13 Replies Latest reply: Jul 4, 2011 9:35 AM by yu.yanfeng

    EE7.9 Update3 output wrong Gencad File(Maybe 7.9.1 also have this severe bug)

    yu.yanfeng

      It have been verified that EE2007.9 Update3 outputs wrong Gencad file, Bottom SMD pads be in the Top side. I guess 7.9.1 may also have this severe bug(I will check it late)
      It seems that Mentor QA seldom do  full verification before release.  Taking a joke, EE means Endless Errors.

       

      Yanfeng

        • 1. Re: EE7.9 Update3 output wrong Gencad File(Maybe 7.9.1 also have this severe bug)
          yu.yanfeng

          I have logged it in the supportnet  #2355560928.

          According to Gencad Specification, if a smd component be placed on the bottom side, it will be recorded as bottom-side placed in Gencad data, and a new footprint(shape description) also be created in the Gencad file with _bottom suffix. In this _bottom shape description, smd pad normally be defined as Top. This make CAM tool will correctly interpret the Gencad data. Below is example data:

          The correct data(From 7.7)

          SHAPE "C0603_BOTTOM"
          LINE -35.000000 -20.000000 35.000000 -20.000000
          LINE 35.000000 -20.000000 35.000000 0.000000
          LINE 35.000000 0.000000 35.000000 20.000000
          LINE 35.000000 20.000000 -35.000000 20.000000
          LINE -35.000000 20.000000 -35.000000 -20.000000
          INSERT SMD
          PIN 1 "Pad Square 40_TOP" -35.000000 0.000000 TOP 0.000000 0
          PIN 2 "Pad Square 40_TOP" 35.000000 0.000000 TOP 0.000000 0
          HEIGHT 44.000000
          ATTRIBUTE COMP_ATTR "Underside space" 0.000000
          $ENDSHAPES

          The wrong data(from 7.9)
          SHAPE "C0603_BOTTOM"
          LINE -35.000000 -20.000000 35.000000 -20.000000
          LINE 35.000000 -20.000000 35.000000 0.000000
          LINE 35.000000 0.000000 35.000000 20.000000
          LINE 35.000000 20.000000 -35.000000 20.000000
          LINE -35.000000 20.000000 -35.000000 -20.000000
          INSERT SMD
          PIN 1 "Pad Square 40_BOTTOM" -35.000000 0.000000 BOTTOM 0.000000 0
          PIN 2 "Pad Square 40_BOTTOM" 35.000000 0.000000 BOTTOM 0.000000 0
          HEIGHT 44.000000
          ATTRIBUTE COMP_ATTR "Underside space" 0.000000
          $ENDSHAPES


          However, the new release EE 7.9 Gencad output program changed all top pad difinitions for those smd component placed in the bottom  to bottom difinition, so it make all CAM tool wrongly interpets the Gencad data.This is a severe bug in EE7.9, I guess this bug may escalate to up-to-date 7.9.1.

           

          Yanfeng

          • 2. Re: EE7.9 Update3 output wrong Gencad File(Maybe 7.9.1 also have this severe bug)
            yu.yanfeng

            Now I have checked it on EE7.9.1 Build Id: 426267 (Build Date: 2010-10-27 11:33:11).  It is there.

            Yanfeng

            • 3. Re: EE7.9 Update3 output wrong Gencad File(Maybe 7.9.1 also have this severe bug)
              jeff.heidel

              Chris Smith from National Instruments called me last week and warned me about this issue.  Fortunately, my company isn't using GenCAD that much anymore.  I think he filed a bug report.  I will forward this thread to him and let him add anything new that he may have found.

              • 4. Re: EE7.9 Update3 output wrong Gencad File(Maybe 7.9.1 also have this severe bug)
                yu.yanfeng

                Jeff,

                Much thanks! I have logged it in the supportnet, and spent much efforts to let Mentor CSE to forward it to engineering team with the DR.

                Currently, Mentor CSE still don't know what is the problem.

                Yanfeng

                • 5. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                  chris.smith

                  We are writing a python script to correct this issue; essentially parsing the bad output and correcting the pad definitions.

                  It's troubling to have write scripts to "correct" a defect in a fundamental deliverable output. I hope that Mentor adds this to a QA checklist for the next release.

                   

                  The output is unusable for our MFG and ICT Engineering teams.

                   

                  As another warning we also saw problems with ODB++ output where mechanical cells had duplicate reference designators. NRD0, NRD0.
                  This causes Valor to crash. The only workaround is to delete the mech cells or add ASSY ref designators to the cells and force renumbers on the cells before ODB++ output.

                   


                  • 6. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                    yu.yanfeng

                    Chris,

                    Much thanks!

                    I have also report  the issues with 7.9 ODB++ output, including same labeles for mechanical cells and  the data  make Trilogy 5000 get wrong roration  bottom pads whiling rorating the data in Trilogy. It has passed months and eventrually Mentor say it not be planned to address in next 7.9.2.

                    I really don't understand how Mentor understand what is the impact to thier custermer.

                    Yanfeng

                    • 7. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                      Andreas.Schaefer

                      Hello Yanfeng,

                       

                      I doublechecked the issue an it is still in 7.9.1.

                      When looking into release notes, it says "dts0100632147 - Opposite side Pad in Padstack is doubly-opposed resulating in Mount side placement in GenCAD output".

                       

                      Do you think, this fix is the root cuase? May be they have presented a bad fix for another issue?

                       

                      kind regards,

                      Andreas

                      • 8. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                        yu.yanfeng

                        Hi Andreas

                         

                        Maybe they maked another error during fixing " dts0100632147". I hope Mentor can correct it a.s.a.p, don't let user to waiting  months.

                        Yanfeng

                        • 9. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                          yu.yanfeng

                          Today I received mail from Mentor CSE, told me that this bug is fixed in 7.9.1 and closed the SR. I downloaded up-to-date update6, the bug is still there.

                           

                          It's sadness that the company runs like a contry club.

                           

                          Subject: SR 2355560928

                           

                          Hi Yanfeng,

                           

                           

                           

                          This is to update you the status of Defect 00632147 logged for issue

                          "EE2007.9 Update 3 output wrong Gencad file."

                           

                           

                           

                          This is fixed in EE7.9.1 per Engineering update received today.

                           

                          Please let us know if you see any issues in EE7.9.1

                           

                           

                           

                          Regards

                          • 10. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                            chris.smith

                            There is an option to reopen logs on supportnet (I thought?)

                             

                            We tend to validate anything they marked as "corrected" also. There has been times where the testcase they used might be ok but other designs can have the same defect.
                            --When in doub't check it out.---

                             

                            Email me Yanfeng if you need help in how we created an external solution to correct the defective output.

                             

                            I would hope Mentor addresses this specific defect in the 7.9.2 release seeing that the deliverable output is completely worthless to MFG and ICT if they rely on Gencad.
                            Without the abillity of having our own Software Engineer and creating a solution outside of Mentor API; we would have been unable to use the new release.

                             

                            Chris Smith

                            Application Support Analyst | PCB Services

                            National Instruments |www.ni.com

                            chris.smith@ni.com

                            • 11. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                              Andreas.Schaefer

                              Hi Yanfeng,

                               

                              my customer support also asked me last week about this issue.

                              They said, the format was changed to help other customers with an import issue.

                              Obviously the format is not quite clear (engoineering said) and now they wanted to know, where I do an import,

                              to make their output valid for my import software.

                               

                              You should reopen yiour DR and give a comment, where you want to import it.

                              An idea here:

                              Do you have Valor Enterprise? There is an licensed GenCad import possible there. Now Enterprise 3000 is vSure from Mentor ;-)

                              If your import is wrong there, they would have to fix it anyway, to get correct results in their own manufacturing suite.

                               

                              kind regards,

                              Andreas

                              • 12. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                                yu.yanfeng

                                Hi Andreas,

                                 

                                When the Gencad file read in to Metor's Trilogy 5000(or Vsure) and CAMCAD whatever which version you use, the result is that all bottom smd pads on topside.

                                Yanfeng

                                • 13. Re: EE7.9.1 Gencad Corrupt values for BTM pads.
                                  yu.yanfeng

                                  EE 7.9.2 now eliminated the bug, the engineering team have adoted old-but-ok program. I wanna ask another question, do you have verified it on CAM350 V10.X?

                                  Thanks CSE in India and me.

                                  Yanfeng