Who proves programs in your shop???
Close
Login to Your Account

Poll: Who proves new programs in your shop?

Page 1 of 4 123 ... LastLast
Results 1 to 20 of 80
  1. #1
    Join Date
    Dec 2019
    Country
    UNITED STATES
    State/Province
    New Hampshire
    Posts
    29
    Post Thanks / Like
    Likes (Given)
    10
    Likes (Received)
    4

    Default Who proves programs in your shop???

    I'm trying to figure out if I'm crazy or not, but currently my shop has setup guys prove programs. So a draft is released, proven by the setup guy, changes are made, and then its saved back into our system. I think this is inefficient and creates alot more opportunity for changes to go unsaved and make things difficult down the line, but im open to being wrong. How does it work in your shop?

  2. #2
    Join Date
    Sep 2006
    Location
    Long Island, New York
    Posts
    765
    Post Thanks / Like
    Likes (Given)
    364
    Likes (Received)
    133

    Default

    Quote Originally Posted by cc22288 View Post
    I'm trying to figure out if I'm crazy or not, but currently my shop has setup guys prove programs. So a draft is released, proven by the setup guy, changes are made, and then its saved back into our system. I think this is inefficient and creates alot more opportunity for changes to go unsaved and make things difficult down the line, but im open to being wrong. How does it work in your shop?
    You mean the G code gets changed at the machine and then saved? Does the cam file ever get updated to reflect what changed at the machine?

  3. Likes wheelieking71, mhajicek liked this post
  4. #3
    Join Date
    Jan 2014
    Country
    UNITED STATES
    State/Province
    Oregon
    Posts
    944
    Post Thanks / Like
    Likes (Given)
    318
    Likes (Received)
    774

    Default

    Depends. Typically it's a bit of a free-for-all. We have 6 machinists that all do varying amounts of programming.

    For anything that has significant quantity, the programmer runs it. I find that programs tend to get much better optimization when the person who programmed it has to actually sit and watch it make chips.

  5. #4
    Join Date
    Aug 2008
    Location
    Sussex, England
    Posts
    3,597
    Post Thanks / Like
    Likes (Given)
    15
    Likes (Received)
    857

    Default

    Surely the right way is to make this a two stage process.

    Set-up guy proves out the initial program making any changes to the original code to get it running safely and at least acceptably fast to be the first version of the shop floor master.

    Machinists then use their experience to improve, speed-up, or whatever the program so it works better on the floor. Operator changes need to be discussed and reviewed before a new improved shop floor master is saved. Review process sounds formal but it doesn't have to be. Main thing is to make a note of the changes relative to the original code. As time goes on you will build up a list of useful changes and when to use them. Such lists can be very informative as they produce in context data to help with tooling et al selection.

    If you have an established and effective way of programming the initial code its often quicker and easier to simply sub in the "works better this way on our machines" after proving out the initial version than it is to try to produce a "works best on our machines" version immediately.

    Naturally you always keep the original as well as the latest shop floor version.

    As with all these organisation things it does depend on the size of the firm, the amount of work done and the types of job undertaken. I figure that once you get past 10 or so guys on the floor you need to start getting formal about things or you never really know where you are or whats going on. Below 4 or 5 you can probably wing it.

    Clive

  6. Likes SteveF1 liked this post
  7. #5
    Join Date
    Oct 2015
    Country
    UNITED STATES
    State/Province
    Oregon
    Posts
    606
    Post Thanks / Like
    Likes (Given)
    253
    Likes (Received)
    462

    Default

    I'm of the opinion that my code coming off my CAM system should not need to be changed at the machine.
    Something doesn't work, back to the office, adjust the model, send out the code and go again.
    I routinely re-punch a program, send it to machine, cycle start.
    Now, I have not got my lathe programming quite to that level,yet. It puts out good code, but lots of open G00 lines with no x or z. So some clean up still.

    I used to save used programs, but it just isn't worth it. I have few programs in my speedio. Program number 2000 is every production program. 1234 is every fixture cutting program.
    Once they are running right as programmed, then I can fiddle with DOC, RPM etc.

  8. Likes Bobw, cc22288, Winterfalke, Slugo, mhajicek liked this post
  9. #6
    Join Date
    Dec 2018
    Country
    UNITED STATES
    State/Province
    Pennsylvania
    Posts
    318
    Post Thanks / Like
    Likes (Given)
    766
    Likes (Received)
    83

    Default

    I am all three, so I program, setup and run my machine. Half of the guys in our shop program their parts and obviously setup and run their machines. The other half sets up and run their machines. We don't have purely operators. 12 CNC guys and 3 manual guys.


    As far as for shops who are high volume, high production, I think the programmer should prove out the program while the setup person is there with him before handing the machine to the operator.

  10. Likes cc22288, mhajicek, Sandoz liked this post
  11. #7
    Join Date
    Feb 2005
    Location
    Hatch, NM Chile capital of the WORLD
    Posts
    9,857
    Post Thanks / Like
    Likes (Given)
    17048
    Likes (Received)
    12077

    Default

    Call me old fashioned, but I can't imagine that a programmer would not prove out,
    and improve a program.

    A computer programmer is making a program.. Do you think they never actually try
    and run the program? This really isn't any different.

    Completely fine to have the setup guy dial in the fixture, and set the tools,
    but that f'n programmer better be standing there. He's the one that *KNOWS*
    what is supposed to happen.

    If its something stupid and simple, F* it. Send it and hit GO. I do it all
    the time. Complicated, my ass is right there at the machine.

    ---------

    I'm also not a fan of saving "The Program".. Save the Cad/Cam, and repost it.
    If something that you might do every couple of years, or its been a long time
    since you last did it. You want to double check it, and looking at the CAM is
    a lot easier than futzing around with a text file. Maybe you have new endmills,
    or you are using a carbide drill now, or whatever...

    I have very very few actual programs saved. And the ones I do have saved are
    really not normal, and required a lot of hand coding, but I still go back to the
    CAM and check it all out.

    I'm with Hounddog.. EVERYTHING goes into the machine as 01000. I have a few
    programs with their own #'s, but its very few. One part I do a lot is 1500,
    I have a program that hoses down the whole table, I think that is 4321. Subs
    are 1001, 1002, 1003.

    I also save 99.5% of my programs as "Delete Me". Which I have never actually deleted,
    I just keep writing it over and over and over..

  12. #8
    Join Date
    Jan 2007
    Location
    Flushing/Flint, Michigan
    Posts
    11,397
    Post Thanks / Like
    Likes (Given)
    695
    Likes (Received)
    9122

    Default

    How big is the program?
    My average on a square shank toolholder on the mill is 500 to 2000 lines and perhaps 10+ tools.
    I would think most lathe stuff much smaller.
    When on a lathe lets say clip grooves with entrance rads or chamfer. These tend to need a tweak with tool changes. Should this get stored in the database?
    Bob

  13. #9
    Join Date
    Feb 2013
    Location
    Madison, WI
    Posts
    1,303
    Post Thanks / Like
    Likes (Given)
    1379
    Likes (Received)
    836

    Default

    I program the majority of the work. The person setting up proofs out the program. I will discuss with them what I did if it's out of the ordinary, but they post it, so they have the ability to view the program before running it. Then if edits need to be made they either talk to me and we work it out together, or they just make the necessary changes and let me know for future information. There's no reason to have two people watching a program run if they can quick walk through it and simulate it in cam(unless they're green).

  14. #10
    Join Date
    Dec 2019
    Country
    UNITED STATES
    State/Province
    New Hampshire
    Posts
    29
    Post Thanks / Like
    Likes (Given)
    10
    Likes (Received)
    4

    Default

    To answer some questions:

    We're a fairly large operation. We have about 30 machines in production and 3 in development. The people operating in development are operator-only. The guys in R&D do their own setups. We have ONE person programming for both the production shop and the new development shop.

    You mean the G code gets changed at the machine and then saved? Does the cam file ever get updated to reflect what changed at the machine?
    Sometimes yes and sometimes no. Its supposed to be sent back to the programmer to be saved but it often doesn't happen that way.

  15. #11
    Join Date
    Dec 2019
    Country
    UNITED STATES
    State/Province
    New Hampshire
    Posts
    29
    Post Thanks / Like
    Likes (Given)
    10
    Likes (Received)
    4

    Default

    Quote Originally Posted by Clive603 View Post
    Surely the right way is to make this a two stage process.

    Set-up guy proves out the initial program making any changes to the original code to get it running safely and at least acceptably fast to be the first version of the shop floor master.

    Machinists then use their experience to improve, speed-up, or whatever the program so it works better on the floor. Operator changes need to be discussed and reviewed before a new improved shop floor master is saved. Review process sounds formal but it doesn't have to be. Main thing is to make a note of the changes relative to the original code. As time goes on you will build up a list of useful changes and when to use them. Such lists can be very informative as they produce in context data to help with tooling et al selection.

    If you have an established and effective way of programming the initial code its often quicker and easier to simply sub in the "works better this way on our machines" after proving out the initial version than it is to try to produce a "works best on our machines" version immediately.

    Naturally you always keep the original as well as the latest shop floor version.

    As with all these organisation things it does depend on the size of the firm, the amount of work done and the types of job undertaken. I figure that once you get past 10 or so guys on the floor you need to start getting formal about things or you never really know where you are or whats going on. Below 4 or 5 you can probably wing it.

    Clive
    Why is it so sure that thats the right way? The problems with that are that the setup guys cant make changes to the programs except at the controller. Then the file on the floor doesnt match the CAM session. That to me is problematic. And we dont have machinists really, other than myself and the programmer. We have programmer, setup, and operators.

    But if the original is kept as the latest shop floor version and a setup guy makes a change at the controller, theres a chance it doesnt get saved back. Then the next time you run the job you have a different cycle time and it screws with your rate, no?

    I agree with your last paragraph though, if I owned a shop with a handful of guys I wouldnt be even making a deal about it. But this is a larger operation and im losing my mind with all of the inconsistencies and lack of version history.

  16. #12
    Join Date
    Mar 2012
    Location
    Minnesota, USA
    Posts
    11
    Post Thanks / Like
    Likes (Given)
    4
    Likes (Received)
    5

    Default

    I voted for all three since I am all three.

  17. Likes mhajicek, Fancuku liked this post
  18. #13
    Join Date
    Sep 2006
    Location
    Long Island, New York
    Posts
    765
    Post Thanks / Like
    Likes (Given)
    364
    Likes (Received)
    133

    Default

    Quote Originally Posted by cc22288 View Post
    To answer some questions:

    We're a fairly large operation. We have about 30 machines in production and 3 in development. The people operating in development are operator-only. The guys in R&D do their own setups. We have ONE person programming for both the production shop and the new development shop.



    Sometimes yes and sometimes no. Its supposed to be sent back to the programmer to be saved but it often doesn't happen that way.
    To answer the question of the thread, the programmer runs the job through on the machines here. That's me. If I make any changes at the control I make sure to update the cad/cam file to reflect that, this way the computer has the latest.

    Having different people change it at the machine and maybe saving it and maybe not, sounds like chaos. Never mind the different cycle times, what if all of a sudden you are burning up tools where you weren't before? Did the last setup person make a change and not save the nc file? What if a setup person saw the program would have the tool hit a clamp, so they changed it and forgot to save it?
    Even better, what if there's a revision to a part so you have to move a hole or make a wall thinner or whatever, now you have to repost the program and then what? Go through all those changes again at the control that have already been done? Or manually add the new changes to the gcode? lol yikes!
    I'm nutty about being able to hit the 'post' button and having good code come out, ready to run. Especially on a repeat job.

    Bleh. Sounds bad. You have my sympathy.
    Last edited by Volitan; 08-06-2021 at 02:10 PM.

  19. Likes cc22288, Bobw, Slugo liked this post
  20. #14
    Join Date
    Jan 2010
    Country
    UNITED STATES
    State/Province
    Idaho
    Posts
    4,517
    Post Thanks / Like
    Likes (Given)
    3248
    Likes (Received)
    2616

    Default

    I've setup my shop that whoever programs the part sets the job up the first time. We have $400 solid state terminal computers with 27" screens at each machine so you can remote desktop into your CAM office computer and make tweaks while proving out the first job. No walking back and forth to the office to change things lol. Got so sick and tired of that !
    If the job has been repeated 1-2 times and we are happy with cycle times we have one of the younger guys who are learning setup and prove out the first piece. 24 machines in the shop. 15 employees . 4 programmers (1 being myself) with 7 operators only. No "setup" guys.

  21. Likes wheelieking71, empwoer, Slugo liked this post
  22. #15
    Join Date
    Sep 2006
    Location
    Long Island, New York
    Posts
    765
    Post Thanks / Like
    Likes (Given)
    364
    Likes (Received)
    133

    Default

    Quote Originally Posted by dstryr View Post
    I've setup my shop that whoever programs the part sets the job up the first time. We have $400 solid state terminal computers with 27" screens at each machine so you can remote desktop into your CAM office computer and make tweaks while proving out the first job.
    Nice! If it's a long involved program I'll take a notepad out to the machine and write down changes as they're made then fix the cad file when I'm done. It's just 4 machines, an operator and me here though.

  23. #16
    Join Date
    Jan 2019
    Country
    UNITED STATES
    State/Province
    North Carolina
    Posts
    257
    Post Thanks / Like
    Likes (Given)
    46
    Likes (Received)
    182

    Default

    Quote Originally Posted by dstryr View Post
    I've setup my shop that whoever programs the part sets the job up the first time. We have $400 solid state terminal computers with 27" screens at each machine so you can remote desktop into your CAM office computer and make tweaks while proving out the first job. No walking back and forth to the office to change things lol. Got so sick and tired of that !
    If the job has been repeated 1-2 times and we are happy with cycle times we have one of the younger guys who are learning setup and prove out the first piece. 24 machines in the shop. 15 employees . 4 programmers (1 being myself) with 7 operators only. No "setup" guys.
    Getting slightly off topic, but what equipment and software are you using for those terminals? Been meaning to set that up for ourselves.

  24. #17
    Join Date
    Mar 2016
    Country
    UNITED STATES
    State/Province
    Illinois
    Posts
    1,161
    Post Thanks / Like
    Likes (Given)
    271
    Likes (Received)
    659

    Default

    I write 99% of all milling programs at my shop. I am what I call the gate keeper of all proven programs so I usually decide what gets saved back to the network and what gets deleted. Most of my cad/cam programs get proven out in the shop and when they are done running the programs I usually save them back into the vault for future runs. I have enough trust in my set up guys to save what ever they change without reviewing it. If there are major changes in the program because of tooling, fixturing or maybe oversize stock the guys will let me know and I will update the set up sheets, programs etc.

  25. #18
    Join Date
    Jun 2012
    Location
    Michigan
    Posts
    6,396
    Post Thanks / Like
    Likes (Given)
    5970
    Likes (Received)
    4090

    Default

    Option 5: It depends

    I program for 6 vmc's.(3,4 and 5 axis) Verify and simulate the programs in Mastercam then the operators load them, setup & run them.
    If any changes need to be made they tell me and I make the changes and they re-load their programs.
    There are extremely rare scenarios where they will make changes such as a carbide drill broke and we need to use a HSS drill and they'll just modify the drill cycle to peck and flood coolant.

  26. Likes mhajicek liked this post
  27. #19
    Join Date
    Jan 2013
    Location
    Gilbert, AZ
    Posts
    7,462
    Post Thanks / Like
    Likes (Given)
    10192
    Likes (Received)
    9598

    Default

    Quote Originally Posted by Houndogforever View Post
    I'm of the opinion that my code coming off my CAM system should not need to be changed at the machine.
    Something doesn't work, back to the office, adjust the model, send out the code and go again.
    This right here!

    I will never understand how/why shops use the .nc as a "master" and not the CAM file.
    I save basically zero .nc files.

  28. Likes Bobw, empwoer, Slugo liked this post
  29. #20
    Join Date
    Jan 2015
    Country
    UNITED STATES
    State/Province
    California
    Posts
    897
    Post Thanks / Like
    Likes (Given)
    131
    Likes (Received)
    464

    Default

    Quote Originally Posted by Volitan View Post
    Even better, what if there's a revision to a part so you have to move a hole or make a wall thinner or whatever, now you have to repost the program and then what? Go through all those changes again at the control that have already been done? Or manually add the new changes to the gcode? lol yikes!
    I'm nutty about being able to hit the 'post' button and having good code come out, ready to run. Especially on a repeat job.

    Bleh. Sounds bad. You have my sympathy.
    Quote Originally Posted by wheelieking71 View Post
    This right here!

    I will never understand how/why shops use the.NC as a "master" and not the CAM file.
    I save basically zero .nc files.
    I'm in this boat.

    What do you do for rev changes using the NC file as master? Cut 'n paste crap every time? I prefer to just modify cam file and repost. Done.



    Also sounds like some guys in this thread have multiple different NC files floating around. Some on server, some on machines, maybe some "hard copies" in binders? (although that guy hasn't replied yet) If people are editing at machines, what happens when you have a mis-match? Rev change doesn't get applied second time and the customer gets a down-rev part? Yikes...


    How about if you want to run a program on a different machine/control? I'd prefer to just click 'post' and have perfect code with all the original edits there. Can't do that so easily with a "NC Master file".

    In my book, editing at the control to use for next time is nothing but a waste of time and adding more potential for error.

  30. Likes empwoer, cc22288, Bobw, wheelieking71 liked this post

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •