A new Backup mode has been added.

The Eclipse Status window is now enabled all the time. This allows constant access to a new option, F5-Start Backup Mode. When selected, the controller reboots with known IP settings, Network Unit ID and Eclipse Source. Network Unit ID will be configured as unit 1. Eclipse Source will be configured for Ethernet. The Ip Address will be configured as xx.xx.xx.xx. The IP Subnet Mask will be configured as 255.255.255.00. The IP Gateway will be configured as 0.0.0.0. This allows a new backup tool to use Eclipse commands on the ethernet port to backup or restore the controller without changing any settings. While in this mode, the controller boots up into the Eclipse Status window and is not allowed to leave. In addition, the only available function key option changes to F5-End Backup Mode.

When F5-End Backup Mode is selected, the controller will reboot again and will use the original Eclipse communication settings, or any new Eclipse communication settings that were downloaded.

On EtherCAT enabled controllers, while in the backup mode, the ethernet port is configured for Eclipse access. When the backup mode ends, the ethernet port will revert to EtherCAT access.

The number of item user fields has increased from 4 to a total of 8.

In the program and status screens, the background color of the Order User Field Screen has changed from Light Grey to Dark Grey, to help distinguish if the Order fields or the Item fields are being displayed.

The number of available expanded gag words in the PLC interface has chnaged from 4 Words to 8 Words. This increases the number of expanded gags from 64 to 128. Depending on the number of presses defined, there are now a total of 139 gags available.

The Press Information diagnostic screen has been modified to correctly display the state of all 128 expanded gags.

Updated the Press Information Diagnostic screen to also show the Local Gags. Local Gags are displayed with a Red heading. The Local heading, for the count of local pressed and gags is also displayed in Red to imply that Red Headings indicate local.

Reorganized the Press/Die table in the Press Information Diagnostic screen to allow for the space used by the additional gags.

A Closed Loop Die Accelerator, for safety reasons, requires that the Manual Shear or Manual Punch input be maintaned during the entire operation. It can take some time for the die to reference and the press to fire. Requiring that the input be maintained allows the operation to be canceled at any time by lifting the button. It also prevents an operator from initiating the operation and then getting in harms way by keeping them close to the button.

SCN 4847 merged the Open Loop and CLosed Loop input tasks togther into one source file. The safety Maintain safety feature on the Manual Press input was lost during this merge. This has been resolved.

In some instances, Motors or Motor Contactors need time to stop or disengauge before switching press directions.

This feature is being implemented only on Open Loop Feed to Stop controllers for now due to the emergency nature of the need. It will be implemented on Closed Loop and Non-Stop applications in the future as we have time or need.

New Up Delay settings have been added for any Press with an Up Output. These settings will provide a delay between turning off the Press Down output and the Press Up output.

SCN 4952 - changed the timing on testing for drives to become fully enabled or disabling themselves.

This change worked on the first drive but relied on memory initialization that failed to work for subsequent drives. This resulted in an infinite loop of error messages that caused lockups and resets during memory test due to memory usage.

This same bug was present with EtherCAT and SERCOS enabled controllers. This has been resolved.

The integer edit window failed to use the Display Format setting when editing and displaying metric values.

This issue was discovered when a customer was trying to edit the Velocity at Max Analog setting.

Error 283 was added to check for multiple presses trying to control the same axis, which is not supported. When implemented, it also mistakenly reported an error if multiple tools on the same press used the same Axis, which is supported. This has now been resolved.

A number of software edits were made to eliminate compiler errors and warnings when compiling the source code with the GCC MinGW toolset. This was done in preperation of porting/sharing source code with the next generation controller that will run on an Arm processor and use the GCC compiler for ARM.

All of the edits were made with the intent that the change is equivelent in functionality or to disable inline assembly that will be relpaced with C++ equivelent code in the next gen controller.

Expanding and then closing a combo box that has no entries in its list resulted in invalid memory use.

In the windows simulation this caused a memory fault. On the XL200 hardware, the results are unpredictable.

This has been resolved.

To facilitate efficient navigation within the tree list of every screen in the XL, there is a timer that is used to delay when the right hand side of the screen is populated based on the current tree list selection. The timer is reset on every key stroke until the user pauses key strokes long enough for the timer to expire and then the right hand window is populated. If this timer did not exist, a user would have to wait a noticable amount of time for the right hand window to populate, in between each keystroke.

This timer, combined with the addition of "Are you Sure?" windows and other window prompts, opened the door for race conditions where code for an object is running with pointers on the stack that can then be deleted during the window prompts execution.

This was resolved by testing to see if the main widow has focus before acting on the timer expiration. If the window does not have focus, a flag is set that will refresh the window once it does.

The Multi-Axis Machine Setup Screen allows the machine to be configured for both Product Code and Material Code selections.

Several changes were made.

1. There must be both Product Codes and Material Codes defined or an appropriate error message will be displayed, preventing entry into the screen.

2. Both the Product Code and Material Code chains must be locked or an appropriate error message will be displayed, blocking entry to the screen.

3. If no Item is available to pre-populate the Product Code and Material Code selections, the current machine setup selections will be pre-populated.

4. Both a Product Code and Material Code selection is required before the Move button can be selected. Any attempt will be blocked by an appropriate error message.

5. At least one machine axis setup had to be defined before the Machine Setup would show complete. This is no longer a requirement.

When clearing Auxilliary Controllers memory, there is a 10 second copuntd down window that self closes. If any other Modal window displayed itself over the top t when the count down window closed, a memory crash would occur.

The Windows simulation would crash. The actual effects on the controller hardware are unpredictable.

This has been resolved by preventing the window from self closing until it has focus.

When making a language selection, a modal window is displayed with all of the possible options. The Main Window and Function Key menu was not being locked out, which allowed the main window to be replaced while the language selection was open. Once the language window was closed, the windows PC simulation would crash. The results are unpredictable on the controller hardware.

This has been resolved by locking out the Main Menu and Function keys as is standard practice in all other Modal Window instances.

The probobility of a tailout has been reduced by 80% in the Million Monkeys test. This increases the amount of time that the screens and user interface can test non-coil related activities.

The controls that populate a Grid were not properly protecting themselves from touchscreen or mouse events when a record chain is locked by Eclipse. They would try to open an edit control with a pointer that was pointing to a string rather than a record class. This resulted in a Crash on the windows PC simulation. The results are upredicatble on the actual controller hardware. This has been resolved.