Troubleshooting
FX3GE RS422 to HMI wiring fix for no response
Short answer
Wire HMI COM1 receive to the FX3GE transmit pair, the HMI transmit pair to the PLC receive pair with matching polarity, and land signal ground on PLC pin 5. The failed map had the transmit pair reversed and ground on pin 3; the crossed, polarity-matched map with ground on pin 5 restored communication.
On this page
Communication between a Mitsubishi FX3GE and a Weintek CMT2058XH panel that shows only no response was, in the confirmed case, a wiring fault on the serial link, and one pin map fixed it. The panel's COM1 runs in RS485 4W mode; the FX3GE peripheral port is RS-422 with a separate transmit and receive pair. Land the PLC transmit pair on the panel's receive pins, the panel's transmit pair on the PLC receive pins with polarity matched, and signal ground on PLC pin 5. The failed attempt had the transmit pair reversed and ground on pin 3; correcting both brought the link up at once.
The pin map that cleared no response #
The link as wired on the machine that recovered:
| Panel COM1 pin | Panel signal | FX3GE RS-422 pin | Port signal |
|---|---|---|---|
| 1 | RX- | 4 | TXD- |
| 2 | RX+ | 7 | TXD+ |
| 3 | TX- | 2 | RXD- |
| 4 | TX+ | 1 | RXD+ |
| 5 | GND | 5 | GND |
Five conductors carry the whole link: two twisted pairs and one signal ground. The RS-422 wiring diagram in Mitsubishi's FX Communication (RS-232C, RS-485, RS-422) manual is the reference for connector numbering on the PLC side. The panel project was built in EasyBuilder Pro (EBPro-6.10.02.300); the software needed no change, because the fault and the fix sat entirely in the cable.
What the failed wiring got wrong #
The failed map and the working map differ on three of the five conductors, and those three deltas are the whole fault:
- The receive pair was already right: panel pins 1 and 2 sat on PLC pins 4 and 7 in both maps.
- The transmit pair was reversed: panel pin 3 (TX-) landed on PLC pin 1 (RXD+) and panel pin 4 (TX+) on PLC pin 2 (RXD-). The working map swaps those two conductors.
- Ground sat on PLC pin 3 in the failed description; the working connection uses pin 5.
Both corrections went in together and the panel came up, so land both. Do not keep the ground on pin 3 while fixing only the pair. Three of five conductors on the wrong pin is 60% of the link (3 ÷ 5 = 0.6), which is why nothing at all crossed.
Why the pairs cross and polarity must match #
RS-422, standardized as TIA/EIA-422, is a differential interface: each direction has its own balanced pair, a compliant driver delivers at least 2 V differential to the line, and a compliant receiver responds to 200 mV. Two wiring rules fall out of that:
- Cross the pairs between devices. PLC transmit must reach panel receive, and panel transmit must reach PLC receive; two transmitters on one pair feed no receiver.
- Match polarity inside each pair, minus to minus and plus to plus, end to end.
An RS485 4W port is the right partner here: in four-wire mode it presents the same two independent differential pairs, so RS-422 and RS485 4W interoperate point to point. The two interface labels describe one electrical arrangement.
Worked example, one reversed pair inverts every bit. The receiver decodes V(+) minus V(-) at its inputs. With panel TX- on PLC RXD+ and panel TX+ on PLC RXD-, a driven +2 V differential arrives as -2 V: the sign flips while the magnitude stays 2 V, still ten times the 200 mV receiver threshold (2 V ÷ 0.2 V = 10). The PLC does not see a marginal signal, it sees confident inverted data, discards every poll the panel sends, and returns nothing down the correctly wired receive pair. A silent no response instead of garbage on screen is the signature of this fault. Pair direction at the topology level is likewise the deciding factor in why one PLC to two HMIs fails over RS-422 multilink.
Rewire and verify #
- Remove power from the panel and the PLC before opening either connector; that is the one moment in this job with live terminals.
- Confirm pin numbering on the FX connector against the documented RS-422 wiring diagram before landing any conductor.
- Land the five conductors per the table, keeping each pair's minus and plus on their own pins; one swapped conductor reproduces the original fault.
- If the cable ends are unmarked, identify each conductor by continuity first — the RXD and TXD conductor identification walkthrough shows the meter procedure.
- Buzz every conductor end to end and confirm there are no pin-to-pin shorts.
- Route the run at least 100 mm (3.94 in) away from main-circuit and power wiring, the separation the FX3G hardware guidance calls for.
- Restore power, download the project, and watch a register you can force from the PLC side.
If the machine carries an FX3G-422-BD expansion board instead of the built-in port, that board exposes the same RS-422 arrangement on an M12 A-coded 8-pin socket; confirm its pinning from its own cable diagram before wiring it identically.
If the panel still reports no response #
The map above is proven on this hardware, so once the cable verifies, the remaining causes are configuration:
- Check that the device definition in the panel project targets the FX family through the RS-422 programming port, with COM1 set to RS485 4W rather than RS485 2W or RS-232.
- Compare the serial parameters at both ends. The working case changed wiring only, so a speed or format mismatch is the next gap to close.
- Re-check pair polarity and the ground pin one conductor at a time against the diagram.
- Declare success only against a value you can change from the PLC side, not a static screen object.
Once data flows but the screen crawls, the cable is done and the bottleneck sits elsewhere, the ground covered in why an HMI freezes or updates slowly over serial.
The decision, plainly: on an FX3GE to CMT2058XH link showing no response, wire the table above with ground on PLC pin 5 and expect communication. If the panel stays silent after the cable verifies, stop rewiring and fix the project's device and serial settings.
Questions asked
Does the HMI port need RS485 4W mode, or will 2W work?
Four-wire mode is required. The FX3GE programming port is RS-422 with a dedicated transmit pair and a dedicated receive pair, so the exchange is full duplex. In two-wire mode the panel puts both directions on one pair, which cannot carry that simultaneous traffic, and the result looks exactly like broken wiring: a flat no response.
Can the FX3GE USB port be used instead of the RS-422 connector?
The FX3GE also has a USB peripheral connector, but its documented role is a programming tool, meaning a PC running the PLC programming software. The RS-422 peripheral connector is the port used for the HMI link described here, so wire that port rather than planning the panel around USB.
Should PLC pins 5 and 6 be bridged to clear odd communication errors?
No. The confirmed fix on this link used a plain signal ground on pin 5 with no bridge to pin 6, and the panel communicated. A conductor bridge across port pins changes the electrical arrangement of the connector, so apply one only when the documentation for that specific port calls for it.
Is no response the same fault as garbled data on the panel?
No, and the difference points at the work. A flat no response means nothing valid crosses the link: wiring, port selection, or a dead driver. Garbled or partly updating values mean bits arrive but the serial format disagrees, so compare speed, parity, and data bits at both ends before touching the cable again.