Troubleshooting a Noisy Encoder Signal

Read the alarm history and identify the exact fault

On the teach pendant press [MENU], select 4 ALARM, then press F3, HIST to view the recorded alarms. Note the SRVO codes and the group and axis numbers shown in the bracket, and use the DETAIL key for the full message. The codes that point at feedback-line noise are SRVO-068 DTERR (the serial Pulsecoder returned no serial data in response to a request), SRVO-069 CRCERR (the serial data was disturbed during communication), and SRVO-070 STBERR (the start and stop bits of the serial data are abnormal).

Separate the primary code from misleading secondary alarms

The communication faults can post alongside other Pulsecoder alarms that are not actually occurring. The manual notes that alarms such as SRVO-063, SRVO-066, SRVO-072, and SRVO-073 may appear together with a DTERR, CRCERR, or STBERR and can be disregarded in that case. Treat the DTERR, CRCERR, or STBERR code as the primary lead and work its remedy first.

Caution: Chasing a secondary code first can send you replacing a motor or Pulsecoder that is fine. Confirm the communication code and clear the cable and grounding path before condemning any part.

Confirm how many axes are affected

The bracket in the message gives the group and axis number for each alarm. If several axes on the same servo amplifier alarm together, suspect a shared cable, connector, or grounding problem rather than a single motor. A fault isolated to one axis points more toward that axis's connection or internal cable.

Correlate the fault with machine events

Check whether the alarm timing lines up with a specific event, such as a welder firing or a large motor starting nearby. Intermittent communication errors that track an external event strongly suggest injected electrical noise rather than a failing encoder.

Reseat the connection cable connectors

The first DTERR remedy is to make sure the robot connection cable (RMP1, RP1) connector CRF8 at the 6-axis servo amplifier and the connector on the motor side are connected tightly. Power down, then check both ends are fully seated and latched. While there, inspect the RMP1 and RP1 cable for cuts or kinks.

Verify shield termination in the cabinet

DTERR remedy action 2 is to check that the shielding of the robot connection cable (RMP1, RP1) is grounded securely in the cabinet. Confirm the shield land is clean, tight, and bonded as the manual specifies. For a line tracking encoder the manual instead directs you to confirm the connection cable shielding is connected securely to the grounding plate.

Check the controller earth path

The CMAL (SRVO-073) and RCAL (SRVO-063) remedies both start with the earth. Confirm the controller earth is connected correctly and that the earth cable between the controller and the robot connection cable is grounded securely to the grounding plate. On an auxiliary axis, reinforce the earth at the motor flange as the manual describes.

Look for a shared conduit with power cabling

Check whether the encoder wiring runs in the same conduit or tray as motor power or welding cables without adequate separation. Coupling from high-current conductors is a common source of the disturbance that produces CRCERR and STBERR.

Understand cross-axis noise from a battery-line ground fault

The DTERR entry carries an important note: an intermittent ground fault on the 6V battery line inside a Pulsecoder cable can make a DTERR appear on a different axis, through noise, when the ground fault occurs. So a DTERR reported on one axis can trace back to a damaged cable on another axis. Inspect the whole feedback harness, not only the axis named in the alarm.

Isolate a suspected ground fault with the connector-off test

The manual sanctions a diagnostic isolation test: turn off the power, disconnect the CRF8 connector on the 6-axis servo amplifier, then power on and see whether the fault recurs. Ignore the SRVO-068 that the disconnection itself raises. If the original alarm no longer occurs with the connector off, the robot connection cable or the internal cable of the robot may be short-circuited to ground.

Caution: The robot cannot be operated in this state; this is a diagnostic check only. Keep everyone clear of the envelope and reconnect CRF8 before returning the robot to service.

Tester-check the mechanical unit cable

Use a tester to check the robot connection cable and the internal cable running to the Pulsecoder for a ground fault or disconnection. The DTERR remedy directs you to replace the mechanical unit cable if the tester check finds an abnormality. Measure with the cable disconnected at both ends so a parallel path does not mask the fault.

Back up the controller before replacing anything

Before pulling cables or boards, perform a complete controller backup so all programs and settings are saved. The maintenance manual calls for this backup ahead of major replacements. Save the backup to USB and confirm it before continuing.

Replace the confirmed faulty cable

If the checks point to the cable, replace the robot connection cable (RMP1, RP1, RM1) or the internal Pulsecoder cable of the robot as indicated. The manual lists motor, Pulsecoder, and servo amplifier replacement only after the cable and grounding path have been cleared. Correct the physical path first so a new part does not simply inherit the same noise.

Caution: Replacing a motor or Pulsecoder requires mastering afterward. Do not swap those parts until the cable, shield, and earth have been ruled out.

Clear the alarm correctly after fixing the cause

The DTERR, CRCERR, and STBERR communication alarms release on a power cycle once the cause is gone. Certain related Pulsecoder alarms (BZAL, RCAL, CMAL) instead need a PULSE RESET followed by a power cycle: press [MENU], select SYSTEM, press F1 [TYPE], select Master/Cal, press F3 RES_PCA, then F4 YES. If Master/Cal is not on the [TYPE] menu, open Variables, set $MASTER_ENB to 1, and reopen the [TYPE] menu.

Re-establish pulse and re-master if absolute position was lost

If a battery-line fault dropped the absolute position, SRVO-075 Pulse not established will appear after the underlying alarm. Reset it and jog the affected axis about one motor revolution in T1 low speed with the deadman held and the cell clear of people, until the alarm stops recurring; it must be cleared before mastering. If position data was lost, re-master the robot before returning to production.

Test through a full production cycle

Run the cell through the cycles that previously triggered the fault, including the external events you correlated earlier. Watch the ALARM history for any return of the DTERR, CRCERR, or STBERR code. Confirm the fault does not reappear across repeated cycles before signing off.

What can go wrong: the noise alarm returns at reduced speed

If the communication fault comes back after a PULSE RESET or power cycle, the manual is explicit that the cause of the alarm may remain. Do not keep clearing and running. Return to the connector, shield, earth, and cable checks and correct the physical source before resuming.

Caution: Never mask an encoder communication fault with gain changes or by repeatedly resetting and running. A defeated feedback fault can lead to uncommanded or abnormal motion.

What can go wrong: a phantom motor alarm sends you replacing the wrong part

An SPHAL (SRVO-071) can be a mis-detection due to electrical noise, and its remedy is to check that the Pulsecoder cable shield is grounded securely before anything is replaced. Likewise, a CMAL (SRVO-073) can mean the Pulsecoder malfunctioned due to noise, and its remedy begins with the controller earth. When these appear with a DTERR, CRCERR, or STBERR, treat the communication code and the grounding path as the real target.

Caution: Replacing a Pulsecoder or motor for a noise-induced SPHAL or CMAL wastes the part and still leaves the noise source live. Verify the earth and shield first.