Troubleshooting
Fixing HMI Read Errors to a Servo and PLC on One RS-485 Line
Short answer
A read error after merging two DOPSoft projects is usually three faults stacked: colliding station numbers, an HMI link using a PLC-only controller driver so servo tags fail to compile, and an RS-485 pair that never reaches CN3. Set the servo to station 1 (P3-00) and the PLC to station 2, select Delta Controller ASCII, then treat a pure timeout as wiring.
On this page
- Three faults stacked in one symptom
- Unique station numbers on one segment
- Station-prefixed tags in DOPSoft
- Clearing the compile-time address error
- ASDA-B2 communication settings and read format
- Worked example: confirming station 1 by reading P3-00
- A pure read timeout points at the physical pair
- Bring-up sequence
- Questions asked
Fix this as three stacked faults, in this order: station-address syntax, the HMI controller driver, then the RS-485 pair. A merged DOPSoft project that reads an ASDA-B2 servo and a DVP-SV2 PLC on one RS-485 Modbus ASCII line fails at compile time with Input address error and Software check failed as soon as a tag's station prefix points at a register the link's controller driver does not define, and those compile errors sit on top of the wiring fault you have not reached yet. Give the servo station 1 (P3-00) and the PLC station 2, set the link's controller to Delta Controller ASCII, address each slave with tags like {Link2}1@SERVO-040A, and treat a project that compiles clean and still throws COM1 Station 1 Read Error as a physical problem at CN3.
Three faults stacked in one symptom #
Each half works alone: a project that jogs the servo and reads torque and speed, and a project that feeds speed and pulse counts to the PLC for external positioning. Merged into one application with two screens, the HMI reports faults that neither half showed. That is the signature of three independent failure classes stacked on one port: the station and address syntax inside the tags, the controller driver selected for the link, and the physical RS-485 topology. Fix them in that order, because the compile-time errors mask the wiring symptom underneath.
One limit is not a fault at all: the ASDA-B2 series has no internal position (indexing) function, so no position table exists to write over Modbus. Positioning must arrive as an external pulse train, here the DVP-SV2 pulse output into the drive pulse input, while Modbus on the drive carries parameter read and write, mode and command registers, and monitoring such as speed, torque and alarms. A merged project is the right architecture, not a workaround.
Unique station numbers on one segment #
Standard Modbus multidrop, in plain terms: one master per segment, a different station number on every slave, and identical baud rate, data bits, parity, stop bits and ASCII or RTU framing on all of them. Two slaves answering the same address collide and return corrupted frames or nothing at all. Assign:
- ASDA-B2 servo drive: station 1, set in
P3-00on the keypad. - DVP-SV2 PLC: station 2, set in the PLC COM port station parameter.
- DOP HMI: master; its settings live on the COM1 link in DOPSoft.
The working line settings for this system are 9600 baud, 7 data bits, even parity, 1 stop bit, ASCII mode. On the drive those arrive as codes: P3-01 = 11 for 9600 baud and P3-02 = 11 for 7, E, 1 ASCII in this configuration. They are coded values, not literal settings, so read them back on the keypad rather than inferring them. The shared-segment rules behind this assignment are the same ones in Choosing the RS-485 protocol for one HMI and multiple PLCs.
Station-prefixed tags in DOPSoft #
In DOPSoft the station number lives in the tag address, not only in the link, so one link object can address several slaves:
{Link2}1@SERVO-040Awrites to servo register 040A on station 1.{Link2}2@SERVO-040Awrites the same register offset on station 2.{Link2}2@D0reaches PLC data register D0 on station 2.
The number before the @ is the Modbus slave address. Everything after it is the register, in the device type the tag names.
Clearing the compile-time address error #
The symptom: change the 1 to a 2 in a tag and DOPSoft answers Input address error, and the build stops with Software check failed. The cause: editing the station prefix is legal only when the link's controller driver defines that device-type and offset combination. A PLC-only driver defines no servo register types, so every re-stationed servo tag fails validation before anything is ever sent on the wire.
Fix it like this:
- Open the communication settings for the port the link uses, COM1.
- Set Controller to Delta Controller ASCII. That driver defines both the PLC device types (D, M, Y and the rest) and the servo register types.
- Confirm the port parameters match every slave on the pair.
- Reopen each tag and re-enter the station prefix; the device list now accepts it.
- Rebuild until both messages clear.
You do not need one project per device. One project, one link, station-qualified addresses is the mechanism that lets the jog screen and the positioning and monitoring screen coexist.
ASDA-B2 communication settings and read format #
| Parameter | Value used | What it does |
|---|---|---|
| P3-00 | 1 | Modbus station address of the drive; must be unique on the segment |
| P3-01 | 11 | Transmission speed code: 9600 baud in this configuration |
| P3-02 | 11 | Protocol code: 7 data bits, even parity, 1 stop bit, ASCII; this parameter also selects between the ASCII and RTU communication modes |
| P3-05 | 0 | Communication interface selection: the RS-485 port on CN3 |
| P3-06 | 0 | Digital-input source: DI states follow the external terminals |
Power-cycle the drive after editing where the manual calls for it. Treat P3-06 with care: it assigns, bit by bit, whether each digital input follows the external wiring or the communication register, so forcing it to 0 silently disables any DI the application drives over Modbus.
Reads fail for format reasons too. Servo parameters are 16-bit or 32-bit depending on the parameter, and reading a 32-bit parameter as a single 16-bit word returns half the value or raises a driver-level read error. Set the tag data length and format to match the parameter definition in the manual's Modbus address table. Wrong word order looks like garbage data rather than a timeout, and Displaying a PLC Real Value Over Modbus on an HMI works through that side of it.
Worked example: confirming station 1 by reading P3-00 #
Inputs:
- Servo on station 1,
P3-00 = 1, verified on the drive keypad. - Link: Delta Controller ASCII, 9600 baud, 7 data bits, even parity, 1 stop bit, ASCII, as set above.
- Tag syntax:
{Link2}station@register. - Address rule, from the ASDA-B2 documentation: a parameter's Modbus register address = (parameter number x 2) minus 1, with the parameter table in the drive manual as the authority for your firmware.
Arithmetic:
- P3-00 is parameter number 300.
- 300 x 2 = 600.
- 600 - 1 = 599 (decimal).
- 599 = (2 x 256) + (5 x 16) + 7 = 0x0257 hexadecimal.
Result: one read-only numeric tag, {Link2}1@SERVO-0257, 16-bit unsigned, decimal format, on a test screen.
Decision: the tag returns 1, so station, driver, format and pair are all proven; move on and poll the PLC with {Link2}2@D0. The tag times out while the keypad still shows P3-00 = 1, so the fault is physical, not software; go straight to the wiring checks below. The tag answers with a value other than 1, which means the register address, not the wiring: look the parameter up in the manual's table before touching a cable.
A pure read timeout points at the physical pair #
COM1 Station 1 Read Error, with settings matched and the project compiling, is a physical-layer fault. The distinction is diagnostic: address errors and write errors mean frames are being exchanged and answered badly, while a total blackout means no frames are crossing at all. For a slave that stays silent on RS-485 in other setups, Fixing HMI Modbus Device No Response Over RS-485 works the same class of fault.
Work this sequence:
- Isolate. Drop the PLC off the pair and cable the HMI straight to the servo's CN3. A servo that answers exonerates the drive and the project; the fault is in the segment you removed.
- Verify the CN3 pin assignment. Check D+, D- and signal ground against the CN3 pin description in the drive manual. A swapped differential pair produces exactly this symptom: perfect configuration, zero traffic.
- Check the chain, not the order. HMI to PLC (station 2) to servo (station 1) is acceptable once the loop is completed; multidrop does not care which slave sits closest to the master. What breaks it is a device that terminates the pair internally, a spur that is not a true pass-through, or an inline connector that transposes the pair. Meter continuity from the HMI connector to the CN3 pins.
- Carry the signal common through the whole chain, on shielded twisted pair with the shield bonded at one end only.
- Terminate at the two physical ends of the segment only, as the drive and PLC manuals specify.
The DC bus inside the drive stays charged after power-off for the discharge time the manual states; let that elapse before re-pinning any connector.
Bring-up sequence #
- On the servo keypad, read back the five P3 values from the table above, then power-cycle the drive.
- On the PLC, confirm station 2 and the same line format.
- In DOPSoft, set Controller to Delta Controller ASCII and rebuild until the compile messages clear.
- Poll one register per station from a test screen: the P3-00 read from the worked example for the servo, one D register for the PLC.
- Reconnect both slaves, direct cable first, then through the PLC, before restoring the full application.
Questions asked
What does the COM1 Station 1 Read Error message actually report?
It names the link and the slave: the master sent a read to station 1 on COM1 and no valid response arrived within the communication timeout. Because the station number is in the message, the poll was addressed correctly, so the failure is on the response side: line settings, register definition, or the physical pair. It is aimed at the servo, not the PLC.
Why would the servo answer with wrong values instead of timing out?
A reply that exists but reads wrong is a format problem, not a wiring one. A 32-bit parameter read as a single 16-bit word returns half the value; wrong word order or a signed and unsigned mismatch scrambles it further. Match the tag data length and format to the parameter entry in the drive manual Modbus address table.
Does setting P3-06 to 0 stop Modbus from controlling inputs?
On the bits you set to 0, yes. P3-06 assigns each digital input, bit by bit, to either the external terminals or the communication register. Zeroing it while the application drives any DI over Modbus, a software servo enable for instance, makes that command silently stop working. Set it per input according to what the PLC actually wires.
Do both slaves really need identical baud and parity?
Yes. Every device on one RS-485 segment must share baud, data bits, parity, stop bits and ASCII or RTU framing; that is the multidrop rule. One slave left in a different format stays silent on every poll, which is also a clue: a timeout on one station only usually points at that station format, not at the pair.
Can one link object in DOPSoft address both slaves?
Yes, because the station number lives in the tag address, not only in the link. Tags like 1@ and 2@ prefixes on the same COM1 link each carry their own slave address, so one link object polls both the servo and the PLC as long as the controller driver accepts every device type used across those tags. The moment a tag names a device type the driver does not define, the build fails at compile time before any frame is sent.