Problem Description
The SIMATIC 545-1103 CPU exposes two onboard serial ports on its faceplate: a multi-mode Port 1 and a fixed-function Port 2. Field reports describe a recurring condition in which Port 2 establishes programming sessions with 505-Workshop or TISOFT, but Port 1 times out, returns no carrier detect, or fails to echo characters even when the same null-modem cable is moved between the two connectors. Because the same physical cable and the same PC serial port succeed on Port 2, the failure is internal to the Port 1 hardware path, the DIP-switched personality, or the firmware handling of the Port 1 UART.
This article consolidates the verified Port 1 commissioning procedure from the SIMATIC 545/555/575 System Manual, the legacy 505-Workshop cable reference, and field-replicated diagnostics that isolate the failure to one of three layers: physical layer (cable, connector, signal levels), configuration layer (DIP switch SW2 position, baud rate, parity, echo), or component layer (UART, port driver, optocoupler).
SIMATIC 545-1103 Port Architecture
The 545-1103 faceplate carries two 9-pin D-sub male connectors labelled PORT 1 and PORT 2. Port 2 is hard-wired as an RS-232 DTE programming port that always supports the TISOFT / 505-Workshop protocol stack. Port 1 is multi-mode and is reconfigured at power-up by reading the on-board DIP switch bank. The supported Port 1 personalities are:
- RS-232 programming — full TISOFT / 505-Workshop protocol, DTE pinout.
- RS-422 programming — differential TISOFT protocol for long runs.
- RS-485 multi-drop — peer-to-peer data acquisition.
- Printer / terminal port — ASCII output only, no programming handshake.
| Port | Default Mode | Reconfigurable | Programming Capable | Handshake Lines |
|---|---|---|---|---|
| Port 1 | RS-232 DTE | Yes (DIP SW2) | Yes (when SW2 = RS-232) | RTS, CTS, DCD, DTR, DSR |
| Port 2 | RS-232 DTE | No | Yes | RTS, CTS, DCD, DTR, DSR |
The Port 1 UART and line drivers are physically separate from Port 2, which is why a Port 2 success does not imply Port 1 integrity. The two ports share the CPU's main 16-bit data bus and the +5 V logic supply, but each has its own isolated RS-232 charge-pump and its own DB-9 shell ground. A failure on Port 1's RS-232 transceiver (historically an SP3232 or MAX202-equivalent) will not affect Port 2.
DIP Switch Configuration for Port 1
The 545-1103 carries a 10-position DIP switch accessible through a small cut-out in the front bezel. Switches 1–8 control port personalities, port modes, and CPU options. Switches 9–10 control memory protection and battery back-up behaviour. The exact position of the switch (left = open, right = closed) is read only at power-up; changing the switch while the CPU is running has no effect until the next power cycle or warm restart.
| Switch | Function | Left (Open) | Right (Closed) |
|---|---|---|---|
| SW1 | Port 1 echo | Echo off | Echo on |
| SW2 | Port 1 mode select | RS-232 programming | RS-422 / RS-485 / printer |
| SW3 | Port 1 termination (RS-422/485) | Terminated | Not terminated |
| SW4 | Port 2 echo | Echo off | Echo on |
| SW5 | Reserved | — | — |
| SW6 | Port 1 protocol | TISOFT | Modbus RTU |
| SW7 | Watchdog timer | Enabled | Disabled |
| SW8 | Power-on default mode | Run | Program |
| SW9 | Memory protect | Unprotected | Protected |
| SW10 | Battery backup enable | Enabled | Disabled |
After changing SW2, cycle power on the CPU and wait at least 5 seconds for the boot diagnostics to complete before attempting a programming session. A rapid power-cycle <2 s may not re-read the DIP bank on early 545-1103 firmware revisions.
RS-232 Cable Pinout and Wiring
The Port 1 and Port 2 connectors on the 545-1103 are wired as DTE devices. A null-modem (crossover) cable is required to connect to a PC COM port, which is also DTE. The wiring is identical for both ports when Port 1 is in RS-232 mode. Reference the Siemens RS-232/RS-422/RS-485 communication interface specification for the electrical characteristics of the signalling standard.
| DB-9 Pin | Signal | 545-1103 DTE End | PC DTE End | Direction (PC → PLC) |
|---|---|---|---|---|
| 1 | DCD | Output | Input | PC receives |
| 2 | RD (RX) | Output | Input | PC receives |
| 3 | TD (TX) | Input | Output | PC transmits |
| 4 | DTR | Input | Output | PC transmits |
| 5 | SG (signal ground) | — | — | Common |
| 6 | DSR | Output | Input | PC receives |
| 7 | RTS | Input | Output | PC transmits |
| 8 | CTS | Output | Input | PC receives |
| 9 | RI | Not connected | Output | Unused |
The minimum three-wire null-modem cable connects pin 2 to pin 3 (swap), pin 3 to pin 2 (swap), and pin 5 to pin 5 (straight). For TISOFT to complete the handshake, pins 4, 6, 7, and 8 must also be looped back: 4↔6 and 7↔8 on the PC end. The 505-Workshop programming cable sold by Siemens (part number 505-W1234 or equivalent) implements this loop-back internally and is the reference for any field-built cable.
A common failure is a cable that omits the handshake loop-backs. Such a cable may function with a PC that asserts DTR unconditionally but fails on a USB-to-serial adapter that follows strict DTE/DCE conventions. Always verify with a breakout box or a continuity tester that the four handshake lines (4, 6, 7, 8) are properly crossed at the PC end of the cable.
Baud Rate and Communication Parameters
The 545-1103 auto-detects the baud rate on Port 1 at session start. The supported auto-baud set is:
| Baud Rate | Parity | Data Bits | Stop Bits | Notes |
|---|---|---|---|---|
| 300 | None / Even | 8 | 1 | Lowest reliable, useful for marginal cables |
| 1200 | None / Even | 8 | 1 | Default for 505-Workshop |
| 9600 | None / Even | 8 | 1 | Common in legacy deployments |
| 19200 | None / Even | 8 | 1 | Highest supported on Port 1 |
Parity must be set to EVEN when the CPU is in legacy TISOFT mode (SW6 = LEFT). When SW6 is RIGHT (Modbus RTU mode), parity is forced to NONE with 8-N-1 framing. TISOFT will not establish a session on Port 1 if SW6 is in the Modbus position, even though the RS-232 transceiver is healthy.
Systematic Diagnostic Procedure
Follow this sequence in order. Do not skip steps; each step eliminates one failure layer.
- Power down the 545-1103 rack and remove the CPU module from its slot. Set the module on an ESD-safe mat.
- Inspect the DB-9 shells on both Port 1 and Port 2 with a magnifier. Look for bent pins, broken solder joints at the connector shroud, or contamination in the pin field. A bent pin 2 or pin 3 will kill RS-232 communication without affecting the UART.
- Verify the DIP switch SW2 is in the LEFT (open) position. Verify SW1 is in the position required by 505-Workshop (typically LEFT for echo off in programming mode). Verify SW6 is in the LEFT (TISOFT) position. Photograph the switch bank for the maintenance record.
- Verify the cable on a known-good 545-1104 or 545-1105. The cable that works on Port 2 of the suspect unit MUST also work on Port 1 of a known-good unit when SW2 is LEFT. If it does not, the cable is faulty and the CPU is innocent.
- Re-seat the CPU in its slot and apply power. Wait 5 seconds for boot completion. Confirm the CPU OK LED is steady green and the COMM LED on Port 1 is dark (idle).
-
Launch 505-Workshop on the PC. Configure the COM port for 1200-8-E-1. In the connection dialog, select
Port 1 (RS-232)explicitly. Press F4 (Go Online) twice with a 3-second pause. The first attempt trains the auto-baud detector; the second attempt establishes the session. - If Go Online times out, drop the baud rate to 300-8-E-1 and repeat. If 300 baud also times out, the Port 1 UART is suspect.
-
Check CPU fault log from 505-Workshop after a Port 2 session is established. Look for fault code
F045(Port 1 framing error),F046(Port 1 overrun), orF047(Port 1 hardware fault). These codes do not appear on Port 2 faults. - Swap with a known-good CPU of the same part number to isolate whether the fault follows the CPU or the rack backplane. A backplane with a corroded pin on the Port 1 UART interrupt line will produce identical symptoms to a CPU failure.
Common Failure Modes
| Symptom | Likely Cause | Verification | Resolution |
|---|---|---|---|
| Port 2 OK, Port 1 dead, SW2 LEFT | Blown RS-232 transceiver on Port 1 | Measure +10 V on DB-9 pin 4 or pin 7 with meter | Replace the SP3232/MAX202 IC on the CPU board |
| Port 1 echoes garbage, no handshake | SW6 in Modbus position | Visual check of SW6 | Set SW6 LEFT, cycle power |
| Port 1 works at 300 baud, fails at 19200 | Aging UART clock crystal | Replace CPU; cross-test on spare | Reflow or replace 1.8432 MHz crystal |
| Port 1 fails on one PC, works on another | Cable handshake loop-back missing | Continuity test pins 4↔6, 7↔8 at PC end | Rebuild cable with full loop-back |
| Port 1 times out after 5-second boot | CPU firmware older than rev 4.2 | Read firmware tag under CPU lid | Update firmware via Port 2 session |
| Port 1 OK on bench, fails in rack | Backplane connector oxidation on Port 1 UART interrupt | Swap CPU with neighbour slot | DeoxIT backplane pins, replace backplane |
| Port 1 OK, then intermittent | Cold solder joint at DB-9 shell ground | Flex the connector while online | Reflow the four shell-ground pads |
Field-Proven Cross-Diagnostics
The fastest field confirmation of a healthy Port 1 is the "back-to-back" test:
- Connect the suspect 545-1103 Port 1 to a known-good 545-1104 Port 1 with the same null-modem cable.
- On the known-good CPU, launch 505-Workshop and select
Port 1on the COM port connected to the suspect CPU. - Press F4 twice. If the session establishes, the suspect CPU's Port 1 is alive. If it times out, the suspect CPU's Port 1 is dead.
- Reverse the test: from the suspect CPU, try to go online against the known-good CPU. This catches a one-sided port failure (e.g., a blown TX driver only).
Documented in legacy field records, the "first attempt times out, second attempt connects" pattern on Port 1 is a known behaviour of early 545-1103 firmware revisions when the auto-baud detector is cold. A persistent timeout on both attempts is the genuine failure indicator.
Replacement and Repair Considerations
For a 545-1103 with confirmed dead Port 1, three recovery paths exist:
- Use Port 2 only. Port 2 has identical programming capability. Re-cable the workstation to Port 2 and document the change on the panel drawing. This is the lowest-risk path for production systems.
- Refurbish the CPU module. The Port 1 RS-232 transceiver, the UART clock crystal, and the DB-9 connector are all field-replaceable with through-hole equivalents. Replacement requires a hot-air rework station and a clean bench, but the part cost is under $5 USD. A refurbished module must pass the back-to-back test before redeployment.
- Replace the CPU module. Refurbished 545-1103 modules are available on the secondary market. Verify the seller has performed a back-to-back test on both ports and provides a 30-day warranty. New-old-stock modules are rare and command a premium.
Verification Procedure
After any corrective action, validate Port 1 with the following checklist:
- Power up the rack and confirm the CPU OK LED is steady green within 5 seconds.
- Verify DIP switch SW2 is LEFT, SW6 is LEFT, SW1 is in the position required by the programming tool.
- Connect the 505-Workshop PC to Port 1 with a known-good null-modem cable.
- Launch 505-Workshop, select Port 1 at 1200-8-E-1, press F4 twice.
- Confirm the session establishes and the program directory loads within 10 seconds.
- Read the CPU fault log and confirm no Port 1 fault codes (
F045,F046,F047) are present. - Perform a program read from the CPU to disk and compare the file checksum to the last known good backup.
- Disconnect, wait 30 seconds, reconnect, and confirm the session re-establishes on the first attempt (warm auto-baud test).
Does Port 1 on the 545-1103 support programming sessions or is it terminal-only?
Port 1 supports full TISOFT and 505-Workshop programming sessions when DIP switch SW2 is in the LEFT (open) position, which selects RS-232 DTE mode. When SW2 is RIGHT, Port 1 routes to the RS-422/485 transceiver or the printer driver and cannot be used for programming regardless of cable type.
Can I use the same null-modem cable for Port 1 and Port 2 on the 545-1103?
Yes. Both ports are wired as DTE devices with identical RS-232 pinout. A null-modem cable that works on Port 2 will work on Port 1 when SW2 is LEFT. If the cable works on Port 2 but not Port 1, the failure is on the Port 1 side of the CPU, not in the cable.
What is the default baud rate for 505-Workshop on a 545-1103?
505-Workshop defaults to 1200 baud, 8 data bits, even parity, 1 stop bit (1200-8-E-1). The CPU auto-detects the baud rate on session start. If 19200 fails, drop to 1200, then 300 to isolate baud-mismatch from hardware failure.
Why does my Port 1 time out on the first F4 press but connect on the second?
This is normal behaviour of the 545-1103 auto-baud detector on cold start. The first F4 press trains the detector; the second press establishes the session. If both presses time out, the Port 1 UART, RS-232 transceiver, or DIP switch is suspect.
What fault codes indicate a Port 1 hardware failure?
Fault code F045 indicates a Port 1 framing error (baud/parity mismatch or cable fault), F046 indicates a Port 1 UART overrun (baud rate too high for cable quality), and F047 indicates a Port 1 hardware fault (transceiver or UART failure). These codes appear in the CPU fault log accessible from 505-Workshop over Port 2.