translation missing: en, page, software_updates, history, title

Now at each flush the eDART writes only up through the latest complete cycle. Then if the job goes down or the job is stopped the eDART writes the all cycles, including the last one with an end point on each curve.

A new feature has been added to the Diverter Controls to cause it to reject a number of parts when the press starts up after it has been down. You can find this on the menu under the "Diverter Settings..." button. The selection is called "Set Reject After Down...".

This feature when you have material that tends to burn or otherwise degrade in the barrel or hot runner system if the machine goes down. Even with cavity pressure sensors it is often impossible to detect the fact that the material has degraded. Thus you could allow parts to be counted as good even though they were streaked or structurally unsound. Now you can force the Diverter Control to discard a certain number of parts after each time the machine goes down.

• Minutes on FTP File Copy:You can now set both the hour and minute at which the FTP file copy sends data to the server. This can be used to stagger the up-copy to the server so that only one eDART is sending data at a time each day. If multiple eDARTs send to the server the files become excessively fragmented on the server.
• The eDART configuration no longer shows the fields "Laptop e-PAK" and "Windows Browse Master". These two functions are no longer supported.


• The eDART configuration now shows the serial number of the eDART box itself if it has been set.


• The eDART configuration now lists some time half-hour time zones. This makes a complete list of all world time zones.

• The Down Time and Scrap Entry tools no longer exist on the Production tab in Architect.
• The Production Monitor and the Production State tools have also been removed from the Production tab in Architect.
• The Insightpro "Data Module" will no longer connect to the eDARTs. Customers that purchased Insightpro reporting will still be supported through the installation of an optional tool.
• The "Production Overview" (a.k.a. "System Overview") on the eDART has been removed.
• Support for the obsolete "Litehammer" interface has been removed (this was not directly a part of Insightpro but a custom software interface).

• We have updated the rules for daylight savings time in the United States to match the changes in the law for the year 2007


• A bug could sometimes allow several screen saver programs to start running, even if it looked like just one. This has been corrected.


• The eDART now saves all data. Previously when it saw that the machine was in manual mode it would not save data. While this eliminates garbage data during setups it can also hide interesting things like purge pressure profiles. Furthermore if the "Seq. Module Input", "Machine in Manual" was wired to the wrong signal or left to pick up noise then the eDART was at risk for not saving any data at all.

  This is a configuration switch that we have turned on with this release.



• Forced FTP File Copy: Many customers have the eDART Configuration set to copy files to the server on a regular schedule. However it is often useful to get the data up to the server immediately. Cases in point are when you have just finished a designed experiment and wish to analyze the data or when you have solved a problem and want to study earlier cycles to determine what went wrong.

  We have added a new selection to the QNX menu titled "Upload Recent Data (FTP)". This immediately starts the copy of new data files up to the server beginning with whatever is new since the last copy. If the copy was already in progress it breaks off the current copy and starts again. This should take little time since it will not even be a whole day's worth of data if the ftp copy schedule was already set up.



• FTP File Copy Fix: In previous releases the ftp copy program would get "stuck" on certain files that it might be trying to send to the server that already exist there. These are not even write protected files. But for some reason the server will not accept an overwrite. Now the ftp file copy program handles the ftp communications better and it also deletes the file on the server before sending the replacement up.


• During installs of new eDART software the program that modifies configuration files would get stuck if it tried to modify the last line of a file. This would prevent the install from completing and "lock" up the install.


• If you were using the "Sequence Settings" integration limit (on the "Other" tab) and the integration limit ended past the end of the cycle the sequencer would not turn it off. This meant that values computed at the end of the integrtation limit (e.g. cycle integrals) would be computed late and applied to the next shot for part sorting, data storage etc. This was not a problem if the integration limit time ended before the end of the cycle (end of cycle is end of mold clamped, or if there is no mold clamped, end of screw run.).


• Screw Run after Mold Clamped: Some machines (Husky among them) can have a shutoff nozzle that allows the screw to run after the mold has opened. If this is the case then, in eDART software versions prior to this, the values for screw run time ("Sequence Time", "Screw Run") and backpressure ("Average Value", "Backpressure") would be delayed two cycles instead of one. This is now corrected.


• Sequencer Integration Limit Before Last Mold Clamped: If you want to end integration at the end of the cycle when the mold un-clamps you can set the integration limit to 0 seconds before previous clamp open. In previous releases this was figured by recording the last mold clamped time and using it as a guess as to when the current cycle's clamp would open. With the value set to zero this is not necessary. Now that integration limit is the mold clamped time if you set the value to zero. If you set the value to some time before previous mold clamped then the software must still use the last cycle's clamped time in order to anticipate the end of the cycle.


• If you had previously clicked "Ignore all sequence inputs", stopped the job and re-started the sequencer would not ignore the fact that the hard wired "Injection Foward" sequence input was "off". So it would assume that Injection Forward was active, set the "off quality" to 5 (hard wired) and then never cycle again because it was ignoring while waiting for the signal to go off. Now if you leave the "Ignore all sequence inputs" checked the sequencer starts with the off quality set to zero an will continue start cycling as soon as it picks up some other sequence method (e.g. stroke or injection pressure).


• If the eDART has only "Injection Forward" for a sequence module input and do not have a stroke or injection pressure sensor then the sequencer logic would turn "Machine Sequence", "Hold" on when Injection Forward went off. Now it does not operate "Machine Sequence", "Hold" if there is no way to know where machine fill ended.


• In the Valve Gate control the open-gate-on-close function would not work if you put spaces in the identifier for the valve gates control. Now the spaces will not matter and you can use them in valve gate Ids.


• In the Valve Gate control the setpoint safety feature that prevents you from changing the numbers by too much would not work on the "Open on Volume" boxes on the main page. Now it does.


• In the previous 7.13.x beta series the removal of the Insightpro support included the removal of the "Last Shift" and "Current Shift" tabs on the Production Monitor tool. This removal caused a bug on the "Since Reset" tab in that when you tried to reset the counts the tool would crash (and re-start). This is now fixed.


• If you ran the Valve Gate control with Basic 3-Stage control it the Valve Gate control would try to start Pack to Hold Transfer over and over again. P->H Transfer is not allowed to run if Basic 3-Stage is already running. Now the Valve Gate Control starts the appropriate machine control based on the valve or transfer you have attached.


• When you tried to move the Production State tool around the outline was larger than the window. When you released the mouse it would "jump" unexpectedly. Now fixed.


• It was possible on rare occasions to get the system stuck on startup if you quickly re-opened the Sensor Locations setup and changed sequencing before the system was fully up and running. A time-out has been added to prevent this.


• Support for Insight Alert: The latest version of Insight Alert displays the alarm status based on what the "Alarm Settings" tool reports, regardless of whether you have a part diverter or not. This requires a small, low level software change to the eDART that is included in this release.


• The Alarm Settings tool would not rename rows with the word "Reject" to "Alarm" for sensors that were not active at the time the Alarm Settings tool started with an older configuration file. Now it does.


• We added some additional logging to the process of installation for internal RJG support purposes.


• The Diverter Control was modified as of release 7.12 to reject parts whenever the machine was in manual mode. ("Seq. Module Input", "Machine in Manual"). However if you switch to manual during the later part of the cycle (e.g. when if the screw is getting starved for material) the part in the mold is still good. Now the Diverter Control only rejects in manual if it discovers that the machine is in manual at the start of the cycle or sometime during injection forward. If the machine goes to manual after injection forward the Diverter Control just operates on the alarm settings for the part that is finishing.


• The list of production values available in the Active-X control and the Raw Data Viewer has been restricted to those that do not involve shifts, parts counts and cavitation. All parts counting work is done by the Intouch system or, for legacy users, Insightpro. The counts involving cycles and the times for the whole job are still visible to third party software through the Active-X control and in the Raw Data Viewer.


• If support uses a separate Phindows sesson (not to the eDART's main console) then the "Remote User Active" display will not pop up on the main console to disturb the user.


• If the current language is still in english then the curve process help appears on the Cycle Graph. We removed this feature when we removed non-translated language text from the eDART. It is now back (English only).


• We removed some extra logs from diags.log (having to do with the queue). These were old items used in the initial development of the eDART and are no longer useful.


• In previous releases when you changed scaling in Sensor Locations, Ejector Pin Sizes or Machine size the new scale factor would not get saved in the data file. Now it does.


• RJG's Alert software and Intouch's part counting could erronously report cycles as "Suspect" when they were actually "Good" depending on how you had used the indicator and sort switches on the Alarm Settings tool. This has been fixed in this release so that it works correctly in all cases.


• The "Status", "Part Alarm Number" and "Production State Number" values were not getting stored as summary values. Now the are stored properly.
• If you start a mold on a different machine and the "Pack to Hold Transfer" control pops up with a selection of machines this is now forced to the front so that you will notice it. This is also true for "Basic 3-Stage Control"


• The "Sensor Locations" program now shows the proper serial number for analog inputs on the offset dialog box. Formerly it just showed a default number.

• Adds several internal values and changes that are used by the System Overview displays on Windows systems. This allows the Windows System Overview to show alarms and warnings whether or not you are using a part diverter to sort parts. Note that the Windows System Overview alarm and warning counts will not increase when the press is in manual mode (if wired to the sequence module).
• Adds network services to transmit real-time summary and cycle data to the data display on the Windows System Overview.
• Ensures that when a job is stopped the transmission to the server also stops. In earlier test versions this was not the case.

• Alarm Settings Texts:
  The Alarm Settings tool now displays its status texts as "Warn" or "Alarm" instead of "Suspect" and "Reject". This is because some customers are montoring alarms without actually rejecting (sorting) the parts with a part diverter (conveyor, flipper or robot). The idea is that the alarms and warnings now represent the problem rather than the destination of the parts. The Diverter Control, when in use, still shows "Reject" or "Suspect" because that is the physical destination of the actual part.
• "Counts" No Longer Listed:
  Formerly when you went to add an alarm the picker would show values that are in counts (e.g. production counts) as well as summary values. This made a confusing list of items that really are not needed for alarms. Now the list only includes summary values and not counts.
• Collapse - Expand Improved:
  Rather than recording the last collapse / expand values the Alarm Settings tool now expands to show all columns and collapses to show just the first four.

• If you were using late computed alarms (such as cycle time or mold clamped time) then and the machine was in manual or the Production State tool in "Setup" then the diverter would not stick in "Reject" at the end of mold clamped. Now the machine in manual or "Setup" state overrides the late computed alarms.
• If you were using delayed diverter outputs then switching the machine into manual (from semi or auto) would cause the diverter queue to advance. Now it does not do this except when the press cycles.

• Calculation without Screw Run: On extruder trials we found that without the screw run signal the energy used per cycle would not be calculated. Thus the eDART would not produce energy summary values when using the "One Minute Cycle" tool in continuous processes like extruders. Now it does.
• Scaling: Formerly if the result came out in watt-seconds it would get the units set to "watt-s". Now any results in "watt-s" get automatically converted to "Joules" (which are watt-s).

• The delayed disk write mechanism could sometimes start multiple clock_sync processes. A "safety valve" has been added to prevent this.
• Job Start "Accept" no longer pops up the message about writing to disk.
• The Job Stop mechanism now ensures that all un-used shared memory is cleaned up, especially the extra spaces left by prematurely stopping the Math Processing tool.
• Fixed problem where Alarms added from the Alarms Settings applications would not display on the Summary Graph when the Sort option was turned on. This was due to the Summary Graph turning off the alarm line display flag if the summary alarm point was invalid, which it is if sorting is turned off.
• Part Sampling was not dealing with security properly. Now fixed.
• The cycle buffer would not always pick up the minimum if the minimum happned to be the first sample in the cycle. Now it does so. This would have an almost negligible effect on most computed minimums. It only appeared a an issue when working with mold / melt temperature computations.
• If the Summary Graph was zoomed it would shift the zoom area and appear to unzoom.
• The Summary Graph cursor would jump back on next shot when zoomed to the end of the data. It will not scroll forward when zoomed.
• The "Peak", "Auxilliary Power" variable was not being computed if you had an Auxilliary Power input defined.
• If the Pack to Hold Transfer (V to P Transfer) tool was vertically squished (to display just one item) it would flip its scrollbar down when you clicked on things in it. Now it stays where it was.
• The Part Sampling tool would store the group count "Start Group: N Shots" translated into the current langugage. This would cause the Windows Analyzer's Part Measurement tool to not recognize groups for data entry. For the present the "Start Group" and "Shots" texts created by the Part Sampling tool are always in English.
• If you used the "Reject Parts" switch on the Part Sampling tool but had no valid alarms (or no alarm settings) then Diverter Control would not reject the parts if your diverter defaults were set to Good or Suspect. Now it rejects them regardless.
• If the serial number of the eDART has been set in the software then it shows up on the Software Version tool from the main menu.
• The sensor locations would sometimes block you from hitting Accept when you had a light tree hooked up. This seemed to be most common when the "Virtual Light Tree" was running. It is now fixed.
• If you started Architect, minimized it and then tried to start it while minimized it would lock up and never appear (until a restart).
• The Architect's "Save" and "Options" buttons are now removed since. The default is now "Auto Save"; i.e. when you drag something on to the Architect the application starts immediately and the attaching to the mold (or machine) is done at that time.
• The Lynx driver would occasionally crash and re-start itself. Now fixed.

On Arburg machines where you may have a hydraulic sensor for both injection and braking the eDART would not let you click "Accept" after re-opening Sensor Locations. You had to stop the job (or the eDART) in order to get out of Sensor Locations.

Also, the Sensor Locations would keep popping up when you tried to change injection pressure back and forth between direct (Plastic Pressure) and indirect (Hydraulic pressure x intensification ratio). This is also fixed.

This release adds the new "More" button to the V->P transfer tool. This allows you to set setpoints on more than one sensor to transfer the machine on whichever sensor arrives at its setpoint first. One method of using this feature is when you have a post gate and end of cavity sensor. You may want to control at the post gate sensor but set a second setpoint under the "More" button to prevent the end of cavity from getting overpressurized. Another technique in mult-cavity applications is to set the pressure control on the main V->P screen to the "#Avg" sensor to transfer off the average. To prevent overpressurization of one of the cavities you can also set additional setpoints under the "More" button to one or more "#High" sensors to prevent overpressurization of any cavity.

This new feature also helpful in shuttle systems where the different bases have different setpoints (was already available to shuttle users).

• When the Summary Graph was refreshed (re-started) it would not re-load notes.
• Notes did not always adjust correctly when zoomed.
• Internally Summary Graph was checking for new notes on each cycle. This extra load has been removed.

Inject Enable Output

• We normally recommend that you set up a "Control Output", "Inject Enable" (OR2-D output) that stays closed as long as the sensors and eDART are operating correctly. In the past, when a control sensor failed the output would go off and you would need to enable the injection manually to continue.

  Now the Inject Enable output will go back on and the message will go away automatically if the sensor recovers. This can be useful in situations where high ejection force triggers an over range but the sensor is still functioning (though it will die eventually).

  If the sensor truly fails or is un-plugged from the Lynx cable then you can re-enable injection by disabling control for that particular sensor. This allows production to continue with an alternate control strategy if required.

  Furthermore we have added new checks for the proper operation of sequence inputs and control outputs. If any of these fail during control the Inject Enable will go off to stop injection.

Mold Id Sensor Has Priority

• If a Lynx Mold Id Sensor is found its serial number will be the preferred identifier for the mold. Thus if you take Lynx cavity pressure sensors out of one mold and put them in another and both molds have a mold id sensor on them then the molds will still be properly identified. Essentially the Mold Id sensor overrides any identification provided by the Lynx cavity pressure sensors themselves.

Late Alarms Computed

• The eDART now allows you to sort parts on three late "Sequence Time" alarms that it previously did not: Cycle Time, Mold Clamped and Plastic Cooling.

  In the past all alarms were computed at the end of screw run. The eDART cannot know the cycle time until the end of the cycle (usually when the mold-unclamps, Mold Clamped OFF). The original design was for sorting equipment that had to know the state of the parts before the mold opened. Now the eDART runs through the above 3 alarm checks after the end of the cycle. If any of the alarm limits are exceeded then it switches the diverter from its former setting to the new alarm at that late time.

  Note: The sorting equipment must be able to handle these late alarms or it will put the parts in the wrong bin. Conveyors generally work if the part has not fallen off between the end of screw run and the end of the cycle. However part flippers and some robots will be problematic if they cannot react before the parts are released.

Touchscreen Keyboard Pop-Up

Diverter Logic Simplified

• The Diverter Control now forces its output to "Reject" when the machine is in manual mode ("Seq. Module Input", "Machine in Manual" is turned on).
• It also forces its output to "Reject" if the Production State tool is in use and is set to "Setup".
• If neither of the above are true then the Diverter Output goes to its default setting (as set from its menu) whenever the "Don't Sort" button is pressed, no Alarm Settings are running or no alarms are set up and set to sort.
• If no Diverter Control is running (i.e. there are no "Sorting Outputs" defined on OR2-D modules) then the parts are counted as "Good" except if the machine is in manual or the Production State tool is in "Setup."
• The maximum diverter times have been increased for extra long conveyors.

Setup Notes

• Formerly the Setup Notes tool stored its text in ".cfg" files that are in a binary format. Now it stores the notes in ".txt" files. These files can be edited with Windows Notepad or Wordpad from a network (e.g. on the server). Any old, binary ".cfg" files will automatically get converted to ".txt" files whenever the particular job is run. Once this happens across the system the old "Setup_Notes.cfg" files can be removed from the system.

Ejector Pin Sizes

• Now forces the units to pop up if you type a number into the area or diameter boxes. This encourages you to pick the correct units instead of using the units selection first (which changes the value) and then having to remember to go back and re-set the value.
• Changing the units on diameter will change the units on area and vice versa.
• Occasionally would "miss" a newly created sensor. Now it should catch all of them.

Data Rollover Message

• The message no longer reports percentages. These are file-dependent and were getting rather confusing. Instead it simply reports the disk "nearly" full or "full" and explains what it will do next.

Cycle Values

• Templates can no longer be selected from the Cycle Values window. Users were getting confused about which one should be use to open the template selector or they were opening two template selectors.

Mold Open Time

• The value "Sequence Time", "Mold Open" is now saved with the next cycle. This value is commonly monitored in semi-automatic (e.g. insert molding) jobs to track the variation in the time it takes to do the manual operation. Since it is a useful tool for monitoring residence time it is reported with the next shot like the screw run time and backpressure.

Sequencing

• Sometimes the auto-invert function (velocity only) would get fooled on Sumitomo machines. We added new intelligence to prevent this from happening.
• The sequencer now automatically locks the stroke direction after it determines that 10 shots in a row have all been injecting in one direction and retracting during screw run in the other direction. This eliminates the need for the user to try to remember to tick the "Prevent auto-stroke detection" box after installation.
• The detection of screw run from "Revolution Rate", "Screw Motor" is now four times as sensitive as it was (triggers at a lower level). This affects mostly users of the Sumitomo Oscillograph interface.
• Formerly the menu line on the Cycle Graph allowing you to "Set Volume Zero at Cursor" would allow you to select it even though screw run was active. This is because it also checked to see if the fill quality "ON" was too low. Now it lets you set fill volume only if there is no screw run signal.
• Now enables injection each time you switch into or out of manual mode ("Seq. Module Input", "Machine in Manual"). This helps in the case where a sprue sticks and you don't run the screw again after switching to manual, opening the mold and cleaning it out.
• Formerly when the sequencer was detecting injection forward on forward motion of the screw but switching injection off on time (as when there is no injection pressure) then the injection forward would not be "re-armed" and the data would not cycle.
• If you clicked the "Ignore all sequence inputs..." box and then answered "No" to verification request the sequencer would actually begin ignoring all sequence inputs even though the box was not checked. To make it work again you would have to click the box and answer yes (turning the ignore on) and then click it again and answer "Yes" again to turn it off. This bug is now fixed.

Types and Locations

• Added "Runner" to the sensors that can be used for pack computations.
• Added a number of sensor locations for gas assist measurements and the summary variables for them.

Summary Values

• Now creates a "Range" for cavity temperatures. The "Range" at each sensor is the maximum minus the minimum between the start of the cycle and the time when sorting begins (usually end of screw run).

File Maintenance with User Interface

• The User Interface restores the rename, delete and copy functions for molds. This is only available on OEM units (e.g. the Incoe Gate Sequence Control) that have no network access.

Port Diagnostic on Raw Data Viewer

• Now shows the number of sensors (under "Value") and sample rate (under "Raw") for the ports without having to open the diagnostic details window.

Mold Deflection

• The auto-zeroing mechanism now adds auto-scaling for RJG's new Mold Delfection sensors. This ensures that the amount of deflection is re-calibrated to the sensor's burned in scale factor on each shot.

Other eDARTs Tool

• This tool that formerly appeared on the eDART main menu is now a part of the "User Interface" on the QNX menu. It has been used very little by most customers so was moved out of the general eDART menu.

eDART Menu

• In the situation where all the menu items appear to have been removed or the menu cannot find any menu items then the "Architect" will always appear on the menu to allow the user to rebuild.

Sequence Lights

• Now shows the manual mode input on the main panel instead of having to open the one that shows the actual sequence input wiring.

Production State Tool

• Now allows you to hide the "QC Approval" and "Pre-Approval" buttons. These are used by only a few customers. So the Production State tool starts up with both buttons hidden. If you want them displayed click the "Settings" button and select the appropriate tick boxes.

Window Positions

• The system now prevents the bottom edges of windows from falling behind the toolbar at the bottom of the screen.

Languages

• The eDART now comes with French translations.
• If translated texts are too wide to fit they automatically get left justified so that you can see at least the left side of the text.
• In former vesions some translated texts were too large to fit in some internal memory buffers of certain programs and they could crash. This is now fixed
• The translations have been bundled into a single language file for each language.

Improvements

• We added the value for the the "Open on Volume" setting as a column on the main control screen. This is because it is used so much in balancing that we found it became a nuisance to keep opening the dialogs each time to you wanted to change the volume to open.
• We have added the ability to open and close on a temperature rise plus a volume value. Since the cavity temperature at the start of injection can drift up and down over time we use a change threshold to tell the valve gate control when the hot material has reached the sensor. You can then set a volume of screw motion from that point until you want the gate to open or close.
• The "V to P Transfer" tool is required in order to get the machine to transfer from velocity to pack when all valve gates close. Some users have forgotten to add the tool to the Architect when they added the Valve Gate tool. So the Valve Gate tool now starts it automatically if the outputs to the machine exist.
• Cause of Transfer Displayed: The V-to-P Transfer tool now shows that the transfer was caused by the valve gate control if the valve gate control transfers the machine when all gates close.
• Open on Mold Open has been removed. This setting was intended for polycarbonate and could be used to pre-heat the pins by pulling them back early. However this could leave the valve gates open with the mold open which we do not want to do. To open the gates early and pre-heat the pins use "Open on Mold Clamped" and set the machine to delay before injection start.
• The previous versions of Valve Gate allowed you to supply a pressure switch input, "Pressure Switch", "Valve Power Unit". If off this would disable injection and put up a message. Now you can also supply a third-party pressure switch and wire it into the sequence module. If you name the input "Seq. Module Input", "Valve Pressure Supply" then it will behave the same as a pressure switch. If the signal goes off then the Valve Gate control will disable injection and inform the user that the power unit is not working.

Valve Gate Bug Fixes

• If the "Pressure Switch", "Valve Power Unit" input was off on startup the Valve Gate Control did not notice it. Now it does.
• If there was a non-existent pressure sensor (e.g. a sensor had been un-plugged) then the "Convert to Pressure" function would not work.
• The "Do you want to turn off all open controls? pop-up message did not allow you to answer "No". Now it does.
• In Sensor Locations if you chose "Valve Gate Output", "Enable Gates" then you could not continue because it would keep asking for an Id to be filled in. This was impossible because the Id column is greyed out for "Enable Gates". Now it only forces you to fill in the Id on "Open Gate" or "Close Gate".

Incoe GSC Only

• The Incoe GSC units now show a banner with the Incoe RJG logos on it.
• Incoe units also have the third line of the Job Setup reading "Cavity/Gates" instead of "Cavities".
• Hide Template Stuff on Cycle Graph: Dedicated valve gate controllers for OEMs (e.g. Incoe) do not have template saving and management. Formerly the template controls would still be selectable on the cycle graph. Now they are grayed out for OEM controllers.

Assignable Sorting Outputs

• The Shuttle Control now allows you to assign specific sorting outputs ("Sorting Output", "Good #x") to specific sensors. For example, a rotary table with 3 bottom halves with 8 sensors in each can have the three corresponding sensors from the three bottoms assigned to a specific OR2-D output. This way you do not need a separate OR2-D output for every sensor in the system.

  Note that you must have the Shuttle Control running for this to work. Preferably the Shuttle Control should be attached to the mold using the Architect.

Cursor on Active Curves Only

• The Cycle Graph cursor now shows values only for the "active" curves.

Proper Template Saving

• Saving a template while using shuttle will now save the inactive curves from the shots back as well as the active ones so that all templates are available.

• Now returns valve to fast speed at the start of screw run. This lets the injection valve open up so that full flow occurs during backpressure.
• Forces the V->P Transfer output to "off" at the start of screw run. In especially wired controls on old machines (mid 1980's) this signal must go off to shut off the injection forward valve. Otherwise it is trying to push against the turning screw. Even if the cavity pressure builds up past the setpoint during screw run (as sometimes happens with mold deflection) the V->P signal will not go on again.
• Now allows control to function even if "Control Output", "Inject Enable" is off. While we strongly discourage it, some customers have added a switch to override the Inject Enable contacts. This is used if they do not wish to use the eDART. But, formerly, the Basic 3-Stage Control would not transfer on any input if the Inject Enable output had gone off on a failed sensor. It assumed that the machine was stopped anyway.

  Now it will function even with Inject Enable off on a failed sensor. However you will not be able to do fill only shots if the Inject Enable is bypassed as described above.

• The "Machine Sequence", "First Stage" signal that the 3-stage control generates was going off at the end of V1 instead of the end of the speed phase (at transfer). This is corrected but had little effect on the system except as a proper record of the speed phase on the machine.

Ejector Pin / Flush Mount Switching

• In previous versions switching between sensor types could cause several internal system errrors. These should now be cleaned up. You must still stop the job when switching between sensor types.
• Related to the above was the instruction on Sensor Locations that you needed to stop the job. In the past, using several senors, you would need to stop the job for each sensor to be changed. Now when Sensor Locations stops the job on a sensor type change all of the new sensor settings are remembered and you only need to stop the job once.

Raw Data Viewer Row Limit Increase

• Formerly this was limited to just 500 lines. Now it does not have a fixed numeric limit.

Sensor Locations Synchronized with Job Setup

• Formerly when you plugged in the sensors on a new mold the Sensor Locations screen would sometimes pop up before the Job Setup window with the new (automatically identified) job. Thus any sensor changes you made on the mold might get saved with the old job if you did not accept the new job first. Now if the Job Setup screen comes up the Sensor Locations is held back until you accept the job.

Data Storage and Summary Graph "Stuck"

• Occasionally, when you opened Job Setup (Change Job Setup) and clicked "Accept" with out any changes, the Summary Graph or file storage (Phlat File Storage) would get "stuck". This is now fixed.

Summary Graph

• Summary Graph would sometimes restart if you tried to auto-scale on a graph that had warning indicators up. It no longer does this.
• When using long cycle setups it now does not have the choppy curves when the cycle graph is zoomed in. Instead it has the regular stairsteps seen with standard cycle curves.
• The maximum that can be entered on the cycle graph time axis matches the longest cycle for which the eDART is configured.
• The legend buttons can no longer be hid.
• The alarm indicators are fixed.

Statistics

• If you changed the units on "Coefficient of Variation" or "Expected Variation" then the others of that type would not change. Now they do.
• The units conversion for standard deviation on temperatures was not correct. Since standard deviations are a range the units conversion should not add or subtract 32 when switching between degrees C and degrees F.

Flow Rate Units

• Units for Flow Rate and summary values computed from it were not getting changed properly across the system. Now they are.

Phlat File Storage Re-Starting

• Occasionally and with very short cycle times the Phlat File storage application would restart itself. While this did not cause a loss of data it was a lingering bug that we were able to find and fix in this release.

Id Available on Auxiliary

• Any sensor that allows the location "Auxiliary" now allows you to name it with an Id number. Formerly the adding of Id numbers to auxiliaries was limited to certain types.

Machine Size

• When started from the eDART main menu the "Machine Size" tool used to always ask for permission to save, even if you did not change anything. Now it only asks for permission to save if you do change something.
• Now prevents an intensification ratio being set over 100 and pops up a message if it gets set that way.
• If you double clicked on the tick boxes that select whether to calculate intensification ratio or cylinder diameter then the program would enable the numberic entry for both. It was supposed to enable only the one that was not being calculated. Now it does.

Translations

• French language translations are now added into the system.
• The multi-cavity id's "# High", "# Low", "# Avg" and "# Rng" are now properly translated. All other cavity id's remain un-translated because we assume that the user has entered them in his native language.
• The AC power loss and recovery messages on UPS equiped eDARTs were not showing up translated. Now they are.

Touchscreen Calibration

• The system now simply resets the calibration on the touchscreen from the first calibration created when it was first set up. Sometimes running the calibration with Phindows or pressing the wrong parts of the screen could make it impossible to recalibrate. The only time the actual touchscreen calibration screen will come up (when selected from the QNX menu) is when you first enable a touchscreen in eDART Configuration.

V-to-P Transfer

• If you re-named or disconnected the OR2-D output that was assigned to "Control Output", "V->P Transfer" the V-to-P Transfer tool would not put up its yellow message box announcing the fact. Now it does.
• If you made the injection pressure transfer value "during pack" higher than overall then the overall value did not work. Now it does.
• The view labeled "RJG Control" brought up the V->P transfer window too small to see the lower line which is the time control. The window could still be pulled open but it was a nuisance.

Alarm Settings

• If you turn off an individual alarm's "sort" switch on Alarm Settings the alarm line would still show up on the Summary Graph. Now this line goes away when you switch sorting off and returns when you switch it on again.

Lynx Driver

• Occasionally the system would start with just one of the two Lynx ports enabled. The system had to be restarted to get the second port to run. This has been fixed.

Sequence Lights

• The drop-question mark help on the raw data lights rows did not translate the key (a wierd set of numbers and letters) into the proper help display. Now it does.

Sensor Locations

• In previous releases the eDART would prevent you from bringing up the Sensor Locations page (from the eDART main menu) until the cycle had ended. This is because doing so can foul up the Lynx network communications. However you could, once the Sensor Locations was up, click "Accept" and have it save all of the sensor changes. Unfortunately this could also confuse communications with the Lynx sensors and force them to reset. This could cause missed control or delayed sorting outputs.

Now clicking "Accept" starts the wait for end of cycle so that the changes do not get saved until the non-critical time has passed.

• When the Sensor Locations offset dialog was added it did not properly log changes to the offsets in /e-dart/logs/changes_node.csv. Now it does so.

Process Restart Memory Leak

• In cases where a process (program) was re-started many times by another process the application in charge of that ("mayor") would eat up a little bit more memory. While this was not a problem under normal circumstances it could be if a program was coming and going regularly, loosing a bit of memory to the mayor each time. This no longer happens because that memory is freed each time.

Job Stop Hang-Up

• Sometimes if a program was locked up and you stopped a job the job would never completely stop and show the "Start Job" button. Now the stop job waits 10 seconds and then kills any reluctant programs that refuse to stop.

Hard Disk Driver Update

• There have been some situations with eDARTs using nearly full large hard drives where the disk driver software would get confused and loose track of the drive. This upgrade takes care of that.