EL72xx

Topics

  1. Diagnostics
  2. Error/warning
  3. Tuning
  4. Electrical installation issues
  5. OCT cable failure

Diagnostics

The diagnostics buffer can be read with the ec_diagnostic_messages.py tool:

Example: EL7211-9014

python3 ec_diagnostic_messages.py -m1 -s3

DEVICE INFORMATION:
===================

name:		EL7211-9014
master id:	1
slave id:	3
vendor id:	0x2
product id:	0x1c2b3052
host time:	2026-02-05 11:33:41.191176


DIAGNOSTIC MESSAGES:
====================
msg_no  time                        text_id  text  flags  diag_code   dynamic
1       2000-01-01 00:00:00         0x4411         0x1    0x1c21e000  0x0
2       2026-02-05 10:30:39.407639  0x8406         0x2    0x1c21e000  0x0
3       2026-02-05 10:30:39.407639  0x8105         0x2    0x1c21e000  0x0

Sometimes the text_id values are converted to readable messages by the tool, and sometimes not. In the case above, we need to look up the text_id values on Beckhoff’s website.

Searching for the slave type and error code normally gives you the information:

  • 0x4411 : Warning “Drive. DC-Link undervoltage. (Warning). The DC link voltage of the terminal is lower than the parameterized minimum voltage”
  • 0x8406 : Error “Drive. DC-Link undervoltage. (Error). The DC link voltage of the terminal is lower than the parameterized minimum voltage”
  • 0x8105 : Error “General. PD-Watchdog. Communication between the fieldbus and the output stage is secured by a Watchdog.”

Here we can see that the drive is missing DC-link voltage (motor power). The watchdog error probably happens during startup of the IOC and is nothing to worry about.

Error/warning

If the drive is in error/warning state and not possible to enable:

  • Missing power supply
  • STO tripped
  • Defect cable
  • Wrong/messy cabling between connector and actual terminal.
  • Phases connected in wrong order resulting in commutation failure
  • PD-Watchdog error (see below)

PD-Watchdog error

PD watchdog error once occurred at restart of the system after a complete power-down event (yearly shutdown test).

Symptoms:

  • Slave goes to OP but stays in fault state (or sometimes not even in fault state).
  • Position of absolute encoder is 0.
  • Refuses to enable.

Diagnostic tool output below:

sandst_a@sls-ec-di22-05-01:~$ python3 /ioc/NeedfulThings/ecmc_ec_scripts/ec_diagnostic_messages.py -m3 -s14

DEVICE INFORMATION:
===================

name:		EL7211-9014
master id:	3
slave id:	14
vendor id:	0x2
product id:	0x1c2b3052
host time:	2026-02-24 15:39:38.197238


DIAGNOSTIC MESSAGES:
====================
time                        text_id  text                 flags  dynamic
2026-02-23 14:56:55.174172  0x8105   (error) PD-Watchdog  0x2    0x0000000000000000000000

Reason seemed to be that the EtherCAT cable was connected to the lower port of the EK1100 coupler, resulting in wrong timing. Steppers worked fine, but servos that use DC clocks did not work. After moving the cable, a power cycle was needed.

Reset with a power cycle of the power bus (communication bus can remain powered). If EL9227-5500 is before this point in the EtherCAT chain, the power bus can be toggled with channel 1 from panels. At PSI, do not toggle channel 2 since that could be connected to communication-bus feed of EK1100; then communication will be lost and can only be recovered by someone physically going to the crate and pressing the button/LED.

** PSI: Do not toggle channel 2 of EL9227-5500, this may break power supply to EtherCAT communication bus**

Note: restarting the IOC or switching the terminal between INIT and OP did not help for this issue.

Tuning

Always be prepared to “KILL” or estop the axes while performing tuning.

  • Make sure scaling factors are correct. Test by setting ecmc position controller parameters to 0 and performing a move. This move will basically be an open-loop positioning move. Normally, the actual value and setpoint still track very well during the move; if not, then scaling of drive or encoder is most likely wrong. Check gear ratios.
  • For most applications the default values for the current loop and velocity loop is good.
  • For slow motion, running slow, accurate and smooth, it could be beneficial to reduce velocity Kp and Ti in the drive.

Electrical installation issues

Strange issues have occurred if the OCT cable shielding is not kept intact until close to the terminal. Issues were identified when the individual wires were mixed randomly in cable management trays inside a control cabinet (with OCT connector on outside, then going to terminals, then single wires to EL72xx). This is not a good concept. Keep shielding of encoder and motor cables also inside the crate. The diagnostic buffer of EL72xx showed encoder related error messages.

OCT cable failure

Once identified defect OCT cable. Symptoms were frequent disabling of the drive, going to fault state.