CANopen + robotec closed/open loop problem

[Gargamel]

Hi.

 We have noticed on a machine we are building, that when the robotec is booting and the plc/canopen is connected, something from the PLC changes the closed loop to open loop on the robotec.
I have been looking around "everywhere" inthe unilogic project, but cannot find anything that seems to relate to the settings for open/closed loop.

Anyone know what this behaviour could occur at and why?

[kratmel]

Please post robotec model # and used unistream hardware.

I think that such behavior is possible only if the robotec works out the emergency algorithm during the loss of communication.

This algorithm (or parameter) should be configurable. That is, if there is a timeout in the connection, the system must do something to protect the user. In your case, it can be a programmed transition to an open loop when communication is lost.

[Gargamel]

13 hours ago, kratmel said:

Please post robotec model # and used unistream hardware.

I think that such behavior is possible only if the robotec works out the emergency algorithm during the loss of communication.

This algorithm (or parameter) should be configurable. That is, if there is a timeout in the connection, the system must do something to protect the user. In your case, it can be a programmed transition to an open loop when communication is lost.

Unilogic USC-x10-T24.

Robotec FBL2360T

[Gargamel]

When out machine goes from auto to manual mode, LoopErr starts to blink.

And the behaviour is, as soon as the plc(?) connects/Starts up with the canopen cable connected. Teh setting changes to OpenLoop in the controllers, where it was ClosedLoop previously.

 

 

[kratmel]

https://www.roboteq.com/docman-list/motor-controllers-documents-and-files/documentation/user-manual/1047-roboteq-controllers-user-manual-v2-1a/file

p.183

For applications demanding the highest operating safety, the controller should be config-
ured to automatically switch to another command mode or to stop the motor (but other-
wise remain fully active) if it fails to receive a valid command on its RS232, RS485, TCP,
USB or CAN ports, or from a MicroBasic Script for more than a predefined period.
By default, the watchdog is enabled with a timeout period of 1 second. Timeout period
can be changed or the watchdog can be disabled by the user. When the watchdog is en-
abled and timeout expires, then the controller will go to a safety stop (force the motor to
stop based on Fault Deceleration configuration command). After that the controller will
accept commands from the next source in the priority list. See “Command Priorities” on
page 181

Please try to disable watchdog for test.

Then try to activate Heartbeat with enabled Watchdog.

 

[Gargamel]

10 minutes ago, kratmel said:

https://www.roboteq.com/docman-list/motor-controllers-documents-and-files/documentation/user-manual/1047-roboteq-controllers-user-manual-v2-1a/file

p.183

For applications demanding the highest operating safety, the controller should be config-
ured to automatically switch to another command mode or to stop the motor (but other-
wise remain fully active) if it fails to receive a valid command on its RS232, RS485, TCP,
USB or CAN ports, or from a MicroBasic Script for more than a predefined period.
By default, the watchdog is enabled with a timeout period of 1 second. Timeout period
can be changed or the watchdog can be disabled by the user. When the watchdog is en-
abled and timeout expires, then the controller will go to a safety stop (force the motor to
stop based on Fault Deceleration configuration command). After that the controller will
accept commands from the next source in the priority list. See “Command Priorities” on
page 181

Please try to disable watchdog for test.

Then try to activate Heartbeat with enabled Watchdog.

 

Hello Kratmel.

The HB is/was already enabled, and set to 100ms

In the controllers CHB Lost Action is set to "No Action".

Both controller and PLC has 100ms HB, but i suspect that the boot up of the PLC takes way longer than 1s, closer to a minute before i can connect, from power on.
Watchdog, is it on the PLC or the controllers side?
In both cases, "CANopen node guarding" is set to 0.

 

 

[kratmel]

20 minutes ago, Gargamel said:

Watchdog, is it on the PLC or the controllers side?

On contoller side.

39 minutes ago, kratmel said:

automatically switch to another command mode or to stop the motor

 

[Gargamel]

A small update.

We are running a beta-version of the current release, and there was a small bug in it, where it listened to all CAN-Id's, where is to happened to be another device recived a message that the drive controller interpeted it to change to open loop.

[kratmel]

Please change temporary watchdog settings in controller via controller Commisioning software. Disable Communication Watchdog for test.

Watchdog must have timer settings and error reaction settings.