Resolving Yaskawa NX100 EA1400N Serial I/O Error on Power-Up

Problem Overview

The Yaskawa Motoman NX100 controller on an EA1400N arc-welding robot raises a persistent Serial I/O fault immediately after every cold start. The alarm appears before any motion is requested and remains latched until the operator manually toggles a specific general-purpose output. On the affected cell, the message clears only when output 1019 (RELAY ON) is forced high and then released, even though no Programmable Logic Controller (PLC) is physically wired to the cell in the repair bay.

This symptom is a classic NX100 behavior: the controller executes its built-in PLC-interface ladder at boot and validates that an external PLC handshake has occurred within a defined window. If the handshake is missing, the system latches a communication-class alarm and waits for an explicit release. The release signal in this installation is hard-coded to output 1019, so toggling it manually clears the latch - but the alarm returns on the next power cycle.

The definitive remediation is a ladder-logic change, not a parameter toggle. This article documents the fault class, the signal assignments, the bench-test workaround, and the verification procedure required to clear the condition permanently once the cell is connected to a real PLC.

Affected System Identification

The EA1400N is normally deployed as a welding cell with a higher-level PLC coordinating the cell sequencer, the weld power source, the safety gate, and the robot's start/stop handshake. When the cell PLC is not present - for example on a single-robot test bench, in a refurbishment staging area, or during pre-shipment inspection - the NX100 still runs its PLC-interface scan and the missing handshake manifests as a Serial I/O alarm.

Root Cause Analysis

The NX100 maintains a resident ladder program (commonly referenced as the system ladder or PLC interface ladder) that arbitrates a defined set of input and output words between the controller and an external PLC. At power-up, the ladder performs a sanity check on the communication path. If the expected inputs from the PLC are not toggled within the configured supervision time, the controller raises a communication-class alarm and latches the associated error-release output low.

For the EA1400N cell, the alarm class is reported as Serial I/O. Per the official Yaskawa NX100 Error Message List, the Serial I/O alarm family includes the following sub-codes:

6

For a bench-test installation with no PLC at all, the dominant sub-code is 7 - Data time out: the NX100 transmits its handshake word and never receives the expected reply, so the supervision timer expires and the alarm latches. Sub-code 8 is the generic category reported when the controller cannot classify the failure more precisely.

The output that releases the latch is mapped by the system ladder. On the affected EA1400N cell, that release line is OUT1019. Toggling 1019 high asserts the release edge to the alarm latch, and toggling it low allows the ladder to re-evaluate. Because the upstream condition (missing PLC handshake) is still present, the alarm does not re-latch immediately - it re-latches on the next supervision cycle or the next cold start.

I/O Signal Map and Output 1019

The NX100 general-purpose I/O area is split into a controller-side block and a user-side block. The signals relevant to this fault live in the controller block, which is normally driven by the system ladder and read by the cell PLC. The relevant signal list for an EA1400N with the standard Motoman PLC interface is summarized below; verify exact numbering against the as-built schematic for the specific cell, because Motoman offers several PLC-interface variants (EtherNet/IP, DeviceNet, Profibus, and discrete hardwired).

NX100 PLC Communication Architecture

The NX100 implements its PLC interface as a resident ladder that is loaded into a protected area of the controller firmware. The ladder is responsible for:

1. Publishing robot status (ready, in motion, error, hold, etc.) to the cell PLC.
2. Consuming cell commands (start, stop, hold, error reset, job select, etc.) from the cell PLC.
3. Supervising the communication link for time-out, framing, parity, and protocol violations.
4. Latching any detected violation to a Serial I/O alarm and exposing the release line.

The supervision block runs at a fixed cycle after power-up. If the expected PLC handshake (a defined bit pattern on the configured input word) does not appear within the supervision window, the ladder declares the link unhealthy, raises the alarm, and holds the controller in a non-motion state. The release line OUT1019 is the only path the operator has to clear the latch without an active PLC reply.

This is by design: in a production cell, a dropped PLC handshake is treated as a hard safety-relevant event, not a nuisance alarm. Forcing the release line acknowledges the operator's awareness of the situation and permits recovery.

Solution Path: Ladder Logic Modification

The permanent fix is to modify the PLC interface ladder so that the Serial I/O supervision block is bypassed, or so that the expected handshake is generated locally when no external PLC is wired. This is a code change in the system ladder, not a parameter change. The user-supplied source confirms this: "No, it would require a ladder change. Contact your local Yaskawa office."

There are two practical approaches:

1. Add a self-handshake rung. Modify the system ladder to force the expected PLC input word to the "healthy" pattern using a constant or a derived internal bit, so the supervision block always sees a valid reply. The release line is then driven to its idle state automatically on every boot.
2. Disable the supervision rung. Comment out or bypass the time-out rung that raises the Serial I/O alarm. This is appropriate only for bench-test, refurbishment, or pre-shipment configurations; it must not be left in place for a production cell, because the Serial I/O check exists to catch real fieldbus faults.

Both changes are made with the Motoman MotoCom32 or the Standard Ladder Editor tool, then downloaded to the NX100 and saved to compact flash. After the change, the controller must be cold-started twice to confirm the change persists across a full power cycle.

Bench Test Workaround (No PLC Connected)

When the ladder cannot be modified - for example, when the technician is on the road with only the teach pendant - the fault can be cleared at every power-up by forcing output 1019 from the teach pendant. The procedure is:

1. Power on the NX100. Wait for the controller to complete POST and the Serial I/O alarm to appear.
2. On the teach pendant, navigate to IN/OUT → OUT → User Outputs.
3. Locate output 1019.
4. Force 1019 to ON. The Serial I/O alarm should clear within one ladder scan.
5. Force 1019 back to OFF. The controller returns to the normal ready state.

For repeatable bench testing, a small program can be written that performs steps 4 and 5 automatically on power-up, but this only delays the underlying ladder change.

Parameter Verification

While the release path is hard-coded into the ladder, several NX100 parameters influence whether the supervision block is even active. Confirm the following from the pendant under SETUP → FUNCTION ENABLE and SETUP → OPTION:

Note: the exact menu paths and parameter mnemonics vary across NX100 firmware revisions. Always cross-check against the NX100 Operator's Manual and the NX100 PLC Interface Manual for the firmware version loaded on the affected controller.

Verification Steps

After either the ladder change or the parameter change, validate the fix with the following sequence:

1. Power off the controller completely (main breaker, not just teach-pendant standby). Wait 30 seconds.
2. Power on the controller. Observe the boot sequence. No Serial I/O alarm should be raised.
3. Confirm the controller reaches the READY state without operator intervention.
4. From the teach pendant, run a low-speed jog in TEACH mode on each axis to confirm motion is enabled.
5. From the teach pendant, run the cell test job in PLAY mode at 10 percent override to confirm the application program loads and executes.
6. Power off and on a second time. Repeat steps 2-5.
7. Power off and on a third time. Repeat steps 2-5.

Three clean cold starts in a row demonstrate that the change is persistent and that no manual output toggle is required.

Troubleshooting Matrix

Safety Considerations

Before any ladder change on an NX100:

• Back up the existing system ladder to a known-good compact flash card.
• Document the change with date, author, controller serial number, and firmware version.
• Verify that the safety-rated I/O (E-stop, safety gate, mode select) is not affected by the bypass. The PLC interface ladder is not part of the safety chain; safety I/O is handled by the safety option board and must remain untouched.
• Confirm that no other cell is in production on the same controller variant while the change is being downloaded.
• After the change, perform a documented risk assessment that explicitly accepts the reduced supervision while the bench is unhooked from a PLC.

Field-Proven Caveats

Three caveats gathered from field experience with the EA1400N and similar NX100 cells:

1. Forcing 1019 from the teach pendant does not survive a cold start; the output returns to its idle state on every reboot. Plan to re-toggle at every power-up until the ladder is modified.
2. If the fieldbus option card is present but no PLC is wired, the option card itself will raise its own diagnostic. Check the option card's LED pattern - a steady or slowly-blinking red often indicates "no link partner," which is the expected state in a bench environment.
3. If the cell has been re-flashed with a newer firmware version, the system ladder format may have changed. The previously edited ladder may not load cleanly, and the Serial I/O fault may return in a different form. Always reload the ladder that matches the installed firmware.

FAQ