Configuring SINAMICS G120C Telegrams on S7-1200 for Fault Codes

Problem Statement: Standard Telegrams Do Not Carry Fault Codes

A SINAMICS G120C PN connected to a SIMATIC S7-1200 (CPU 1215C) is typically operated using one of the Siemens standard telegrams. Telegram 1 carries speed setpoint and actual value plus the standard control/status words. Telegram 350 adds a torque limit word and the actual current/torque feedback, which is attractive when a torque cap must be enforced from the PLC. Neither telegram, however, includes the drive's fault number (r0945) or warning number (r2122) in the cyclic process image. The PLC sees the bits in the status word that indicate "fault active" or "warning active," but it cannot read the actual numeric fault code without additional configuration.

This article documents three viable approaches to surface fault and warning codes from a G120C on an S7-1200 in TIA Portal V15.1 Update 3 with Startdrive V15.1 Update 2:

1. Keep the standard telegram and add a free telegram (telegram 999) on slot 1 containing the relevant r-parameters mapped to PZDs.
2. Use supplementary telegram data on the same slot to extend a standard telegram with additional words that the user tailors.
3. Read fault and warning records acyclically with WRREC and RDREC against the PROFIdrive parameter channel.

The first two options push the values into the cyclic I/O image and update every bus cycle. The third option is decoupled from the bus cycle and reads on demand, which is the technique used by the Siemens "S7-1200 with SINAMICS G120" application example (entry ID 70155469).

Telegram Selection Matrix for SINAMICS G120C PN

The following telegrams are selectable for the G120C PN. Only the words that matter for a typical speed-controlled application with torque limit are shown. PZD widths are fixed at 16 bits unless otherwise noted.

None of telegrams 1, 2, 3, 20, 350, or 352 contain r0945, r0947, r2122, or r2124. Telegram 999 is the only built-in way to get those values into the cyclic frame without changing the firmware or using an acyclic service.

Prerequisites

• SIMATIC S7-1200 CPU 1215C (DC/DC/DC or AC/DC/RLY) with PROFINET interface, firmware V4.4 or later (recommended for compatibility with TIA V15.1 UPD 3).
• Three SINAMICS G120C PN drives, each with a valid PROFINET device name and IP address assigned by the topology editor.
• SINAMICS G120C firmware V4.7 SP10 or later to match the Startdrive V15.1 UPD 2 device library.
• TIA Portal V15.1 Update 3, Startdrive V15.1 Update 2, Openness V15.1 installed on the engineering station.
• GSD file for the G120C PN imported if the project was migrated from an older TIA version.
• Functional PROFINET network with the PLC and all three drives reachable from the engineering station.

Step 1 — Configure Telegram 1 for Speed Control

Telegram 1 is the lowest-complexity cyclic configuration and is the recommended base for any application that does not require encoder-based position control. To assign it:

1. In the project tree, expand SINAMICS_G120C_PN_1 > Device configuration and open the Properties inspector.
2. Select the PROFINET interface → Telegram configuration.
3. In Add telegram, choose Standard telegram 1.
4. Map the resulting I/O addresses into the PLC's process image (for example, %IW64 for ZSW1/NIST and %QW64 for STW1/NSOLL).
5. Create a global DB, for example DB_Drive1_Cyclic, with the structure STW1 : WORD; NSOLL : INT; ZSW1 : WORD; NIST : INT; and assign it to the cyclic I/O using the PLC variable mapping table.

The STW1 control bits 0 (ON/OFF1), 1 (OFF2), 2 (OFF3), and 3 (enable pulse) must be set in the order required by the drive's ramp-up sequence. Bit 6 must be toggled to acknowledge faults. The ZSW1 status word reports drive state in bits 0–2, fault in bit 3, warning in bit 7, and speed deviation in bit 8.

Step 2 — Add Torque Limiting with Telegram 350

When the application must cap motor torque (for example, to protect a gearbox or to perform a soft start against a known load), telegram 350 is the correct choice. It adds the torque limit word M_LIM on the output side and the actual current (IA_IST) plus actual torque (M_IST) on the input side.

1. Replace the telegram assignment from Step 1 with Standard telegram 350.
2. Update the cyclic DB layout to:
   STW1 : WORD; NSOLL : INT; M_LIM : INT; SupplSetp : INT; (output)
   ZSW1 : WORD; NIST : INT; IA_IST : INT; M_IST : INT; SupplAct : INT; (input)
3. Scale M_LIM as a signed 16-bit value where 16384 (0x4000) represents 100 percent of the rated motor torque (parameter p2003 reference torque).
4. Scale NSOLL as a signed 16-bit value where 16384 (0x4000) represents the reference speed (parameter p2000).

Step 3 — Extend the Telegram with Free Telegram 999

Free telegram 999 lets you map any read-only r-parameter to a PZD position in the cyclic frame. Telegram 999 can be used either as a standalone replacement for the standard telegram or as an additional slot attached to the drive's PROFINET interface.

For a configuration that keeps standard telegram 1 (speed control) and adds fault codes, add a second telegram slot with telegram 999 and map the following parameters to the PZDs in this order:

Configuring the mapping in Startdrive:

1. In the device view, open Telegram configuration on the PROFINET interface.
2. Add a second telegram and select Free telegram (999).
3. For each PZD slot, click Insert signal and select the parameter number (e.g., 945) and the array index 0.
4. Set the signal type to Input (drive → PLC) for the fault/warning values.
5. Compile the project and download the drive configuration. The cyclic I/O addresses of slot 2 will appear in the device view next to the standard telegram.

Free telegram 999 is also useful when the application needs additional process values such as r0021 (smoothed actual speed), r0031 (actual torque), or r0080 (torque-generating current Iq) without burning an extra acyclic service.

Step 4 — Use Supplementary Telegram Data (Option B)

Where Startdrive permits it, the standard telegram can be extended with supplementary data that appends additional PZDs directly to the same telegram rather than creating a second slot. The exact mechanics differ slightly between firmware versions, but the effect is the same: a single AR carries both the standard telegram and the user-defined words.

To add fault/warning numbers using supplementary data:

1. Open the telegram configuration and switch from "Fixed telegram" to "User-defined / extended".
2. After the standard telegram words (STW1/NSOLL), append two extra output words and four extra input words.
3. Bind the appended input words to parameters r0945[0], r0947[0], r2122[0], and r2124[0].
4. Reserve the appended output words for future writes (e.g., p0840 as ON/OFF1) if needed.

This approach has the advantage of a single PROFINET AR, a single configuration in the topology editor, and one set of I/O addresses for the PLC. The disadvantage is loss of strict conformance with the PROFIdrive profile; some third-party controllers may not accept the extended layout.

Step 5 — Acyclic Parameter Access with WRREC and RDREC

If you cannot or do not want to consume cyclic bandwidth for fault and warning codes, the acyclic route is the right choice. The drive exposes its parameter list over PROFINET record read/write services. The S7-1200 provides two system function blocks for this: RDREC (read record) and WRREC (write record). The Siemens application example cited above wraps these in an FB (FB20) that orchestrates the request/response handshake.

For PROFIdrive parameter access, the data record index is typically 0xB02E (decimal 45102) on the drive's PROFINET slot. The request payload follows the PROFIdrive parameter channel specification:

The response on a successful read contains: byte 0 = function echo (0x01), byte 1 = number of parameters echoed, byte 2–3 = format/type of the value, byte 4–5 = number of values returned, byte 6–7 = parameter value (U16) for r0945.

Sample ST code to issue a read for the current fault number on drive 1:

// DB_Acyclic parameter request record for r0945, single value
DATA_BLOCK DB_G120C_ReadFault
  STRUCT
    Request  : ARRAY[0..15] OF BYTE := 8(#0);
    Response : ARRAY[0..31] OF BYTE := 32(#0);
    State    : INT  := 0;
    Done     : BOOL := FALSE;
    Busy     : BOOL := FALSE;
    Error    : BOOL := FALSE;
    Status   : WORD := 16#0000;
  END_STRUCT;
END_DATA_BLOCK

// OB1 cyclic call
// Pre-fill request:
DB_G120C_ReadFault.Request[0] := 16#01;    // Function: read parameter
DB_G120C_ReadFault.Request[1] := 16#01;    // Number of parameters
DB_G120C_ReadFault.Request[2] := 16#10;    // Attribute: value
DB_G120C_ReadFault.Request[3] := 16#01;    // Number of elements
DB_G120C_ReadFault.Request[4] := 16#B1;    // Parameter number low byte
DB_G120C_ReadFault.Request[5] := 16#03;    // Parameter number high byte (= 945)
DB_G120C_ReadFault.Request[6] := 16#00;    // Subindex 0
DB_G120C_ReadFault.Request[7] := 16#00;    // Reserved

// Trigger RDREC
RDREC_DB(REQ := TRUE,
         ID  := "G120C_1_Interface".HwId,    // HW ID of PROFINET slot
         INDEX := 16#B02E,                  // PROFIdrive parameter channel
         MLEN := 32,                        // Max response length
         VALID => DB_G120C_ReadFault.Done,
         BUSY  => DB_G120C_ReadFault.Busy,
         ERROR => DB_G120C_ReadFault.Error,
         STATUS => DB_G120C_ReadFault.Status,
         RECORD := DB_G120C_ReadFault.Response);

The current fault number (drive-side index 0) appears at byte 6–7 of the response. To read the full buffer, change byte 3 to the desired count (e.g., 8) and increase MLEN accordingly; each fault occupies 8 bytes (U16 number, U16 reserved, REAL time received, REAL time resolved).

Step 6 — Reading the Fault Buffer (r0945, r0947, r0949)

SINAMICS G120C exposes the fault buffer as a circular array of 8 entries. After a fault occurs, the drive moves the new entry to index 0 and shifts older entries up. The relevant parameters are:

Reading the full buffer with one acyclic call is not possible because each parameter has a different data type and width. A practical strategy is to issue one read per index (0..7) using RDREC, with each call returning one fault number (2 bytes) plus its time received (4 bytes). On the PLC, store the values in a ring buffer DB keyed by drive number and slot index.

A complete index sweep from index 0 to index 7 fits in 8 sequential RDREC calls. Each call can be 200 ms apart on a stable PROFINET network, which keeps the bus load negligible. On a 1 ms PROFINET update time, the RDREC completes typically within 10–30 ms.

Step 7 — Reading the Warning Buffer (r2122, r2124)

Warnings follow the same circular scheme but live in a separate array. The parameter map is:

Note that the warning buffer is larger (64 entries) than the fault buffer (8 entries). When sweeping via RDREC, plan accordingly. A common simplification is to read only index 0 plus the count parameter r2109; if the count is non-zero, escalate the polling frequency or read the additional indices.

Step 8 — PROFINET Diagnostics Channel and FB20 Reference

Beyond the parameter channel, PROFINET offers a diagnostics channel that emits asynchronous alarm telegrams when the drive enters a fault state. The relevant alarm indices on a SINAMICS G120C are summarized below:

Siemens ships a freely-available example project that demonstrates the diagnostics channel in combination with WRREC/RDREC for fault and warning reads. The example is published under entry ID 70155469. Inside the example, FB20 wraps the parameter channel handshake and exposes typed outputs such as FaultNumber, FaultTime, WarningNumber, and WarningTime for each of the three drives.

A complementary reference that shows the speed-control path on the S7-300/400 platform (still valid as a logic pattern for the parameter channel and telegram layout) is the application document at entry 58820849 (PDF).

Step 9 — Commissioning Procedure

1. Download the PLC hardware configuration with all three G120C PN devices in the topology. Confirm that the PROFINET device names resolve from the engineering station.
2. In Startdrive, open each drive's online portal and verify the firmware version with parameter r0018. Accept if the version is V4.7 SP10 or later.
3. Restore the drive factory settings if commissioning for the first time (parameter p0010 = 30, then p0970 = 1). Wait for the drive to complete the ramp-down; the process can take up to 30 seconds.
4. Run basic commissioning: enter motor data from the nameplate into p0300..p0311, set p0010 = 0 to exit commissioning, then p0971 = 1 to save.
5. Configure the telegram (standard 1 or 350) on the PROFINET slot. Download the drive configuration from Startdrive.
6. If using telegram 999 or supplementary data, add the parameter mappings and download again.
7. Compile and download the PLC program. Place the CPU in Run.
8. Send a single ON command (set STW1 = 0x047E, then 0x047F) and verify that the motor turns and that NIST follows NSOLL.
9. Trigger the acyclic reads one drive at a time and confirm that RDREC returns VALID = TRUE and STATUS = 16#0000.

Step 10 — Verification Checklist

Troubleshooting Matrix

Performance and Bandwidth Considerations

Each free telegram 999 PZD consumes 2 bytes of cyclic payload. The standard PROFINET frame on the S7-1215C has a maximum of 256 bytes input plus 256 bytes output per device, so a 5-word free telegram leaves ample headroom. The cyclic update time can stay at the default 1 ms without bus saturation even with all three drives each carrying a 5-word telegram 999.

Acyclic reads do not consume cyclic bandwidth at all but are rate-limited to roughly 50–100 parameter requests per second per drive on a G120C. For fault and warning sweeps of 8 entries, the worst-case latency is 8 × 20 ms ≈ 160 ms. This is acceptable for HMI display but too slow for fast interlock logic. For interlocks, prefer the cyclic status word bit 3 (fault present) and bit 7 (warning present), and use the acyclic read only to populate the fault/warning number for the operator.

Safety and Operational Notes

Frequently Asked Questions