A customer complained that the Stitching feature was not putting in a full cut on the last part.

The customer sent in an example order. The final Stitch parts were 36" long and the last item had six of these parts in it. The Min Auto Stitch Length was 72" and the Max Auto Stitch Length was 144". The controller stitched four parts. The last two parts combined to equal the Min Auto Stitch Length.

At the end of an Order or when the Stitch Mode setting requires a full cut, the controller will add scrap if it needs to in order to bring up the resulting blank to the Min Auto Stitch Length. The controller had a bug in the code that tries to add the scrap piece. It tried to add a short sliver that I believe got ORed together with the last stitch cut.

Two changes were made. If no scrap cut is necessary, the controller will clear the stitch gag of the last part added to the queue. If a scrap cut is necessary the controller will compare the scrap cut first to the Min Auto Stitch Trailer Length. The scrap will be no less than this length. If the Min Auto Stitch Trailer Length is zero the Minimum Part Length will be used to set the minimum scrap length part that gets added to do the full cut.

While at a customer site we were using the Closed Loop Data screen to monitor the Maximum Analog signal. While watching the screen it appeared to be changing and fluctuating incorrectly between a wide range of values for several cycles of the die before coming to a final value.

A code review showed that the values were getting reset during the entire time that the graphics screen was capturing data. It can be capturing data for up to 30 seconds which is why we were seeing it take numerous die cycles before it settled on a value.

The code was modified to only reset the analog and error max values one time at the beginning of a capture so that it would work as expected.

The controller tries to minimize the times when it needs to preserve the part queue when the controller is halted. For example, if the controller has one press and one gag. If it has no tools defined to use the gag there is no reason to maintain the queue when halting.

In the above example there is one reason the controller may need to preserve the queue and that is if there is a stitch gag defined and there are stitched parts in the part queue. The controller had a test to detect the stitched parts but it was implemented incorrectly and failed accurately detect them. This has been fixed.

We have seen a big benefit in having Velocity Feed Forward on Hydraulic Folding Machine Systems where compressible fluid and flexible hoses make closed loop tuning more difficult. Velocity Feed Forward helps provide stability by allowing the majority of the command signal to be provided through Open Loop means.

We sometimes have difficulty on customer machines that use chain driven roll-formers to position in closed loop. The chains create backlash. It is very similar to the hydraulics problem in that there is not a rigid zero backlash system that we are controlling. These chain driven systems generally also have a line encoder in part because of the inaccuracies introduced by the chain.

We already have Velocity and Acceleration Feed Forward parameters and code available in the Selective Closed Loop configuration. This change makes these settings and functionality available for Two Encoder Feed to Stop systems.

To avoid causing issues for existing customers receiving software updates, a new Setup has been added called Enable Feed Forward Settings. It defaults to No. When set to Yes it enables the two pre-existing settings;

1. Max Velocity at 10V.
2. Acceleration Feed Forward Gain.

The Max Velocity at 10V setup parameter is a Velocity Feed forward setup that relies on the accurate calculation of the velocity of the position set-points for the Closed Loop algorithm. This is a filtered velocity calculation.

There were no issues with the calculations or the filtering but there was an issue with the code that calculates the set-points and re-references the loop position between parts. The referencing and the new set-point were being calculated correctly but the old set-point, which is used to calculate the instantaneous velocity was not getting set properly between parts. This resulted in a large negative instantaneous velocity that took some time to filter out. This cause the acceleration ramp to look odd and miss-shaped.

This problem would only be noticeable when the controller is set for Feed to Stop and Selective Closed Loop. It has now been fixed.

Neither Baud Rate setting in a 6370 should ever be modified by a tech or a customer.

Techs and customers try to set the RS232 baud rate in the 6370 to match the baud rate the printer has configured. However, the XL controls both baud rates, RS232 and RS485, in the 6370. Each Print driver in the XL requires a specific RS232 setting that the XL will configure the 6370 for. That is the setting the printer MUST use.

If the settings in the 6370 are changed to match what is in the printer rather than using the setting the XL configures, the printer may work for a while. Eventually, after a power cycle or driver selection, the XL will change it back to the setting the printer is supposed to be configured for and printer communication will cease.

Two new Read Only setups have been added to display the required RS232 printer baud rate. The new setups are called Bundle RS232 Baud Rate (Read Only) and Part RS232 Baud Rate (Read Only). The printer must be configured to use the displayed setting.

As the names imply the settings are Read Only.

SCN 4561 added a "Remaining Coil is Scrap" button that is only enabled when the UART version of Eclipse is 3.59 or higher.

This change allows Eclipse to enable or disable access to this button with a setting configured in Eclipse. The button will be disabled by default unless Eclipse enables it.

Some ERP systems cannot handle scrap codes that have not been pre-defined. This change allows Eclipse to force the controller to validate Bundle Id's above 900 against the Scrap Code list it defines, which is intended to reflect all of the Scrap Codes the ERP system can handle.

If enabled, when calculating parts, the controller will test that the Scrap Code represented in a Scrap Bundle Id exists in the Scrap Code List downloaded by Eclipse. If the Scrap Code cannot be found an error message 264 will be displayed that indicates the invalid Bundle Id and Scrap Code.

SCN 2908 added the ability to track scrap that was cut by a Front Shear prior to the coil being threaded into the machine. If a Manual Front Shear operation is detected while the coil was tailed out, after the new coil loaded the controller prompts the operator with a window where the scrap can be manually entered. This window is displayed after the new coil number is entered.

This same window will now be displayed any time a new coil is loaded, after the coil number has been entered.

Any Customer who is bothered by this new Window can disable it with a new Setup Parameter. The new Setup is called Coil Pre-Thread-Up Window. It has two settings, Enabled and Disabled. It defaults to Enabled. The Setting is not visible if a Front Shear Tool is enabled. In this case the Window is still selectively displayed as described in SCN 2908.

The new setup can also be configured by Eclipse using the ME UART command.