
| Drive output voltage max. | ±25 V | –9..+14 V |
| Drive output voltage min. (with load) | ±5 V | ±1.5 V |
| Drive output short circuit current limit | 500 mA to Vdc or Ground | 150 mA to Ground 250 mA to Vdc |
| Receiver Hysteresis | 1.15 V | 50 mV |
Modbus/RTU wiring
For Modbus/RTU communication, use RS-485 with either a 3-wire with shield or a 4-wire cable. To improve noise immunity, using shielded cables and twisted pairs is recommended. The units should be connected in a daisy-chain layout, though star topology can also work in some cases with a few units.
In a Modbus RTU network over RS-485, it is vital to terminate the bus to prevent signal reflections that can cause communication errors and CRC failures.
Guidelines on Proper Terminator Placement
- Both Ends Only: Install a termination resistor at the first and last device of the trunk cable.
- Do Not Terminate Middle Devices: Adding resistors at intermediate nodes can overload drivers and cause signal loss.
- Daisy Chain Topology: RS-485 should be wired in a continuous line. Avoid star or tree configurations with long branches or stubs, as these cannot be correctly terminated.
Resistor Specifications
- Value: Usually 120 Ω to match the characteristic impedance of standard twisted-pair cable.
- Power Rating: Select a resistor with a power rating of at least 0.25W to 0.5W.
- Connection: Attach the resistor directly between the A (D0/-) and B (D1/+) data lines.
Modbus/RTU RS-485 Signaling
In a Modbus/RTU network (which uses the RS-485 physical layer), A and B represent the two wires of a differential pair used to transmit data.
- Differential Signaling: The receiver measures the voltage difference between two lines instead of against a common ground. This helps the signal resist electrical noise because interference usually impacts both wires equally.
- Half-Duplex: Modbus/RTU uses these two wires for transmission and reception, but only one device can communicate at a time.
The Naming Confusion (A vs. B)
Since there is no universal naming standard, wiring errors are common. Terminal labels vary by manufacturer as follows:
| Labeling Convention | Non-Inverting Signal | Inverting Signal |
|---|---|---|
| Common/Modbus | B or B+ | A or A- |
| TIA/EIA-485 Standard | A (Negative) | B (Positive) |
| Alternate Labels | D+, Data+, Tx+ | D-, Data-, Tx- |
- Standard Rule: In the official TIA-485 standard, A is negative (A-) and B is positive (B+).
- Industry Practice: Many Modbus device manufacturers (such as those following Modbus Organization guidelines) use A for negative (-) and B for positive (+).
Voltage Levels
- Logic 1 (Idle/Marking): Occurs when the voltage at B is higher than A by at least 200mV.
- Logic 0 (Active/Spacing): Occurs when the voltage at A is higher than B by at least 200mV.
- Idle State: When no device is transmitting, the bus is “idle.” Biasing resistors are often used to keep the B line slightly more positive than the A line, preventing noise from being misinterpreted as data.
If your devices have the correct baud rate and address but still can't communicate, try swapping the A and B wires on one side. This is a common cause of Modbus RTU setup failure and won't harm your hardware.
Modbus/RTU Speed (Baud Rate)
This is the transmission speed in bits per second (bps).
- Standard speeds are typically 9600 and 19200 bps. For quicker data updates over short distances, higher speeds such as 38400, 57600, or 115200 bps are often used.
- Lower baud rates, such as 9600, provide greater stability over lengthy cable runs—up to 1200 meters—because they are less affected by signal reflection and noise.
- Timing Requirement: Modbus RTU depends on silent intervals (3.5 character times) to mark the end of a frame. If the baud rate is incorrect, devices cannot “see” where one message ends and the next begins.
- Manufacturer specification: The primary focus should be on the speed setting. If not specified, it is recommended to start at 9600 baud and 8N1.
Modbus/RTU Data Format ("8N1")
Data Format 8N1:
This describes the structure of a single byte (character) sent over the wire. Totaling 10 bits per character:
- 8 (Data Bits): Modbus RTU consistently employs 8 bits per byte, enabling direct transmission of binary data, whereas Modbus ASCII converts data to text.
- N (Parity): No parity bit is employed for error detection at the byte level. Note: While the Modbus standard technically specifies Even Parity as the default, the industry predominantly uses No Parity (8N1).
- 1 (Stop Bit): A single stop bit indicates the conclusion of the byte.
- Constraint: If “No Parity” is selected, the standard recommends using 2 stop bits (8N2) to keep the character length at 11 bits, but most modern devices default to 8N1 regardless.
Modbus/RTU Error Checking (CRC)
Even though “No Parity” (8N1) lacks bit-level checking, Modbus RTU is still secure because every full packet ends with a 16-bit CRC (Cyclic Redundancy Check). If a single bit is flipped during transmission due to noise, the CRC will fail, and the receiving device will ignore the command.
The CRC (Cyclic Redundancy Check) is an error-detection method that ensures data integrity in Modbus RTU. It is a 16-bit (2-byte) value appended to each message.
Key Characteristics
- Algorithm: Modbus uses the CRC-16-ANSI (also known as CRC-16-IBM) polynomial: X16+X15+X2+1
- Verification: The sender computes the CRC and adds it to the frame. When the receiver gets the frame, it recalculates the CRC; if the values differ, the receiver discards the packet and does not reply.
- Efficiency: It is far more dependable than a basic “Checksum,” as it can identify all single and double-bit errors and the majority of burst errors.
Modbus/TCP
Modbus/TCP (also known as Modbus TCP/IP) is a version of the Modbus protocol designed for Ethernet network communication. It wraps standard Modbus messages into TCP/IP packets, enabling reliable, high-speed data exchange.
Key Characteristics
- Architecture: Uses a Client-Server model (previously called Master-Slave). The Client sends requests, and the Server replies with data or action confirmations. Network Port: Default is TCP Port 502.
- Reliability: Depends on the TCP/IP stack for error detection and reliable delivery, removing the necessity of CRC (Cyclic Redundancy Check) used in serial Modbus RTU.
- Addressing: Uses IP addresses to identify devices instead of the 1-byte Slave ID used in serial versions.
Modbus TCP Frame Structure
The Modbus TCP frame (Application Data Unit or ADU) consists of a specialized header (MBAP) followed by the standard Modbus PDU.
| Field | Length | Description |
|---|---|---|
| Transaction ID | 2 Bytes | A request number used to match responses with requests. |
| Protocol ID | 2 Bytes | Always 0 for Modbus services. |
| Length | 2 Bytes | Count of remaining bytes (Unit ID + PDU). |
| Unit ID | 1 Byte | Used for routing through gateways to serial Modbus devices; otherwise often set to 255 or 0. |
| Function Code | 1 Byte | Defines the operation (e.g., Read Holding Registers). |
| Data Field | Variable | Contains the actual payload (register addresses, values, etc.). |
Modbus Registers and Coins
Modbus data is structured into four main “data banks” or “storage units”, categorized by whether the data is a single bit or a 16-bit word, and whether it is read-only or read-write.
Coils and Discrete Inputs are 1-bit variables. While a Coil can be read and written, a Discrete Input is only a read-only indicator of status.
Registers are 16-bit storage units: Holding registers are both writable and readable, and Input registers are read-only.
Although the Modbus standard allows Holding registers to be both read and written, many manufacturers restrict them to read-only. Always check the manufacturer’s specifications for accurate details.
Key Characteristics
- 16-Bit Architecture: All Modbus registers are 16 bits (2 bytes) in size. To handle larger data types such as 32-bit floating points or integers, systems usually allocate two registers in sequence.
- Big-Endian Format: Modbus typically transmits the most significant byte (MSB) first.
- Zero-Based Addressing: Although documentation often refers to addresses using “1-based” numbering (e.g., Holding Register 40001), the actual transmission uses 0-based addresses (e.g., Address 0).
- Manufacturer Flexibility: Devices are not required to support all four register types. Many modern devices simplify data handling by mapping all information, including status bits, into Holding Registers.
- Read/Write Permissions: “Input” types (Discrete Inputs and Input Registers) are strictly for data sent from field devices to the controller and cannot be modified by a Modbus master.
Modbus Coin and Register Addressing
To index Modbus address ranges, a 5-digit address (e.g., 40001) was initially used. Over time, this was insufficient, as it allowed only 9,999 addresses per type. This was expanded to 6-digit addresses (e.g., 400001), offering 65,536 addresses aligned with Word boundaries. The first digit indicates the area type: 0x for Coils, 1x for Discrete Inputs, 3x for Input Registers, and 4x for Holding Registers.
Addressing within each range begins at 1, so the first holding register is 40001 or 400001. Users should note that many manufacturers use hexadecimal addresses, while Modbus/RTU uses decimal addresses. Larger data types (>16 bit, such as REAL, LREAL, DT, or STRING, WSTRING) span multiple registers for a single variable; thus, both the start address (offset, e.g., 400012) and the area length are specified.
