CC-Link IE: A Station Will Not Join or Exchange Data

intermediate · 45 min

A field procedure for Mitsubishi CC-Link IE networks when a station refuses to join or cyclic data stops moving. Separate wiring faults from configuration faults using module LEDs, then confirm station numbers, the master's network configuration and link refresh in GX Works3.

ℹ️ A CC-Link IE station that will not join is almost never a broken device. It is a disagreement — the master's configured picture of the network versus what is actually plugged in. Your job is not to repair a station; it is to find where those two pictures differ. The LEDs tell you which of the two to look at first.

Two Mitsubishi CC-Link IE Field remote stations are set to station 02 on their rotary switches — the master raises a station-number duplication error, the duplicate is held out of the cyclic link, and its slice of the RX / RY / RWr / RWw link devices never refreshes.

First: which CC-Link IE are you on?

CC-Link IE is Mitsubishi Electric's industrial Ethernet family. Mitsubishi developed CC-Link in 1996, released the specification openly in 2000, and the CC-Link Partner Association was founded in November 2000; CLPA now manages the specification. It is the dominant automation network in Japan and much of Asia and uncommon elsewhere. Everything in this procedure is Mitsubishi-specific. Names, menus and LEDs will not match a Siemens or Rockwell machine.

Four family members fail in different ways, and — this matters more than anything else in this article — their troubleshooting lists are not interchangeable. Most bad CC-Link IE advice on the internet is one variant's check list applied to another.

• CC-Link IE Control (Controller) Network — controller to controller. 1 Gbps. Twisted pair (Cat5e or better, double-shielded STP, 100 m between stations) or multimode optical fibre (550 m between stations with a 50 µm core). Line, star and ring; the optical duplex-loop version loops back on a break. Up to 120 stations: 1 control station and 119 normal stations. Link devices are LB, LW, LX, LY.
• CC-Link IE Field — controller to field devices. 1 Gbps over Cat5e or better, double-shielded (STP), straight Ethernet cable, 100 m between stations. Line, star and ring. 121 stations: 1 master plus 120 slave stations. Link devices are RX/RY/RWr/RWw. Needs a dedicated network module (e.g. RJ71GF11-T2) and its own cable run. It is addressed by network number + station number, not by IP address.
• CC-Link IE Field Basic — a software protocol running from a CPU's built-in Ethernet port. No special module, no special cable. 100 Mbps, 100BASE-TX, star topology only. On MELSEC iQ-R the master handles up to 64 slave stations (16 occupied stations × 4 groups); a slave occupies 1 to 4 stations. Cyclic data rides UDP port 61450; the slave listens on UDP 61451 for automatic device detection. This one is addressed by IP address and subnet mask.
• CC-Link IE TSN — the current generation. Time-sensitive networking lets control traffic and normal Ethernet coexist. Devices are ranked by CC-Link IE TSN Class (older manuals say "Authentication Class"; "Slave station" also became "Device station" in TSN documentation). Addressed by IP address, and it supports 100 Mbps device stations as well as 1 Gbps ones.

The words that explain most faults

Station roles. On CC-Link IE Field, the master station controls the network and only one station may be set as master. A submaster station is a real, legitimate second master-capable station that takes over if the master is disconnected — the MST LED flashes on it. Below the master, slave station is the generic term covering the local station (has its own CPU program), the remote I/O station (bit I/O only), the remote device station (bit and word data) and the intelligent device station (also initiates transient transmission). Mitsubishi's TSN documentation uses device station for the same idea.

Network number and station number. The network number identifies the whole network (1 to 239); the station number identifies one station inside it. On IE Field the master is station 0 and slave stations are 1 to 120. Two stations with the same station number on the same network is a duplication error; two stations both set as master is an overlap error. In both cases you remove the cause, then reset or power-cycle the module to recover.

Link devices. RX = bit data in from a device, RY = bit data out to a device, RWr = word data in, RWw = word data out. These live in the network module, not in your ladder.

Link refresh. The automatic copy between link devices and CPU devices. Configure it in GX Works3 at [Navigation window] → [Parameter] → [Module Information] → your module → [Basic Settings] → [Refresh Settings]. A typical iQ-R job maps RX to X, RY to Y, and RWr/RWw to W or D. If refresh is wrong, the network can be perfectly healthy and your program still sees nothing.

Interlock bits, and their polarity. For cyclic transmission, interlock on SB0049 (data link error status of own station) and the per-station bits of SW00B0 to SW00B7 (SW00B0.0 = station 1, SW00B0.1 = station 2, …). For transient transmission, use SB0047 (baton pass status of own station) and SW00A0 to SW00A7. These bits are ON when there is an error — Mitsubishi's module labels are bSts_DataLinkError and bnSts_DataLinkError_Station[n]. Gate your logic on the bit being off. Getting this backwards means your program runs on stale link data at exactly the moment the link is down.

Tell the causes apart before you change anything

Read the module LEDs first, then branch — do not work through the steps linearly. The LED tells you whether you are in the wiring branch or the parameter branch, and only the parameter branch involves stopping the machine.

• LINK LED off on a port — no Ethernet link at all. Wiring branch. Mitsubishi's checks: cable does not conform to the Ethernet standard, station-to-station distance out of range, bending radius violated, cable disconnected, switching hub not operating normally, or the station at the other end not operating.
• L ER LED on (port) or L ERR LED on (module) — abnormal data received, or loopback being performed. On a ring topology, L ER on at a port is the ring looping back around a break — the ring is working, and the break is elsewhere. Checks for both: cable conformance, station-to-station distance, disconnected cable, switching hub, noise near the module or cables, and whether "Network Topology" under Basic Settings is set to "Ring" with the ring correctly wired at that port. The module-level L ERR list adds: are other stations operating normally, and is "Module Operation Mode" under Application Settings of the master set to "Online".
• D LINK LED off or flashing — cyclic transmission is not running for this station. Mitsubishi's single check list for this state spans both layers, so do not assume it is a parameter problem. Physical items: cable conformance, station-to-station distance, bending radius, disconnected cable, switching hub. Topology items: a switching hub wired into a ring; both P1 and P2 into the same hub; the actual wiring not matching the "Network Topology" parameter (wired as a ring but set to Line/Star, or wired as a star but set to Ring); devices of different network types sharing one cable run. Configuration items: 121 or more device stations connected; duplicated station numbers; the master itself not operating or not connected. After correcting any of these, power off and on or reset the CPU modules in every station where the error was detected.
• ERR LED — read it together with the RUN LED, and note the flash interval. RUN off with ERR on or flashing is a major error (hardware or memory), and the module stops. RUN on with ERR flashing is a moderate error, typically a parameter error, and the module stops operating. RUN on with ERR on is a minor error such as a communication failure, and the module keeps operating. On the RJ71GF11-T2 a 500 ms flash specifically means a data link faulty station was detected; a 200 ms flash means a module error. Open module diagnostics and act on the displayed cause. If diagnostics shows a disconnected station, the documented action is to correct "Network Configuration Settings" under "Basic Settings" of the master to match the device stations actually connected, then run the D LINK troubleshooting above.

Then go to the engineering tool. In GX Works3 the menu is [Diagnostics] → [CC-Link IE TSN/CC-Link IE Field Diagnostics] on current versions, and [Diagnostics] → [CC-Link IE Field Diagnostics] on older ones. It opens a live network map plus a Selected Station Communication Status Monitor. For Field Basic the menu is [Diagnostics] → [CC-Link IEF Basic Diagnostics], which reports each slave's Transmission Status as Transmitting, Unfixed (communications with the master never established) or Disconnecting, plus Disconnections and Time-out Count accumulators (0, then 1 to 65535, rolling back to 1 on overflow). Unfixed counts exclude reserved stations.

Read the map, not the device. A station drawn but greyed out is configured and absent. A station present on the wire but absent from the map was never configured — the master will not exchange data with a station it does not know about.

Step-by-step

⚠️ Steps 5b and 6 write parameters and reset the CPU. Outputs drop and the machine stops. Get permission, and make sure nothing is mid-cycle.

1. Photograph the LEDs on the failed station and on the master before touching anything.
2. Open the CC-Link IE diagnostics from the master. Note which stations are missing and which are in error.
3. Branch on the LED reading.

• LINK LED off, or L ER / L ERR on → go to step 4. Do not write parameters yet.
• LINK on, D LINK off or flashing, or ERR set → go to step 5.

1. Wiring branch. ⚠️ Reseating a cable on a line topology disconnects every station downstream, and on a ring it forces a loopback. Confirm the cable is Cat5e-or-better double-shielded STP, under 100 m, bend radius within spec, and clicks home at both ends. Check the hub is powered, conforms to the standard, and is not wired into a ring. Check that P1 and P2 of one module do not go to the same hub. Check for a noise source beside the module or cable run. If the LED clears, stop here — you never needed to touch a parameter.
2. Configuration branch.

• a. Compare the master's Network Configuration Settings list against what is physically installed — model, station number (STA#), occupied stations, and, on IE TSN or Field Basic, IP address. Every row must match a real box. Confirm no duplicate station numbers, and that the actual wiring matches the "Network Topology" setting. Confirm you are not over 120 device stations. Check whether the station in question is set as a Reserved Station or an Error Invalid Station — either will make it silently absent.
• b. Check the station number on the device itself. Many remote modules set it with rotary switches; the switch value must equal STA# in the configuration, and switch changes usually take effect only at power-up.
• c. On CC-Link IE TSN or Field Basic only, confirm no duplicate IP addresses, one shared subnet mask and matching network address, and that the master's IP Filter Settings are not blocking the station. These checks do not apply to classic CC-Link IE Field, which has no IP addressing in its cyclic path.

1. Verify Refresh Settings map the station's RX/RY/RWr/RWw to CPU devices you actually read, and that those CPU devices are not also written by something else. ⚠️ Writing corrected parameters requires [Online] → [Write to PLC] and a CPU reset or power cycle — outputs drop. After a duplication or overlap error, every affected station needs a reset or power cycle before it will rejoin.
2. Re-check the diagnostics window. Watch Disconnections and Time-out Count: rising counters mean intermittent cabling, noise, hub load or a timeout set too short, not configuration. On Field Basic, a timeout of roughly five times the link scan time is the usual starting point.
3. If a Field Basic slave still shows an error and your GX Works3 predates 1.040S, the window will not show the slave's own error code. Read it from buffer memory: set the target station number in Diagnostic request information (Un\G1051), turn on Diagnostic information display request (Un\G1050.0), then read Diagnostic information 1 (Un\G1053 to Un\G1064) and Diagnostic information 2 (Un\G1068 to Un\G1077). The request bit clears itself once the data is stored.

Never do this

Do not "fix" a station that will not join by deleting it from the master's network configuration, or by hiding it behind Reserved Station or Error Invalid Station.

• A Reserved Station is counted in the network but not expected on the wire, and is not detected as faulty. Its real purpose is to hold link device assignments for equipment you have not installed yet, so that adding it later does not shift everyone else's addresses.
• An Error Invalid Station is the more dangerous one, because it is defined precisely as a station the master will not report as faulty even if it drops out during data link. That is exactly the symptom you were called out for.

Both can also be applied temporarily, at runtime, without changing the written parameters — via Reserved Station Function Enable and Enable/Disable Ignore Station Errors. So a machine can be running in a silenced state that does not show up in the project file. When a station is genuinely absent by design, the alarm disappears, the machine looks healthy, and that station's outputs are simply never driven; weeks later somebody finds an interlock that has been dead the whole time.

Fix the duplicate station number, the topology mismatch, or the missing configuration entry instead — and if a station really is decommissioned, remove it from the configuration deliberately and record it, rather than masking it.