IMDTouchKnowledge BaseShop main store

How-to

Connecting multiple S7-300 PLCs to WinCC stations

Industrial networks Published 8 min read

Short answer

Connect the six S7-300 PLCs and the three WinCC stations over Industrial Ethernet with configured S7 connections; PROFIBUS DP allows only one cyclic master, so it cannot host three independent operator stations. A CP 343-1 Lean gives each PLC 8 S7 connections and the full CP 343-1 up to 16. Keep everything in one STEP 7 project so NetPro configures all connections in one place.

On this page
  1. Can three WinCC stations share one PROFIBUS DP bus to six S7-300 PLCs?
  2. Which network should carry the three WinCC stations instead?
  3. How many S7 connections does each CP 343-1 Lean have left?
  4. Should all six PLCs go into one STEP 7 project?
  5. What steps configure the connections in NetPro and WinCC?
  6. Do the six CPUs stay configured as DP masters?
  7. When one PLC fails, do the other stations fault too?
  8. Does a WinCC station still need a CP5611 or extra software?
  9. Questions asked

Connect the PLC-to-PLC exchanges and all three WinCC stations over Industrial Ethernet with configured S7 connections, and leave PROFIBUS DP on the local slave segments where it already runs. One STEP 7 project holds the six SIMATIC 300 stations plus one SIMATIC PC station per WinCC machine, so NetPro configures every connection in one place and each operator position keeps its own independent path to all six PLCs. The sections below give the per-CP connection budget, the NetPro and WinCC settings, and the failure behavior you should expect before you buy a single card.

Can three WinCC stations share one PROFIBUS DP bus to six S7-300 PLCs? #

No — PROFIBUS DP is a master/slave protocol that allows exactly one cyclic (class 1) master per line, so of your three WinCC stations only one could ever poll the six PLCs and the other two would receive no cyclic data at all.

The PROFIBUS DP specification (IEC 61158, Type 3) assigns each cyclic data exchange to one class 1 master; a second master on the line may run acyclic diagnostics, never the cyclic operator traffic. A PC or panel placed on a DP bus is configured as that master — the documented interface setting is the "Only master on the bus" check — and only one station can hold the role. The workaround that follows, one WinCC server as the sole DP master with two clients hanging off it, ties all three operator positions to a single machine: when that server dies, every seat loses the six PLCs at once. FMS was the peer-to-peer service on PROFIBUS, but it needs a dedicated communication processor; the CPU's integrated DP port does not serve it. That ends DP as the WinCC layer, and it says nothing against the DP segments each CPU already runs for its own remote I/O.

Which network should carry the three WinCC stations instead? #

Industrial Ethernet with S7 connections: fit an Ethernet CP into each S7-300 rack — a CP 343-1 Lean serves 8 S7 connections and the full CP 343-1 up to 16 — and each of the three WinCC stations opens its own connections to all six PLCs.

S7 communication runs connection-oriented over Ethernet, and each WinCC-to-PLC link is single-sided: the PC side is configured and the PLC acts as the S7 server, so no communication blocks go into the PLC programs. Connections are independent of each other, which is why three stations can read the same data block at the same time without any master election. Put the six CPs, one small managed industrial switch and the three PCs on one flat machine subnet in a star. Once the links run, tag update speed is decided by poll sizing and tag layout rather than by this topology — why HMI point updates run slow on an Ethernet PLC link covers that side with its own numbers.

How many S7 connections does each CP 343-1 Lean have left? #

Three of the Lean's 8 go to your three WinCC stations; add the five peer links of a full mesh between the six PLCs and you have 3 + 5 = 8 — exactly the limit, with zero left for engineering or another seat.

Worked example, per PLC, full mesh:

Connection consumer per PLC Connections Running total
Three WinCC stations, single-sided S7 3 3
Peer S7 connections to the other five PLCs (full mesh) 5 3 + 5 = 8
Spare on a CP 343-1 Lean (limit 8) 0 8 of 8 used
Same load on a full CP 343-1 (limit 16) — 8 used, 8 spare

Decision: if every PLC genuinely exchanges data with every other, the Lean runs at its ceiling — take the full CP 343-1 and keep 8 connections in reserve for an engineering station, diagnostics or a fourth operator seat. If only some pairs exchange data — two peer links per PLC, say — the count is 3 + 2 = 5 of 8 with 3 spare, and the Lean is the right card. Trim the mesh to the pairs that actually move data before stepping up the hardware.

Should all six PLCs go into one STEP 7 project? #

Yes — create one project, for example named Total, containing the six SIMATIC 300 stations taken from your existing projects plus one SIMATIC PC station per WinCC machine, so NetPro holds and configures every S7 connection in a single editor.

In SIMATIC Manager you insert the WinCC machines with Insert > Station and the SIMATIC PC Station object; the connection work then happens in NetPro, where Insert > New Connection adds each link. With everything in one project, the connection tables compile and download together and the PLC symbol names transfer directly into WinCC tag management. Six separate projects can still work — a standalone WinCC opens a single-sided connection to any PLC by IP address — but you lose the single connection view, the symbol transfer, and a home for the peer-to-peer PLC links. Each station you copy in carries its hardware configuration unchanged, local DP segment included. If a controller is later replaced by an S7-1500, the same single-sided S7 connection pattern carries the HMI link, as connecting an HMI panel to an S7-1500 PLC shows.

What steps configure the connections in NetPro and WinCC? #

Follow six steps — address the network, slot the CPs, insert the PC stations, insert the connections, download in planned downtime, then create the tags — and you end with 18 configured WinCC connections across the three machines.

  1. Plan one flat IP subnet for the machine network; as an illustrative scheme, put the six CPs at 192.168.0.11 through 192.168.0.16 and the three WinCC PCs at 192.168.0.21 through 192.168.0.23 — any private subnet works as long as every partner shares it.
  2. In each PLC station's hardware configuration, slot the CP 343-1 Lean (or the full CP 343-1) into a free slot in the rack and enter its fixed IP address.
  3. Insert each WinCC machine as a SIMATIC PC station (Insert > Station) and install the matching SIMATIC Net PC software so the station has the Ethernet communication application the project references.
  4. In NetPro, insert the PLC-to-PLC S7 connections between the CPUs that exchange data, plus the 3 × 6 = 18 single-sided S7 connections from the three PC stations' WinCC applications to the CPs, each with the remote TSAP 03.02 — the rack 0, slot 2 access point of the S7-300 CPU.
  5. Download the changed stations — the CPUs and the PC stations — from NetPro. Do this during planned downtime: a connection-configuration download to a CPU in RUN restarts its communication services on a live line.
  6. In WinCC, create the tags on the configured connections; if you skipped PC-station integration, add the same connections by hand under the SIMATIC S7 protocol suite's TCP/IP channel, entering each PLC's IP address and rack 0, slot 2.

Do the six CPUs stay configured as DP masters? #

Each CPU keeps its DP-master role for the slaves on its own local segment — configuring all six as masters with the Standard bus profile creates no PLC-to-PLC traffic and never puts the WinCC layer on a shared DP bus.

A DP-master configuration makes the CPU cycle its own PROFIBUS slaves, and that segment stays exactly as it runs today. Two class 1 masters wired onto one PROFIBUS line do not hand each other slave data: even with all six CPUs on one cable, each would still master only its own slaves and would have no path into the others' data. The exchange this project needs is built as configured connections instead — the S7 connections of the sections above, with each server CPU granting its partners PUT/GET access to the data blocks agreed for exchange. Keep the Standard profile on every local segment; no multi-master DP profile is needed anywhere.

When one PLC fails, do the other stations fault too? #

No — in a switched Ethernet star a dead PLC drops only its own links, so the other five PLCs and all three WinCC stations keep running; a shared DP line instead raises a bus fault on every healthy station on the segment.

That propagation is exactly what the shared-cable idea would have cost you. On one PROFIBUS line threaded through all six racks, a cable break or a failing transceiver brings bus diagnostics up on every master that shares the segment, healthy or not. In the star, the switch isolates the failure to one port: the five surviving CPUs keep their cycles, and in WinCC only the tags on the failed connection stop updating. Configure connection monitoring on each WinCC connection and show a connection-status bit on the screens that use that PLC, so an operator reads a dead station as no data instead of a frozen value that looks alive.

Does a WinCC station still need a CP5611 or extra software? #

Not on this architecture: the Ethernet path uses each PC's standard network interface plus the S7 protocol suite included with WinCC, and a CP5611 is needed only if you deliberately keep a station on MPI or PROFIBUS.

What the PCs do need is the SIMATIC Net PC software that provides the Ethernet communication application a SIMATIC PC station references — install it before downloading the PC-station configuration, and each machine then takes its connections straight from NetPro. If you skip PC-station integration, the same S7 protocol suite channel reaches every CP by IP address with rack 0, slot 2 — no extra card anywhere. The CP5611 belongs to the serial path you are rejecting: three fully independent WinCC stations on MPI or PROFIBUS would each need their own card, while folding them into one server with two clients saves two cards only by re-creating the single-master, single-point-of-failure layout that DP forces on you.

Questions asked

How many S7 connections can one WinCC station carry?

WinCC V8.1 documentation lists 64 S7 connections on the PROFIBUS DP channel, with a typical configuration around 60, and up to 510 S7 connections in Advanced configurations, so six PLCs per station sits far inside the channel limit. The binding constraint in this project is the PLC side: a CP 343-1 Lean serves 8 S7 connections and the full CP 343-1 up to 16, so the card in each rack decides how many operator stations and peer PLCs can attach.

Can the six OP170B panels stay where they are?

Yes, they keep running as they are. Each OP170B stays on its existing link to its own PLC with its current panel project, while the new Ethernet CP sits alongside in the rack and carries only the WinCC and PLC-to-PLC traffic. Nothing in the panels' configuration changes, so operators keep the same screens through the whole migration.

What does single-sided mean on the WinCC-to-PLC connections?

Single-sided means only the WinCC side is configured: the PC opens the connection and the PLC acts as the S7 server answering its read and write services, with no communication blocks added to the PLC program. A PLC-to-PLC exchange is double-sided: both ends are configured in NetPro, and the server CPU grants its partner PUT/GET access to the agreed data blocks.

Why not hang the three WinCC stations on the CPUs' MPI ports?

MPI would work electrically, since every S7-300 CPU has the port and a CP5611 fits a PC, but MPI runs at 187.5 kbit/s and the CPUs' small MPI connection budget is already shared with the existing panels. Three stations polling six PLCs would load it heavily. Treat MPI as a fallback for one panel per PLC, not as the backbone for eighteen cross-links.

What bus profile should the DP masters use after the change?

Leave each CPU on the Standard profile it already runs. The bus profile shapes timing on that CPU's own PROFIBUS segment: token rotation when several masters share a line, and slave polling. None of that touches the new Ethernet layer, and because the six local segments stay separate, no multi-master DP profile is needed anywhere in the project.

Keep reading

  • Troubleshooting

    Why One PLC to Two HMIs Fails Over RS-422 Multilink

    A 1:1 serial test proves one cable and one framing set; it cannot prove a shared n:1 multilink. Fix two HMIs on one RS-422 trunk by running the multilink driver entry on both panels, giving each a unique link identity, matching every serial parameter, and terminating only the trunk ends. The documented limit is 16 displays, so two panels are a configuration problem, never a capacity one.

    Industrial networks

  • How-to

    Connecting an HMI Panel to an S7-1500 PLC

    Confirm the HMI's documented driver covers the S7-1500 for your exact panel model and software revision, put the panel and CPU on the same subnet, permit the connection in the CPU protection settings, then validate one known PLC tag before building screens. A working Ethernet link alone proves nothing about tag access.

    HMI and touchscreens

  • Troubleshooting

    Connecting a CompactLogix PLC to an HMI over EtherNet/IP

    A CIP transport unsupported fault on a Generic Ethernet/IP module in a CompactLogix project means the connection was built the wrong way around. For HMI tag access, delete the module, select the panel's Rockwell PLC driver, point it at the controller IP, and rebuild the tags — no assembly instances, I/O sizes, or EDS file.

    HMI and touchscreens