584,808 active members*
5,216 visitors online*
Register for free
Login
IndustryArena Forum > MetalWorking Machines > Daewoo/Doosan > I never understood what Fanuc people meant by Fansuck until this 5700.
Page 1 of 3 123
Results 1 to 20 of 47
  1. #1
    Join Date
    Jun 2006
    Posts
    424

    I never understood what Fanuc people meant by Fansuck until this 5700.

    This is our first Fanuc mill. We have a Hurco mill that we previously had. We asked some questions of the sales guy who suggested tape mode would allow us to run programs of enormous size from RS232 with normal functionality- specifically search, replace, edit, restart.

    In reality you lose, search, edit, re-acquiring the program and can only regain some function by programming the single path machine like a cumbersome multi-path machine using 198 calls for multiple programs by tool -(programs that also cannot start in the middle because Fanuc won't allow you to search the 232 program). It's a waste of time, now I realize why people don't want to buy fanuc controls on mills. This is probably the side of the house where Fanuc gets the **** kicked out of them. Our other 5 Fanuc controls are multiaxis turning, and we don't seem to have problems getting programs to fit in the control memory.

    When the solution to the problem is do **** harder and get more headaches to get the job done that's a motivator to buy something else. I had about two days of workholding jobs - 4 different jobs all settup with programs supplied in 2 hours on the first day. Dancing around with G68 angle skews and tape mode has probably added two days to that work. I can see the operator body language saying, "I'm really frustrated with this crippled operating condition and these hurdles", and that guy has a good attitude. I dislike strongly when outside conditions effect my employees enjoyment of work because that tells me this stuff is pushing those people closer to a place where I can't compete and keep them.

  2. #2
    Join Date
    Jul 2005
    Posts
    380

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    OK, running from "TAPE" mode is one way to run a large program. But somewhat balky. I prefer running from the PC card slot, I/O channel 4. Much easier.
    You make each tool a "program" on the card, ie - Center drill is O0001, Drill, O0002, etc.
    Then, On your CNC memory, you would write a main program -

    %
    O1000(MAIN 1)
    M198 P1;
    M198 P2;
    etc.
    M30;

    Now, you can run LARGE programs without a balky RS 232 connection, plus MUCH easier program restart and re-run.
    The G68 function would be in one of those programs with your angle tolerances inserted, and you can run.

  3. #3
    Join Date
    Jul 2005
    Posts
    380

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    When you and your guys are ready. Training is free. For life. As long as you have a Doosan on your floor.
    All you have to do is get here. Lunch and coffee are free. Make your own travel lodging arrangements, and two days, hands on, here in the showroom.
    You register for free online:
    https://doosanmt-training.coursestorm.com/

    Knowledge is power. Come get some.

  4. #4
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Quote Originally Posted by DouglasR View Post
    OK, running from "TAPE" mode is one way to run a large program. But somewhat balky. I prefer running from the PC card slot, I/O channel 4. Much easier.
    You make each tool a "program" on the card, ie - Center drill is O0001, Drill, O0002, etc.
    Then, On your CNC memory, you would write a main program -

    %
    O1000(MAIN 1)
    M198 P1;
    M198 P2;
    etc.
    M30;

    Now, you can run LARGE programs without a balky RS 232 connection, plus MUCH easier program restart and re-run.
    The G68 function would be in one of those programs with your angle tolerances inserted, and you can run.
    Based on the short time we've been working with this, it doesn't seem there is a training solution to this problem. I do appreciate the offer, but our people are busy trying to produce work with the machine.

    It is a pretty large weakness in the machine, and that's obvious, but this is difficult to figure out a solution to. We did call fanuc, and our Ellison application person, who basically confirmed there is no way to run a somewhat normal CAM produced mill program in the machine without using some form of drip feed and awkward method of running the program. We did figure out the


    Program start of main program
    M198 P0004 (program O0004 ending in M99)
    M01

    Program continues ending in M30

    Method of programming the machine, but that was the less handicapped version that still represents an abnormal way to run the program that costs extra time and takes most of the control away from the operator. (edit, restart, search etc)

    In this case we specifically asked a pre-sale question about the small control memory and being able to load and run large programs normally with search, edit, restart, and other functions normally in easy guide I. We were told by sales rep that that was possible through the fanuc card slot and the only weakness of the method was literally the card being in the slot.

    That wasn't true- probably a lack of understanding by the rep. "No weakness" doesn't cost 25-50% of the time negotiating 2 days of prove outs. - Litterally oh **** moments where the program is stopped, and we have to re-run large sections of code over just to get back to where we were. - machine movement, spindle rotation, needless wear etc. Fanuc did that. They sucker punched the customer with the lack of control from the card.

    Now apparently maybe the Fanuc data server is a way to confront the reality that the machine has insufficient memory for normal CAM programs on mundane parts. I don't know what that option costs, but it should be something Fanuc just does to be competitive, because the machine has nearly no memory. Solid state 250gig drives cost $60-100 and shouldn't reduce the reliability of Fanuc's hardware only machine because they are solid state hardware.

    Doosan did a good job making the machine. Fanuc is just cutting its balls off. Fast tool changes are negated if your people are spending hours to proove out parts because they hit a button and have to restart from the beginning a long section of drip fed code, or because they have to load several programs and weave them together like a programmer to get the machine running.

  5. #5
    Join Date
    Dec 2013
    Posts
    5717

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Or you could do what I did with my lathe and rip out every piece of hardware that says Fanuc on it, and replace it with something that works and is user friendly.
    Jim Dawson
    Sandy, Oregon, USA

  6. #6
    Join Date
    Jul 2005
    Posts
    380

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    There is a data server option that can be field installed.
    Also, if you are aware of the M198 function then perhaps you can exploit it to your advantage. You can divvy up the roughing sequences etc to make it easier to tackle.
    Just tryin' to help.

  7. #7
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    I was doing really pathetic stuff like cutting one set of step jaws on a two station vise and having it consume 30000 lines of code roughing with a dynamic mill path, and that was more than the control could let me load, so we were spending more time jockying with files than actually machining parts. I talked to some CAM gurus on a Mastercam forum who gave me tips to condense the paths to 25% of their size.

    That doesn't take away the fact that Fanuc's memory is embarassing, weak, and pathetic, but it does possibly reduce the extent to which I will run into the problem. We intend to do production work, up to 24 parts in the machine at a time with workholding we already have, so 25% of the size probably won't eliminate the problem.

    I think the Doosan machine is better than the Fanuc control at this point. In my couple of years shopping around, program memory is the worst beating Fanuc takes. The I series Oi's are behind the times in block read speed, look ahead, and program memory. It is easy for people not to notice block read speed, but look ahead and program memory are pretty important today.

  8. #8
    Join Date
    Jan 2009
    Posts
    52

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Quote Originally Posted by Green0 View Post
    I was doing really pathetic stuff like cutting one set of step jaws on a two station vise and having it consume 30000 lines of code roughing with a dynamic mill path, and that was more than the control could let me load, so we were spending more time jockying with files than actually machining parts. I talked to some CAM gurus on a Mastercam forum who gave me tips to condense the paths to 25% of their size.

    That doesn't take away the fact that Fanuc's memory is embarrassing, weak, and pathetic, but it does possibly reduce the extent to which I will run into the problem. We intend to do production work, up to 24 parts in the machine at a time with work holding we already have, so 25% of the size probably won't eliminate the problem.

    I think the Doosan machine is better than the Fanuc control at this point. In my couple of years shopping around, program memory is the worst beating Fanuc takes. The I series Oi's are behind the times in block read speed, look ahead, and program memory. It is easy for people not to notice block read speed, but look ahead and program memory are pretty important today.
    I am sorry your having an issue with the Fanuc. I have a couple of comments. First, Doosan picks the CNC they offer on the machine including all the options and storage. The person who purchased the machine also had a say on what the machine specification were. Finally if a Cam system produces 30k lines of code to do some roughing on step jaws, then you should really look for a better cam system. To me it look like someone dropped the ball before you started working on the machine. Fanuc can be a pain some days, but day in and day out if you want to make money with your machine, Fanuc just works, day after day after day......

    Best of luck I hope you fine an answer that fits your needs. While data server works it is not a direct replacement for part program storage.

  9. #9
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Dynamic milling is awesome. Here are the tips you probably already know, but the defaults are suck for short programs.
    Also remember that the entire point of dynamic paths is to reduce depth cuts, so make your depth cuts your entire cutting length of the end mill. I have seen programs in the wild that are 3MB when they only needed to be 300K.

    In the example bellow I leave .02" on the XY. This will guarantee no over cutting. Might want to leave more if more aggressive, tool flex will take off more material than you expect at large depth cuts.
    TIP: the total tolerance is what makes toolpaths really small. There is no reason to hold it at .001 especially if you are leaving lots of stock. Open that up and let mastercam make huge arc moves to replace many many line moves.

    Attachment 408252

  10. #10
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    I'm pretty sure we actually forgot how to drip feed from the card from a program call. I paid someone to figure it out again and write the procedure down. I guess we should have made that a written procedure. These secondary problems are such that in hindsight I would have just bought a Hurco or DMG mill if I had known what a pain in the ass it would be to run the machine. We have a macro now that does some memory saving function from some CAM guru but we don't even know what it means or how to use it- I'll drop that in below.

    We called about a memory upgrade today, and Fanuc only offers 2MB (so 4 times not **** for memory) max storage on this low numbered control (there was no higher numbered option for the DNM5700). With our 31I B lathes we did a memory upgrade to 4MB (smaller than the 8 max because lathes don't get too crazy for lines of code) for $1700 per machine $425 per MB. Ellison/Fanuc wants $2000 for these 2MB $1000 per MB. In 1996 Nintendo released Tales of Phantasia and Star Ocean for $69 and it had 48MB of ROM memory- $2.30 per MB when adjusted for inflation. Fanuc apparently punishes customers for their failure to design adequate memory into their control. The program memory is just embarrassing.

    here's the Greek macro that is intended to allow recycling code on different offsets to help the small minded Fanuc keep up.


    O1000 (MULTI-OFFSET MACRO)
    (MACRO RUNS SUB PROGRAMS STARTING AT G54 SUB PROG O1001)

    N10 (MACRO SETUP/INITIALIZE)
    IF[#510EQ1]GOTO20 (SKIP IF INITIALIZED)
    G4X.1 (STOP LOOK AHEAD)
    #500=10. (TOTAL NUMBER OF OFFSETS)
    #501=1. (FIRST OFFSET, 1=G54.1P1)
    #502=1001. (FIRST SUB PROGRAM)
    #510=1. (SET INITIALIZED)

    N20 (CALL OFFSET)
    IF[#501GT#500]GOTO9000
    G90G54P#501
    G4X.1

    N30 (SETUP SUB PROG NUMBER)
    G65P#502
    #501=[#501+1]
    #502=[#502+1]
    M01
    IF[#501LT#500]GOTO20
    G4 X.1

    N40 (END ALL)
    #501=1.
    #510=0.

    G4X.1
    N9000 (ERROR - OFFSET OUT OF RANGE)
    #300=1(OUT*OF*RANGE)
    M30

    I just got an update, and that remembering how to wrestle with Fanuc and squeeze some capability out hasn't happened yet. I'm just really unhappy with the **** piss memory of this machine. It just handicaps the hell out of a decent machine. You can't for any reasonable cost buy an actual solution to this problem. 2MB is apparently it. Everything else runs differently as far as people are telling me. You just can't pretend this machine is competitive with anything out there like this. Even HAAS has 1GB standard.

    Fanuc should send everyone an apology letter with the 2MB for free and say, "We're sorry we're just incapable of being competitive on this."

  11. #11
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Can you post a picture of the part you are doing with a ruler next to it. Or better yet the entire fixture on the machine. Then we can get an idea if its a programming issue that can be solved with CAM or if the part is just too complex for the control.

    2MB is a fair amount of memory. Unless you are doing something very memory intensive like 5x milling or intense 3D surfacing it will suffice for the majority of 2D and 3D parts. A picture of all your tool paths in backplot with the dots turned on will also help.

  12. #12
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Quote Originally Posted by avongil View Post
    Dynamic milling is awesome. Here are the tips you probably already know, but the defaults are suck for short programs.
    Also remember that the entire point of dynamic paths is to reduce depth cuts, so make your depth cuts your entire cutting length of the end mill. I have seen programs in the wild that are 3MB when they only needed to be 300K.

    In the example bellow I leave .02" on the XY. This will guarantee no over cutting. Might want to leave more if more aggressive, tool flex will take off more material than you expect at large depth cuts.
    TIP: the total tolerance is what makes toolpaths really small. There is no reason to hold it at .001 especially if you are leaving lots of stock. Open that up and let mastercam make huge arc moves to replace many many line moves.

    Attachment 408252
    Thanks for that - I'm going to try that to help make the modern programs off the most popular CAM system in the US CNC industry more compatible with Fanuc's abysmal memory.

  13. #13
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Filtering your tool path like that will make it run faster on any machine since it will have to process 10X less code. Please try it. The cam system defaults are set up to work in 95% of cases. Unfortunately that also kills performance. If you can send me your mastercam file and I can inspect it for you.

    AG

  14. #14
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Quote Originally Posted by avongil View Post
    Can you post a picture of the part you are doing with a ruler next to it. Or better yet the entire fixture on the machine. Then we can get an idea if its a programming issue that can be solved with CAM or if the part is just too complex for the control.

    2MB is a fair amount of memory. Unless you are doing something very memory intensive like 5x milling or intense 3D surfacing it will suffice for the majority of 2D and 3D parts. A picture of all your tool paths in backplot with the dots turned on will also help.
    I was just making 6 1.25" size approximate parts in aluminum from extrusion per load- that's 2 6 part fixtures half op 1 half op 2, and we used arc filter to get that down to 266KB of the ~512KB so that we could fit it in the control. We originally were thinking of using 2 more stations but that would double the amount of code and wouldn't fit so we ditched half the stations from the start. Everything we've done in the machine has been a battle with memory- I think every job has gone in, needed to come out and be modified or something to get it in and running.

    Click image for larger version. 

Name:	thumbnail.jpg 
Views:	0 
Size:	106.3 KB 
ID:	412360

  15. #15
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Are you using sub programs? Mastercam will spit them out automatically with a transform operation. If I had to guess with the very limited info I have, you should be in around 32K.

  16. #16
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Don't get me wrong - I do agree 512K these days sucks.

    What you need to do is:

    1 - dynamic toolpath for roughing out all 6 parts in one shot. This should be done with with no depth cuts! Filtered.
    2 - the program for every part should be wrapped in a transform operation. The transform operation should be in incremental - the button on the right.
    I happen to be working on a similar example right now: https://i.imgur.com/rna8q5H.png



    Post the dynamic and the transform operation.

  17. #17
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Quote Originally Posted by avongil View Post
    Don't get me wrong - I do agree 256K these days sucks.

    What you need to do is:

    1 - dynamic toolpath for roughing out all 6 parts in one shot. This should be done with with no depth cuts! Filtered.
    2 - the program for every part should be wrapped in a transform operation. The transform operation should be in incremental - the button on the right.
    I happen to be working on a similar example right now: https://i.imgur.com/rna8q5H.png



    Post the dynamic and the transform operation.
    I don't honestly know what memory the machine has. Maybe it is 256. I thought it was 512, but I can only get about 28,000 or so lines and it seems like on the milling side that's not enough a lot of the time for even one program to be stored. Our lathe programs go up to about 14,000 and our lathes have 4MB, so we can store many programs in the control.

    The program right now is running we're just running it on half the vises in the machine. I got a quote on the 2MB today, but I wasn't expecting it to cost more than the 4MB upgrades for the 31I Yama Seiki lathes we have at $2000 vs $1700. I have a friend who said he would ask a Fanuc guru if the data server option allowed Search up down, find replace, edit, optional stop, and restart capabilities like the ROM memory. It was my understanding Fanuc had no answer to this problem except to say, we can give you 4 times the nothing you have for $2000 so you can have more nothing, or drip feed.

  18. #18
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    I just saw the picture of the parts. That should be a 32K-64K program. Use my example to get it there with sub programs. I think using a main program and subs in incremental is better because it removes any possibility of a stacking error. (I need to test this idea - don't quote me yet)

    Also if you have to make some kind of minor change, you just make it in the sub on the controller.. For example a federate in a problem corner.

    I do agree it sucks that you need to change your method for this particular part but its always great to learn new methods.

  19. #19
    Join Date
    Sep 2008
    Posts
    87

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    Hey, just saw the crazy macro bellow...

    Try the sub program method in incremental. Removes all this hogwash....
    Main program goes to X, Y location and runs sub in incremental code. Then the new X Y location and runs it again.... So its say....

    :1001(MAIN PROG)


    G54 (VISE 001)

    G90(ABSOLUTE MODE)
    G0 X0 Y0;
    M98P1234;

    G90(ABSOLUTE MODE)
    G0 X2 Y0;
    M98P1234;

    G90(ABSOLUTE MODE)
    G0 X4 Y0;
    M98P1234;


    G55 (VISE 002)

    G90(ABSOLUTE MODE)
    G0 X0 Y0;
    M98P1234;

    G90(ABSOLUTE MODE)
    G0 X2 Y0;
    M98P1234;

    G90(ABSOLUTE MODE)
    G0 X4 Y0;
    M98P1234;

    M30;

    :1234(SUB PROG )
    G91 (INCREMENTAL MODE)
    lots of g-code here...
    M99;




    Does that make sense?

  20. #20
    Join Date
    Jun 2006
    Posts
    424

    Re: I never understood what Fanuc people meant by Fansuck until this 5700.

    I ran those filter settings by the guys on the Mastercam Koolaid forum on Facebook, and most of those guys were saying they run .001 to .003" on the tolerance setting typically less than half or severely less than half the finish stock value, nobody confirmed .011 would work for .006" of finish stock without gouges. I'll have to just try that your way at some point and see if it works. Those guys obviously mostly have controls with 1GB storage (Okuma, Hurco, DMG etc), and don't ever have to crush a program into a tiny amount of lines of code. This is aluminum so we are taking like 35% stepover where steel would be 12% or so and triple the number of lines of code. I've been told not to use operation transforms in Mastercam because sometimes unintended results happen. I didn't know Mastercam could post subprograms. Mastercam doesn't really push a free and available training doctrine for the product, so every user knows something another user doesn't and no one knows it all, or really even most of it. I've got something like $26,000 invested in Mastercam training and mastercam training video product from third party providers and I realize I'm going to be slowly learning Mastercam for the rest of my life, or until I switch to a different product.

    Obviously if the Fanuc data server doesn't operate like program memory we're going to have to start using sub program calls- which is like the second programming task on top of making the program. My macro was intended to allow the same code to be used over and over to condense file size but I don't understand it. This simple memory problem costs a lot of time, it's been probably 10 hours on the floor already in only a few months of machine ownership. I'm going to try to streamline that into a word document explaining to the operator how to run a program, a folder for every job with multiple programs, this will cost some time on every programming solution, and some time for every machine settup and prove out. It's probably going to cost hundreds of hours of lost time over the life of the machine.

    Fanuc is just not thinking if they think this isn't a huge problem for their sales on the mill side.

Page 1 of 3 123

Similar Threads

  1. Doosan DNM 5700
    By Greegor in forum Daewoo/Doosan
    Replies: 4
    Last Post: 09-06-2018, 05:26 AM
  2. New Software for Fanuc People
    By Fanuc Mate in forum Fanuc
    Replies: 2
    Last Post: 10-19-2010, 11:43 PM
  3. Any Fanuc Karel people here?
    By BobM3 in forum RC Robotics and Autonomous Robots
    Replies: 4
    Last Post: 09-12-2008, 06:53 PM

Tags for this Thread

Posting Permissions

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