July 7, 2009

What's New in Release 8.8

New "Balance Viewer" Tool

The new optional software tool "Balance Viewer" is available with this release and after. This tool presents the cavity fill balance as well as all other multi-cavity values in bar graph form. This makes balancing with tip temperature adjustment much easier. It also provides a diagnostic display for multi-cavity tools that leads you quickly to other problems such as cooling or pressure differences.

Contact RJG sales or support for more information.

Support for Process Stability Dashboard

This release supports the new eDART optional tool "Process Stability Dashboard". When purchased and installed that tool causes the eDART to calculate match values for all of the machine and pressure curves in the eDART. These match values may then be displayed on the EDM's Advanced System Overview (version 1.4).

Plant managers, supervisors, quality people and engineers find value having an instant view from their offices showing whether process setups are being used. This overview works both with and without cavity pressure. Furthermore the technician can see at a glance whether the process is running to template and restore the process as necessary.

Support for eDART Conx (machine-Only)

This release also is the first to support the eDART "Conx" that works with machine-only inputs (no cavity sensors). The Conx version of the eDART software is designed to be installed on a network and includes the Process Stability Dashboard mechanism. This makes its primary purpose one of ensuring that quality control procedures are being followed from the machine's point of view. These eDARTs can be upgraded to a full eDART like the eDART-flx can.

Job Setup Restructured

We found that many problems occured when technicians or mold setters accepted the job with the last material or cavity layout used and did not pay attention to the changes. The Job Setup now forces the operator to select a material or cavities if there is more than one choice. If there are no more than one choice it simply fills it in and allows him to accept the only job possible.

We also no longer allow a name to be created simpy by typing a new one in the text field. You now must push the "Create New" button if you want to truly create a new name. These are secured as before if you are using system security.

New "Save Notes" Tool

In order to encourage active addition of notes we added a new tool under "Applications" (in the Architect) called "Save Notes". This is just a simple text box into which you can type and then save a note. It avoids having non-technical users needing to learn how to navigate the summary graph and then later remember how to add a note. You might drag the Save Notes tool onto "Molds" in the Architect and then position it in a standardized place on the screen (say under Cycle Values). Then save a view to keep it there.

The note is always saved with the last full cycle completed when at the time you click "Save".

Various Improvements and Bug Fixes

• Templates are now saved with a prefix of the machine name. If you have previously saved templates they will still be used automatically as before. When you save a new template the machine name is prefixed. To transfer a mold to a new machine use the old template or systematic molding (including matching cavity pressure curves if available) to create the proper new process for that machine. Then save the new template using the same name. You now have a template for each machine that will be active whenever that mold runs in that machine.
• Refer to the documentation and video training lessons on template matching for more details. If you have a need to have this feature removed contact RJG for help.
• A part of the new template mechanism is that the fill volume used for calculating viscosity is saved with the template instead of with the process. Often a machine's transfer point needs to be adjusted differently to create a new process on the new machine. You then save a new fill volume. Once the template is saved the fill volume will be restored each time that template is run.
• Also along with the new template mechanism the template selector shows a column for machine name and a wider column for the date / time stamp.
• The Pack to Hold Transfer tool had a bug in which if you disabled a cavity pressure row in the "More" dialog AND there was a late spike in pressure (e.g. ejection) above the still working (enabled) setpoint on another row then the V->P output would go on and stay on until the start of the next cycle. This could cause a machine to transfer at the start of the cycle. The bug fix prevents disabled rows from influencing control on late pressure spikes.
• Each time you closed the "More" box the Pack to Hold Transfer also was logging in the changes_node.csv file that the rows in it had been removed even though they were still there and working. Now it only logs the removal on a true remove of a row.
• The eDART was supposed to compute a range and balance for cavity pressure peaks. But it did not do so properly if you had multiple post gate and end of cavity sensors. This bug is now fixed. It now computes peak balance and range for each group.
• Math Processing Optional Tool: If you have purchased this optional eDART tool and wanted to attach it to a machine instead of to a mold it would attach but would not allow you to edit the setup file under the machine. Now it works so that if attached to the machine the machine setup file is edited, if to a mold the mold file is edited.
• We fixed a bug that applied in cases where a mold temperature was attached (through a Lynx Quad Temp module) and the temperature signal was so well shielded and grounded that it had absolutely no noise. In these cases the temperature calculations would cause a repeating error (square root of negative number) such that the calculator would use up all the eDART's CPU time and it would appear to be locked up. This is now fixed.
• Formerly the "Effective Melt Temperature" correction factor would average the correction for all sensors at one general location (e.g. End of Cavity). This made the temperatures look different because of different heat removal rates, installation of sensors, slight location differences etc. Now the Effective Melt Temeprature is normalized for each sensor so that they all read the same.
• The Advanced System Overview would get too much data sent to it when a machine was down. This put the network into continuous receive on the ASO side. The eDART now quits sending cycle curve data to the EDM when the machine goes down.
• The template select window is no longer re-sizeable because the columns are now wide enough for dates.
• The manual File Copy utility sometimes did not show all the eDARTs to which you copy, even if they were active and connected to the network. Now it does
• V to P Transfer Tool Bug: When using temperature transfer and one of the transfer temperatures did not rise until the second cycle the control would not operate on that sensor. This would likely be seen only in a shuttle application with a controlling temperature sensor each moving base. Now the V to P Transfer tool will pick up the sensors regardless of when they become active.
• The data file storage now saves a flag for those shots that were taken while the machine was in manual mode (if wired to the sequence module). The Analzyer, release 8.9 or later, recognizes the flag and allows you to choose whether to export shots taken in manual or not.
• The eDART blocks the deletion of folders in the currently running job when the EDM (running cross-copy) instructs it to do so.
• The "Sequence Settings" "Other" tab now contains a switch for temperature sensing in LSR molds. It is named "Falling Temperature". If the mold is hot and the temperature falling this switch needs to be turned on. This forces the eDART to look for falling edges for flow front arrival time and compute peak values instead of minimums for the mold temperature.
• Sequencing: Plastic Cooling now ends at the end of mold clamped where it used to end at mold open. If you are using the value Sequence Time / Plastic Cooling for an alarm setting, read on.
• We found a situation in which the late computation of the plastic cooling summary value caused a double output on the diverter control. The cycle officially ends at the end of mold clamped. If you also have Mold Open (or Opening and Closing) then the eDART would wait to compute the plastic cooling value at the end of mold open. If the diverter had timed out and gone off by that time it would go back on again causing confusion for some downstream automation. Now, with the cooing time ending at the end of mold clamped all computations are finished at the end of the cycle.
• Now the process data are stored with all shots saved. This release adds an internal flag indicating whether the machine was in manual when the shot was taken. (Seq. Module Input / Machine in Manual on). The new Analyzer (3.9) has switches to export or compute statistics including the manual shots or ignoring them.
• The new optional tool "Double Shot Injection" is released and mantained with eDART release 8.8. This was a custom tool for a particular customer shooting two sections of a part from the same barrel on staggered shots.
• Summary Graph: Previous to this release you had to set two security settings to prevent alarm level changes: one for the Alarm Settings tool ("Change Alarm Levels") and one for the Summary Graph ("Slide Alarms"). Some customers missed the second and assumed the first would take care of it. Now if you secure the Alarm Settings level changes then they are secured in both places.
• Alpha point only: The change to the data interface to ASO that prevented sending data after a machine was down used the wrong number for "down". So the ASO would not display anything on the process viewer. Be sure to test the ASO process viewer with release 8.7.2d.
• Excessive Rejects: This tool is supposed to start when you define the Sorting Output / Excessive Rejects in Sensor Locations. This works the same as the Diverter Controls. Unfortunately, after a job stop and re-start it would not always restart itself. Now it makes a clean exit so that it starts properly.
• With no screw delay set on a machine or a machine with the Screw Run signal coming on before the motor actually starts (i.e. at the end of injection forward) then the eDART could see the screw run signal before Injection Forward went off. When that happened it assumed that it could not create a proper Machine Sequence / Hold signal. This in turn would make some values such as Sequence Time / Hold Time not get computed. Now the eDART makes a valid hold signal end at the start of screw run even if screw run comes on before injection forward ends.
• Valve Gates: If you are using timers to control the valve gates and a glitch or spike came in on Seq. Module Input / Machine in Manual that signal would shut down the timers and cause the gates to not function. Now the timers will not be stopped nor the manual mode recognized for 1/5 of a second (0.2 s). If a shorter glitch comes in it is ignored.
• Statistics: Now has a "Good Only" switch. Tick this box to show statistics only for parts that were made when the sorting output was in the "Good" position. For example, with good alarms set the CPk would show the process capability shipped to your customer. With the switch un-ticked CPk would show your internal process capability. If no sorting output is defined (no Diverter Control) then the switch is not available and the statistics computes on all parts.
• Statistics: If you have a part diverter output defined and functioning then the values in the statistics now include only those parts made when the diverter output was in the "Good" position. This includes switching to "Don't Sort", reject after down or any other function that causes the diverter output to go to not good. Thus the statistics such as CPk etc. include only those parts that were actually kept and not discarded.
• If you need to have this feature returned to the previous method (all parts included, good or not) there is a switch in the system configuration that we can help you change. Contact customer support.
• The Part Sampling tool's function to start a new data file pops up a message with the name of the file that is ending. Previously it would not actually break the old file and start a new one until you clicked "Continue". If you did not click continue the break did not happen. Now it will break right away, even if you don't click continue immediately.
• Occasionally the summary value Average Value / Stable Hold Pressure would not get computed. This usually happened when the injection forward signal goes off during hold instead of at the end of hold as it is supposed to do. Now the Average Value / Stable Hold Pressure should get computed whenever the hold pressure is stable enough to judge what the hold pressure should be.
• If you turned the screen saver off in eDART Configuration and did not re-start the eDART then the screen saver could come on again on the next cycle and get stuck on.

February 10, 2008

What's New in Release 8.7a

Maintenance Upgrade

• One of the improvements in 8.7 was supposed to make sure all sensors showed up properly on "Sensor Locations." The 8.7 work did not entirely solve that problem and in some cases made things worse. This release makes all the sensors show up properly on a job start or when you select "Sensor Locations" from the main menu. There are still some instances when plugging in a single sensor will not show all sensors. We are still working on that problem but thought it important to get this update out in the interim.
  
• The data file upload to the server (FTP upload) would sometimes not send certain files depending on the order. This has been fixed.
  
• When changing the "Diverter Disabled" settings the Diverter Controls would set its state in memory to "Don't Sort". This did not affect the operation of the diverter control but it did cause the Advanced System Overview to show "Disabled" in the "Sorting" column. This is now fixed.

 

January 8, 2008

What's New in Release 8.7

eDART Data Manager Release 1.1 Support

This release supports new features in the eDART Data Manager version 1.1. These include better control of job starts and the ability for the cross-copy function to work in poor quality networks. If you have an Insight Server running cross-copy it will still work as before, though without the new robustness of EDM 1.1.

General Improvements and Bug Fixes

• The eDART now does a better job of recognizing 100 MB Ethernet networks and switching speeds to match.
• The Template Select window has been expanded to better show the whole name and the date and time without overlap.
• When using mold id sensors mixed with molds with just plain Lynx sensors the Job Setup would get confused. If you went from a mold with just mold id to just Lynx and back it would think you had a duplicate mold id sensor. This bug is now fixed.
• If you opened the Job Setup window with the "Start Job" button on it or when a new Lynx sensor came in and then added another sensor then the other software tools would start coming up behind Sensor Locations after you accepted the job. Now they wait, as they should, until you click "Accept" on Sensor Locations
• Sometimes the Sensor Locations window would not show all sensors at job startup. This has been fixed.
• Sometimes when you clicked Accept on Sensor Locations the Job Setup window would pop up again. This should now be fixed.
• When using the Part Sampling tool and the data file reached a break point it would break the sample group into two files. This made it impossible to enter measurements in the Analyzer's Measurement Editor tool. Now if you start a sample it waits until the sample is completed before breaking the file.
• If you had security enabled on the Sequence Settings and clicked one of the tick boxes on the "Other Tab" (of Sequence Settings) without logging in then the Sequence Settings tool would crash and re-start. This is now fixed.
• In the Part Sampling tool you now cannot break the file during a sample group. The "Start New Data" file is grayed out.
• Also the ftp file copy to the eDART Data Manager (EDM) would break any files that had notes in them. Since the new EDM copies files more often than in the past this could break files into many, many pieces. Now the file breaks only when you use the manual "Upload Recent Data" function or when you specifically force it to do so in Part Sampling with "Start New Data File" (except when sampling, as above).
• When using individual sorting outputs (Sorting Output / Good Control with an id filled in) the cavity id's were case sensitive relative to the cavity sensor names. Now they are case in-sensitive. For example, if you named the id "LH" ("Left Hand") for a cavity pressure sensor but "lh" for the sorting output then the sorting output would not function. This is fixed.
• When the Cycle Values would convert the template values for "Rise" from internal Celsius to Fahrenheit it would incorrectly apply the 32 degree offset. This also happened with the "#Rng" "phantom" temperature curves.
• If you use the "Upload Recent Data" function on the QNX menu to force data to the server it now immediately overrides any ftp file copy already in progress. This ensures that the data that reach the server are guaranteed to be the most recent.
• Summary Graph: Sometimes when adding new curves the auto-scale percentage would get set to 1 which would make the curve scale so small it looked like a straight line. Now it sets the default 75% auto-scale.
• The Flow Animation tool did not make a very good moving picture if you had very low post gate sensor pressures during filling; i.e. falling below 100 psi as with polypropylene or nylon. This is now cleaned up.
• Also the flow animation calculations were running at a little bit of a low priority which could make the flow front or screw position look jerky. The calculations now get slightly more priority so that the display goes smoother.
• The "Pack to Hold Transfer" tool now logs changes in the items on the "More" button whereas before it did not. (These changes go into /e-dart/logs/changes_node.csv).
• The Diverter Controls now logs changes in the default diverter mode.
• The Statistics tool would sometimes not remove rows when you tried to remove them and then stopped the job and started it again. It might keep duplicates in its internal list so that the removed items would reappear. This is now fixed but the order may not be retained on the first job start after installing this upgrade.
• For some time the logging of hard drive wear data was being done automatically at boot-up time. We have made this a manual function at this time because some motherboards we use had a timing problem doing this at boot time. We can still get the hard drive wear data using a command.

 

August 25, 2008

What's New in Release 8.6

Flow Animation

This is a new tool found under the "Analysis" tab on the Architect. It makes a rough estimation of the flow front position in the cavity. It also demonstrates where the pressure curves come from by showing a pressure bar (manometer) and then graphs by clicking on selected buttons. It requires a minimum set of inputs of Volume, Injection Pressure and Injection Forward and Screw Run sequence inputs. Without them the Flow Animation tool will not start.

You may drag this tool onto the Architect's current mold page (or "Molds") if you want to run it all the time. If you want to just use it for teaching or demonstration you can drag it onto the Architect’s "System" tab, making it part of the main menu.

"Effective Melt Temperature"

If you have cavity temperature sensors installed the eDART now computes this new summary value. Like "Effective Viscosity" this value is a relative, not absolute; i.e. it cannot tell you the exact melt temperature but it can tell you that it has changed. The changes in melt temperature relate to the material delivered to the cavity temperature sensor, not to the barrel or hot runners.

Along with the computation the eDART has a new tool in the Architect's "Setup" tab that lets you "normalize" the Effective Melt Temperature. This tool can be dragged onto the current mold or all molds if you want it to run every time you start the job. You run the process until it is stable and then enter a temperature close to what you think the melt temperature is (typically from a purge or barrel temperature). Then, whenever that mold and material are run, the Effective Melt Temperature will report out a number that changes relative to the number you set. You may set alarms or warnings on this number to look for problems with barrel heaters, hot runner heating or process setup entry. It is also a useful tool for research.

For those of you who have been using Effective Melt Temperature thus far this version adds "Runner" and "Sprue" to the list of in mold sensors that it compensates with the Effective Melt Temperature tool.

The theory and use of this value are too detailed to cover here. A document is provided with this release that describes the use of Effective Melt Temperature in more detail.

Alarm Stability Checking

When you add alarms or warnings to the Alarm Settings tool it uses (by default) the previous 20 shots to compute the normal variation and set the levels to sensible numbers. The problem in the past has been that if the process is not stable then the calculated levels are useless. The stability calculation is developed from methods defined by the NIST to look for "random variation". The process stability goes up when there are no obvious trends or cyclic behavior over the selected number of shots.

The new stability checking feature lets you wait after setting alarms until the process stabilizes. This ensures a good estimate of what a "normal" process is and reasonable alarm levels. If you choose not to wait for the process to stabilize you can override the stability check, set the alarms immediately and take the consequences.

Various Bug Fixes and Improvements

Lynx Sensor Inputs

• The eDART-flx is no longer limited to 8 cavity sensors. It can now use as many as the Lynx ports will support.
  
• If a Lynx device, such as an OR2-D was once assigned to a mold (e.g. for valve gate outputs) then Sensor Locations would not allow that serial number to be used as a machine output. If you tried to change it to something like Control Output / V->P Transfer it would always revert back to what it was on the mold. This is now fixed.
  
• Sometimes sensors would be missing from the Sensor Locations page at job startup. A bug fix forces all of them to now appear properly.
  

Values Computation

• A new summary value type called Recovery Integral has been created. This value is computed for all cavity pressures and for Volume. The Recovery Integral for shot volume is useful for detecting changes in the way the melt is mixing. If the screw is slipping or stalling differently from shot to shot this could indicate a material problem. We have heard that this is particularly interesting in structural foam applications. The Recovery Integral / Shot Volume value is delayed one shot. This is because the value computed applies to the next part. So any alarms set on it will reject the parts that had material improperly mixed.

  The Recovery Integral for cavity pressures is useful for detecting changes in late cycle activity such as gas pressure.

• The summary value "Range" when used for the change in temperature from minimum to peak has been renamed. It now is called "Rise". The word "Range" was getting confused with statistics.
  
• The Statistics tool now calculates and displays the same stability number used by the Alarm Settings stability check.
  

Sequencing Improvements

• On some machines that supply the eDART with both an injection forward and 1st stage sequence signals and the injection forward signal would "bounce" at the end of injection the eDART would sometimes interpret the bounce as another start of fill. This was due to a "race" condition between first stage and injection forward at the start of the cycle. Depending on which came on first it would not ignore the bounce at the end.

  Note that if your machine does not create a second injection forward signal (off-on-off or "bounce") at the end of injection then this does not affect your system.

• If the Volume input was not active by the time the Sequence Settings started then the "Fill" tab on the Sequence Settings would be blank. This is now fixed.
  
• If you had the V to P Transfer tool running and did not turn any controls on or did not define a control output then the eDART would not create a Machine Sequence / Hold signal. This in turn prevented the summary values for hold pressure from being computed. It also prevented the Flow Animation tool from running. This is now fixed.
  
• On machines that do not have Seq. Module Input for injection forward but only for 1st and 2nd stage the eDART figures that Machine Sequence / Injection Forward should go on at the start of 1st stage and off at the end of 2nd stage. However many of these machines, when you make a short shot by turning off hold time the 2nd stage signal does not come on at all. So the eDART would get confused since it could not find the end of injection.

  This "bug" has been fixed by having the eDART notice that 2nd stage is defined by Sensor Locations, that 1st stage went off and that 2nd stage did not go on. The second stage must go on within the sensor update interval (determined by the number of sensors on the Lynx port) plus 5 mS. If the second stage signal is does not go on within that time interval then the eDART now assumes that the end of 1st stage is the end of Machine Sequence / Injection Forward.

Valve Gate Control

• Valve Gate bug fix: During purges the volume used to do temperature + volume could get confused and turn the gate off immediately after turning it on.
  
• Valve Gate bug fix: If you changed the temperature sensor used for controlling the gate the valve gate control would go on using the old one.
  
• Valve Gate Improvement: If temperature + volume = 0 and the screw went backwards the temperature + volume control would not fire even if the temperature came up later. Now it will.
  
• Valve Gate Improvement: The Open on Temperature + volume function will now only work during injection forward as it does for other open controls.
  
• Valve Gate bug fix: If there were spaces in the sensor id's then the "Open on Close" selections would not work properly.
  
• Valve Gate bug fix: Sumitmomo machines do not have a "hard" injection forward signal (Seq. Module Input / Injection Forward ). So the valve gate control would not operate the gates if the machine was in manual during injection when trying to purge through the manifold.
  
• Valve Gate Improvement: If the eDART does not have a stroke / volume sensor hooked up the temperature + volume control will still work. It will just assume the +volume = 0 and fire the control as soon as the flow front reaches the sensor.
  

General

• The "eDART Configuration" tool (QNX menu) now lets you turn on the "virtual keyboard". This allows you to operate the eDART with a mouse only and no keyboard, even without a touch screen. Formerly you could only get the virtual keyboard if you enabled the touch screen interface.
  
• The "More" button on the "Velocity to Pressure Transfer" tool now allows you to select temperature sensors. When you pick a temperature sensor the number you put in for a setpoint is the amount of screw travel (in volume) that you want the screw to move after the flow front contacts the sensor (2 degree temperature rise).
  
• The Cycle Graph now bases its auto-scale on groups of units by kind (e.g. mold temperature).
  
• In Cycle Graph and other tools (e.g. Alarm Settings) the temperature values were rounding off to the nearest 10 degrees instead of 0.1 degree. This is fixed.
  
• The Summary Graph now logs the change in the logs/changes_node.csv file whenever anyone edits a note. Thus if someone makes a change in a prior note and security is enabled he is forced to log in and the record of the change is kept.
  
• Temperature signals can now be output to analog output modules to follow the temperatures.
  
• There is now a summary variable computed for clamp pressure if you use it. If you define in Sensor Locations Hydraulic Pressure / Clamp the summary value Peak / Hydraulic Clamp will now be computed. Customers have found this useful for detecting problems with clamps not building full tonnage because of cold oil or other causes.
  
• Sets the security system to default to being turned off on eDART-flx units.
  
• If you removed a mold with a Mold Id on it and put in one with just Lynx sensors the Job Setup would not accept the new Lynx sensor as an identifier. Now it checks to see if the previous Mold Id sensor has gone away. If it has then it accepts the new Lynx sensor or sensors as an identifiers and changes the job. If the Mold Id is still active it overrides any Lynx sensors also attached to the mold.
  
• The "Reset Mouse and Keyboard" function (QNX menu) now works from Phindows. Formerly you had to use a keyboard or mouse at the eDART.
  
• Certain system configuration settings could not be set because of the way brackets and comments got parsed in the configuration files.
  
• The FTP functions in the eDART and some other hidden interfaces now support new functionality included in the eDART Data Manager.
  
• The pop-up message asking you to accept new files from the cross-copy server (eDART Data Manager) has now been skipped. If there are new files it simply brings them to the eDART. There was no way to skip the new files anyway so the message was a nuisance.
  
• In some instances when using Mold Id sensors the Job Setup would blank out the machine fields even though there was an identifying machine sensor (ID-7 sequence module) properly detected and listed. This has now been fixed.
  
• Sometimes particularly noisy data would have missing cycle curves stored in the analyzer (Phlat File) data. This was because the eDART did not reserve quite enough memory for curves that would not compress well. We have increased this value by 36%. This should handle most data. Please contact RJG if the symptom still occurs after installing this release.
  
• Older Insight Pro Systems may not have picked up a job change when the job was changed automatically by the changing of the Lynx sensors when molds were switched. Now the eDART sends stop and start signals to the Insightpro software.

January 17, 2008

What's New in Release 8.5

Auto-Set Fill Volume

The Sequence Settings now automatically sets the fill volume if you have a stroke encoder. It looks for the first dramatic decrease in speed. After 20 stable shots it selects the volume at that slow-down point and sets it as the fill volume. This stabilizes your viscosity, fill flow rate and speed and other summary values. In the past most people forgot to set the fill volume and so the eDART would use the injection pressure peak for end of fill. This caused irregular summary values if there was a cold-slug peak or a high peak during pack.

This new function is mostly a buried function. You do not need to do anything. If you choose force the function to run again you can open "Sequence Settings", select the "Fill" tab and press the "Auto Set Fill" button. This starts the process again and, after 20 shots, will override your current fill volume settings. When the auto-set process is complete the button pops up. You can stop it any time by clicking the "Auto Set" button so that it pops up.

A PDF document is provided on the Release 8.5 disk that explains the "Auto Set Fill Volume" in detail.

Excessive Rejects Tool

"Excessive Rejects" is included as a standard tool in this release on full eDARTs (not flx). This tool starts itself when you define an OR2 output as Sorting Output / Excessive Rejects in Sensor Locations. Then it counts shots rejected by the part diverter and turns that output on if more than X parts are rejected in Y shots.

The simplified users manual is included on the upgrade CD.

Extended Cycle Time

"Extended Cycle" optional software tool for eDARTs is now availaible as an optional tool for full eDARTs. This tool can be put on the main menu or on "Molds" to start on every mold. It allows you to define each jobs maximum cycle time between the standard 4 minute 20 second or maximum 35 minute time. The longer cycle time jobs will save fewer samples (100 / second max) but can store up to 35 minutes per cycle.

Contact RJG for more information on purchansing this tool.

Various Bug Fixes and Improvements

• Summary Graph Alarm Lines: These now move along with the sliders when you click and drag them. Thus you can see how close to the data you are moving the alarm line. As before, when you let go of the slider the Summary Graph asks if you are sure you want to set the alarm as you have moved it and, if not, puts it back where it was.
• The Cycle Graph now has security (password request) on saving templates on the eDART flx
• The "Negative Cushion Bug" has been fixed. To wit: if the stroke sensor was installed in the reverse direction (which is quite legal) and the eDART detected and inverted the stroke (to make it positive) then the cushion computation was in error. It would go in the opposite direction with changes in shot size and would also sometimes show a negative cushion. This has now been fixed for stroke mounted in both directions.
• The above fix also includes the addition of the summary value Cushion / Shot Volume. This is the "normalized" cushion. If you are using cushion for an alarm value (e.g. when there is no cavity pressure) then using this value instead of Shot Stroke makes the value work the same when the mold runs in different machines regardless of screw diameter.
• When using the "Part Sampling" tool for correlation studies the tool would not save the group note in the data until after the next cycle after the last part taken out of the press for the group. Customers would tend to start typing the new group name before the note was saved so that the previously completed group would have the name of the new group instead. This has been fixed.
• When an application like "Part Sampling" or the sequencer's new Auto-Set Fill Volume function saved a note the Summary Graph was not getting informed about it so it would not put the little note indicator on the bottom of the graph. Now it does so.
• The Sensor Locations improvment in 8.4 would override Seq. Module Input #1 with "Injection Forward". This would make the Sumitomo Oscillograph Interface unit no longer read the correct sequence input which should be "First Stage". Now any "Seq. Module Input" signal that comes in locked (grey) will not be overridden by the defaults.
• The eDART flx now allows networking as an installable option.
• If you are using temperature values there are now summary values for "Process Time" for each temperature sensor. This is the time from the start of fill until the flow front reaches the sensor (sudden temperature rise). This may be the best value to set alarms for detecting short shots with temperature sensors. If the temperature does not rise (the flow front does not reach the sensor) then the value will not be computed. It will show up as "Suspect" in the Alarm Settings, for example.
• Remote User Active now pops up in upper right hand corner of screen instead of upper left so as not to obscure controls on graph menus.
• Revised the "Sequence Settings" "Other" tab to make the wording for end of integration simpler to understand.
• If the sequencer is computing end of fill and start of hold from a set volume (set by cursor or other means) and the screw never reaches the set volume, then the sequencer will fall back to computing end of fill from injection pressure peak. However, in previous releases it would show fill ending but not show hold starting. This would cause the system to fail to compute an Average Value/Hold Pressure or Average Value/Stable Hold Pressure. Now if the screws does not reach the set fill volume it will still not compute the hold numbers on the first shot since it must determine that there has been a change. Then, after that point it will begin computing hold again.
• Also, on the "Sequence Settings" "Other" tab we returned the default integration limit to the end of screw run for new molds. We found that using Mold Clamped could and often did lead to a "race" condition with the robot or other automation equipment. This way at least the race condition will not happen when first starting up. If you wish to change to a fixed integration limit (or mold clamped) you may still do so by changing the settings on the "Sequence Settings" "Other" tab.
• The Valve Gate control forces V to P Transfer to start. However if V to P is supposed to be minimized in the view it would sometimes pop up anyway (over the top of other things). This was generally only a nusiance and mostly on the OEM units we build for Incoe's Gate Sequence Control.
  If you had the eDART set up for direct file transfer to windows then the "eDART Configuration" would lock up when you clicked "Save". This is now fixed.
• On the Cycle Graph an invalid sensor (disconnected or re-named) will now display a question mark for the cursor value.
• When the Alarm Settings tool first started and it had alarm rows on it already but the eDART had never yet created the data values (e.g. peak) then the units column was blank. This caused the level value to be shown in the internal units (e.g. psi) instead of the current units, such as bar. Now when the system first starts it looks up the expected units type and converts the level value even before the press starts running and the data is created.
• The eDART's program that copies files to the server using FTP had a bug: If you changed the job by stopping and changing material but had the same number of cavities (or cavity layout) AND you did this between scheduled copying to the server then the ftp file copy program would not copy any data after that change. This bug was introduced in Release 8.0 but is now fixed.
• The Valve Gate Control in previous releases has the ability to open and close gates on temperature + volume. If you did not have a stroke sensor hooked up you could not use this feature at all. Now if you have no stroke sensor and use the temperature sensor valve gate control it will operate the gate at the temperature rise. This is like the "Temperature + volume" feature with the "+ volume" value always set to zero.
• The "V to P Transfer" tool now allows you to pick "Mold Deflection" under the "More" button. This supports the forthcoming RJG Mold Deflection sensor. Among several ideas suggested one is that you may use it as a backup to ensure the mold is not overpressurized.
• Some relative temperature units (like Average Value, Water Temp. Drop) were getting the units converted with the 32 degree offset for Fahrenheit. This is now removed for those special values.
• In the Alarm Settings tool the default values for the initial high and low alarms are now set to mean +/- 4.5 sigma instead of mean +/- 6 sigma. We found that many customers were adjusting the alarms after the fact to about 4.5 sigma anyway.
• The defaults editing (using the Architect) was always secured. Also the translation did not work in the defaults editing. These are now both fixed.
• By default the Cycle Graph cursor will now display the curve value and the template value (if a template is displayed) instead of just the curve value.
• If you were using parts sampling in the mode were you are doing regular samples (e.g. hourly) and making a new file for each group then the last sample of the group would end up in the file that started after the group. Now all samples in the group stay with the single file.
• In the previous release the pick list for sequence inputs was limited for new machines. Now it is expanded back to allow the full choice of inputs.

---

September 24, 2007

What's New in Release 8.4

Excessive Rejects Tool

The optional (purhased) tool "Excessive Rejects" has been developed to work with this release. This tool starts itself when you define Sorting Output / Excessive Rejects. Then it counts shots rejected by the part diverter and turns that output on if more than X parts are rejected in Y shots.

Contact RJG for more information on purchansing this tool.

Support for the eDART "flx" Added

These are changes that apply to the flx eDART for its reduced size and functionality. Included with these changes is the removal of some extra menus and also help links that are no longer active.

Support for Intouch Added

For customers who purchase the Intouch production system this release now supports an optional tool called "Intouch Support". This tool installs the virtual keypad that is used on the eDART to enter information into the Intouch system.

Various Bug Fixes and Improvements

• Sensor Locations Improvements:
  • The first time a customer sets up an eDART on a machine the Sensor Locations screen just shows the most important sequence signals on their respective terminals. If other signals are available they may be selected from a shortened list. After the first setup any signal may be chosen at a future date.
  • Also the multiple warning messages that you have changed sequence signals have been removed. Only one message is shown when you change a sequence signal (and only if the job is already running). If you accept the change then the sensor locations are accepted immediately (unless something is not filled in).
• eDART Configuration Re-Boot Removed
  For most changes on the eDART Configuration window (from the QNX button menu) you no-longer need to restart the eDART. If you are making configuration changes from a remote connection (using Pindows) and change a network setting the remote connection will be closed. Only the lower upper right and lower left panes of the eDART Configuration still require a reboot.
• Video Montitor Screen Sizing
  The width of the video image would sometimes slide off to one side or the other (or both) on some video montiors; particularly LCD types. Often pressing the monitor's "Auto" button would fix it. But this was a nuisance and did not work on some monitors. So this version sets the eDART to put out video timing that is the most compatable with the most montitors (judging from experience with the previous eDART).
• Touchscreen Virtual Keyapad Fix
  The touchscreen keypad could stop working if Phindows was being used to view the eDART at the same time as a touchscreen was being used. This is now fixed.
• Touchscreen Calibration
  Some customers tried to calibrate touchscreens while using Phindows. Now the eDART prevents this by closing Phindows when you try to calibrate a touchscreen. This fix also improves the reliability of touchscreen calibration.
• Summary Graph Notes
  The Summary Graph now allows notes to be added when zoomed in.
• FTP to Server Fixes
  In some erratic network installations the FTP on the eDART could create two directories on the server as named in the path on "eDART Configuration" (one above the other). This is now fixed.
  
  Also some FTP servers send unusual codes to the FTP file copy program. If this happens the program will now stop (instead of hanging up) until the next secheduled FTP file copy time.
  
• We uncovered and fixed a potential sequence of events that could prevent data storage. We are unaware of this occuring outside of our own lab at this time. It is fixed, nevertheless, just in case.

---

 

August 10, 2007

What's New in Release 8.3

Added Hidden System Items

This release adds some updates to the QNX system (including the debugger) that did not get distributed to some eDARTs with the 8.2 upgrade. It mostly affects the newer, faster eDARTs and therefore should not be noticable on the original eDARTs.

---

 

July 19, 2007

What's New in Release 8.2

Cycle Counter Added

The Cycle Counter is a simple new tool under "Production" in the Architect. It shows total cycles and reject cycles since the job started. It is a replacement for the former "Production Monitor" tool that was removed for non-Insightpro legacy system users.

Diverter Controls

The dialog for diverter timing has been simplified. Now you can simply choose whether to hold the output until the next alarm or not. If not then you can set a time to hold the alarm on.
This release also cleans up certain issues with limited sequence inputs and the reject on down function.

Various Bug Fixes (and small improvements)

• When taking data if you stop the machine and then enter a note at the last point on the Summary Graph the eDART will not take data again when the machine starts up. This will continue until you stop the job and start again. This problem has been fixed in this release.
• On eDARTs that are not on a network or do not have a time server you would double click the time in the lower right corner to set it. However it would not let you set the date nor would it set the battery backed-up clock. Now the new clock setter (accessed in the same way) does both of these.
• When working with Lynx Quad Temperature modules (or any module that can go invalid) and the data was invalid (e.g. broken thermocouple wire) when the job started you could not add the curve to the Cycle Graph later. Now you can.
• If you have downloaded new software to the eDART from Windows using the eDART Installer and that software is waiting to be installed the eDART now checks for the new software when you stop the job. If it finds new software it asks if you want to restart to install it and does so if you select "Yes". This allows you to update all eDARTs in the plant and then wait until a convenient time to do the software install.
• The Cycle Values and cursor on the cycle graph would add 32 degrees when showing differences in deg. F. This is now cleaned up to show a straight temperature difference.
• The units of precision are set more sensibly for temperature and percent: 1 digit to the right of the decimal point for temperatures and 2 for percent.
• The Statistics tool adds the "Max - Min" statistic so you that can study the range of a value over a time period.
• On the Summary Graph the cursor now snaps to the latest displayed value when you click in the whitespace area.
• On the Summary Graph the cursor will now track with the data point selected as new data comes in and is scrolled across the display. This makes it much easier to add a note at a specific shot time.
• Some other items on the Summary Graph are cleaned up with the alarm indicator sliders sliding off with more curves displayed etc.
• The ftp file copy sometimes got stuck copying to certain kinds of FTP file servers (other than the RJG server; e.g. "Proftp" ). This is now fixed.
• If you used the "Setup Notes" tool in versions prior to Release 7.11 then Release 7.11 or later of the eDART software would convert the older binary form of the notes file (Setup_Notes.cfg) to a text file that can be edited with any computer. However, if you did not select a material at job setup then then the configuration text would be lost. This is now fixed. It should only apply to jobs that were created with no material and have not been run since Release 7.11.
• We had a nuisance bug in which if you maximized a window (e.g. Cycle Graph), then used some other tool from the toolbar and then returned to the first tool you could not restore it to its original size by clicking the square button at the upper right. This is now fixed.
• When the cross-copy service is running on the server it prevents you from starting duplicate jobs. However if a new mold is hung in a machine and the cable is plugged in and it self identifies and you accept the new job... then the eDART would not tell the server that the old job is released and can be run somewhere else. Now as soon as you click "Accept" the old job is free to run again.
• On Incoe Gate Sequence Controller eDARTs the User Interface (on the QNX menu) would let you delete molds as needed because there is no Architect to support that function. However it would not refresh the molds list after you deleted or created one. Now it does.
• The Alarm Settings tool would immediately switch the "Sort" or "Alert" switches when you clicked the boxes. Then, if you changed your mind in the pop-up question it would switch it back. But during the time it was switched the function would change before you intended it to. Now it only switches if you answer "Yes" to the confirmation question.
• Job Setup now alows some longer names because these may be needed for the Simon systems that have the only choice being the mold name (no material etc.)
• The Architect no longer displays the "logs", "tmp" and "update" directories on its "System" tab because these make no sense in that context.
• With a server running the cross-copy service and have very bad network communications the eDART could possibly use up its memory trying to make connections over and over. This release prevents that from happening. This fix also makes it likely that if you have TWO cross-copy servers running at the same time the eDARTs will not talk to either one of them. The cross-copy function is not defined for two cross-copy servers and you should never have two of them running anyway so this should not be a problem.
• The User Interface (on the QNX menu) would not do directory listings properly if the names had spaces in them. Now this is fixed.
  

---

April 19, 2007

What's New in Release 8.1

Various Bug Fixes (and small improvements)

• Support for Virtual Modem Removed
  This was an old serial networking interface that was created in the days when most laptops did not have an ethernet port. It is no longer needed.
• Restored Cavitation to Insightpro Legacy Users
  We had inadvertently removed the shift tabs and standard and actual cavities when we supplied the "Insightpro Support" optional tool. These are now restored to those users who own Insightpro Reports
• Now allows cross-copy of configurations (e.g. alarm settings, templates etc.) between different software versions. Previously this was blocked because we initially allowed cross-copy of eDART software. Since this is no longer done we can open up the cross-copy of configurations between verisons.
• On flx-APC units we hid the adjustments for time from mold clamped to injection forward and injection forward on time. This made it impossible to use the flx-APC units with just a Mold Clamped sequence input. The sequence settings now displays the timing items necessary to use just Mold Clamped as a sequence input.
• Updated the QNX operating system to the latest release