Announcement

Collapse
No announcement yet.

A Small CNC Mill - A Huge Education

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • A Small CNC Mill - A Huge Education

    October 2006 - 11 years ago - I really knew nothing about machining. I had a small manual lathe my wife gave me for Christmas, a bench grinder I wallowed out parts with, and a drill press I wallowed out holes with. I thought I knew how to use them, but I was wrong. That's when I ordered a mill. I ordered something I could afford and I thought would be easier to use. A Taig 2019-CRER.

    I was wrong.

    It was sold as turn key. "That's for me," I thought. "I can just plug it in and start learning how to use it." Boy that was a mistaken assumption. It arrived in two boxes. One from the factory containing the basic mill, and one from the reseller containing the stepper motors and all the electronics. That's sure not turn key in my opinion.

    After several days of cussing swearing and assembly in between doing my day job I finally got it assembled and set it on a work bench. There it sat for a couple weeks until I ponied up for Mach 3 to run the machine and picked up a used computer to run it. I was starting to understand that "turn key" was just sales puffery at best. "Finally I can start cranking out parts," I thought. I was wrong.

    At first I tried to use the Wizards. A sort of short cut conversational programming macro capability in Mach 3. I thought it was me or my machine at first, but some of them are quite bad. I had circles come out like lemons and had people tell me it was the machine. The thing is the tool path on the screen looked like a lemon as well. I started to think I had a lemon. I was wrong.

    It turned out the wizards were made by a variety of people learning as they went and with varying degrees of knowledge. The parameter inputs standards weren't consistent from one to the next making them harder to learn, but once you did some of them were quite good. There is one that is really a group of tools called NFS Wizard. It allows you set setup multiple operations and string them together.

    Then there was my machine. In a lot of ways it was quite good, and in others it was quite bad. Its a kind of confusing Frankenstein's monster. The Z axis ways are a sort of box clamping mechanism with a square tapered brass gib that pushes the head to one side on the ways. The Y axis ways are two square rods with half squares on the saddle that engage them. The X axis has the only "normal" ways with a dovetail and a tapered brass gibb. I didn't have a clue about any of that, and because everything was different I was a bit afraid of it. All the axis have 1/2-20 "precision" v-lead screws with brass pinch nuts. Learning how to adjust all of that, knowing nothing, and thinking that machines had to be "perfect" left me struggling for months. The seller wasn't much help. Taig was responsive to questions however. Eventually I learned how to tear it down and put it back together... because I had done it so many times. I could take one apart and put it back together easily now. Probably less than an hour. Maybe a little more if I adjust everything how I want it as I do. Adjusting a Taig is a balance between speed and precision. I eventually I came to accept what it was and what it could and could not do.

    The electronics and control were another story. The reseller left me thinking it was my computer so I upgrade it.. a couple times. The breakout board and drives looked suspiciously like a knockoff of an old Xylotek control. Of course I didn't learn that until years later. It was slow. 20 inches per minute max and even after I had the machine dialed to the best of my abilities it still would flake out.

    I cut parts like that. My first parts were more a matter or turning perfectly good material into piles of chips and broken end mills. I'm sure my wife got quite tired of me showing her some really ****ty parts or engraving on a piece trash, and expecting validation of my prowess as a machinist. LOL. Eventually I started to find a balance between what it could do and what I wanted it to do.

    One day I was reading a tackle making forum and a machine shop owner who made molds went on a long winded rant about what was involved in making molds for rubber worms. It just sort of rubbed me the wrong way. I went out to the shop and started working on a Friday night and didn't stop until I had produced my very first mold. I didn't have any real CAD software. I didn't have CAM software. I did get ripped off for $75 for LazyCAM, but if you can produce anything other than the most rudimentary parts with it my hat is off to you. Of course knowing nothing I thought it was my lack of knowledge, not the software. I coded most of it by hand. Simple engraving type operations. I used Excell and some macros to write iterative incremental code. I learned how to write simple code for cutting arcs and then learned how to change the plane so I could cut arcs in X & Z instead of X&Y. It was very much hand coding, but atleast I had a computer to do the drudge work. 3 days later I had a mold I thought would work. It was a total piece of crap compared to the work I do today, but it would produce usable baits. I caught fish with them.

    Over the ensuing months I learned to make other molds, discovered CamBam, and wrote bigger and bigger programs. All the while I fought with my machine and my crappy controller thinking I was really just asking to much of it. I would hover over it ready to hit the e-stop when things went wrong... and they did. I would weld up holes in a piece of aluminum right there on the table and start over, because I didn't want to waste metal over and over again.

    The best thing I ever did for the machine was spend $600 on a package of steppers, cables, controller from Ahren Johnson over at CNC Router Parts. I bolted it on and it was like a different machine. I was screaming along at 30-40 IPM with the machine adjusted tight, and with it loosened up bit I could cut parts at 60 IPM. No kidding.

    I started pushing the machine to see what I could do. If I had settled for less I could have probably stopped there, but I was learning about HSM. Something a Taig is best at, but not really suited for. Its got a 10K 1/4HP spindle, but as documented the feeds didn't match that. So again it was a balancing act between over powering the spindle, pushing the steppers to hard, and adjusting the machine for precision or speed.

    The whole time I was cutting parts. More and more often I was getting finished parts that looked ok. I still sometimes had issues with the machine, but it was because I was asking more and more out of it. One Taig owner and reseller in a group all but called me a liar when I told him how many hours I had on my machine. I experimented. I ran wood routers as spindles for more power and less weight so I didn't have to worry about the spindle when pushing the machine. I easily burned up a dozen of them. Some failed because they got packed with aluminum chips and shorted out. Others got so hot the plastic spindle noses that held the bearings would melt. Some I actually went through brushes on. I started rating them by how many hundred hours I would get out of one.

    Somewhere in all of that I wrote code file for a mold that was 1.3 million lines of code and took over 30 hrs to run. I hovered over the machine as long as I could, and then I slept on the floor next to it while it ran. I was startled awake at any odd sound from the machine, and I was startled awake a lot. I wound up cutting that job twice because I made a mistake and ruined the first run. That 1.3 million lines of code was one half of a mold. I spent a solid week in the shop doing that one job. I only went in the house to grab something to eat, and to use the restroom. I still have that mold, and I still use it. People keep asking me to reproduce it and sell it, but I am pretty attached to the design.

    A couple months ago I ran 5.6 million lines of code for a mold program. 2.3 million lines per side. I ran each side on a different machine at the same time and it took a little over 8 hours. It was done right the first time. The next day after a little hand finish work deburring, pressing in alignment pins, and tapping clamping screw holes it shipped out. I spent all the time the job was running in the office doing CAD and CAM for other jobs and ignored the machines until they needed a tool change. (Sorry no ATC machines in my shop... yet.)

    Now instead of knowing nothing I feel like I know a little bit.

    That Taig taught me to machine, rebuild machines, design machines, and understand that there is no such thing as perfect. If two parts are a "pefect" fit they become one part. I retrofit a KMB1 from a non-working Randtronics control to a modern PC based control. The first time it took me a year. When I upgrade my controls it took me a day. I repaired everything that was wrong or went wrong it it. Mechanical, electrical, electronic, setup... I took a cheap flimsy Chinese noodle router and retrofit it with a robust control system and motors in a day. I redesigned the leads, motors and control system on a MaxNC and cut the first mold I was really really proud of on it. One I consider a kind of work of art. A mold I reproduce and sell today. I made parts with my Taig for several of my other machines. The Hurco KMB1 has a companion spindle mount, and an encoder cover made on the Taig. The spindle mount on the noddle route was made on the Taig. Bidirectionally adjustable captive bearing carriers for the MaxNC were made on the Taig, and its saddle was modified for spring loaded anti backlash nuts on the Taig.

    I have the Taig all apart on a shelf in the shop because I am to busy to mess with it, but I plan to put it back together again. When I do it will be a fully functional and usable display piece in my office.
    Last edited by Bob La Londe; 11-08-2017, 10:59 AM.
    *** I always wanted a welding stinger that looked like the north end of a south bound chicken. Often my welds look like somebody pointed the wrong end of a chicken at the joint and squeezed until something came out. Might as well look the part.

  • #2
    If you have read all of that and think that a Taig may be more trouble than its worth... then you would be wrong.

    The knowledge to do everything I had to do or learn is out there. Nobody currently sells it with as crappy of a controller as my first one. People will gladly guide you with software choices and control software to get started. Its not a production machine and it was never intended to be one but I used it like one. Its intended to be a hobby machine for a guy piddling around with it for a few hours on a weekend or in the evening. It would last and run for years and years like that. Maybe a lifetime. I got years of actual run time out of mine running way beyond what it was designed for. I literally made (and sold) tens of thousands of dollars worth of parts with mine. It paid for three other machines and their retrofits, and it made parts for them before I put it on the shelf.
    *** I always wanted a welding stinger that looked like the north end of a south bound chicken. Often my welds look like somebody pointed the wrong end of a chicken at the joint and squeezed until something came out. Might as well look the part.

    Comment


    • #3
      That was a very interesting and enlightening story Bob. My hat is off to you!
      Kansas City area

      Comment


      • #4
        great story - persistence is the key to learning

        Comment


        • #5
          Cool story. Thanks for sharing. I had seriously considered a Taig with the CNC retrofit ready option, but a seriously-used Hurco BMC-20 showed up on Craigslist for not much more than the Taig "kit". Spent about a year fixing what was not working, broken or needed attention and now have a very capable CNC mill. During my year of repairs, I learned a LOT about the machine and about CAM software. I also knew nothing about CNC machining when I got the Hurco and it was a big learning curve, but worth every second of time invested.
          Cheers,
          Gary

          Comment


          • #6
            Originally posted by gcude View Post
            Cool story. Thanks for sharing. I had seriously considered a Taig with the CNC retrofit ready option, but a seriously-used Hurco BMC-20 showed up on Craigslist for not much more than the Taig "kit". Spent about a year fixing what was not working, broken or needed attention and now have a very capable CNC mill. During my year of repairs, I learned a LOT about the machine and about CAM software. I also knew nothing about CNC machining when I got the Hurco and it was a big learning curve, but worth every second of time invested.
            I don't know anything about the BMC20, but it looks like a serious VMC. Definitely not a toy.
            *** I always wanted a welding stinger that looked like the north end of a south bound chicken. Often my welds look like somebody pointed the wrong end of a chicken at the joint and squeezed until something came out. Might as well look the part.

            Comment


            • #7
              Originally posted by Bob La Londe View Post
              I don't know anything about the BMC20, but it looks like a serious VMC. Definitely not a toy.
              Haha! No, it is far from a toy, and quite a scary eye opener for me (10hp spindle and 400ipm rapids), coming from my manual Bridgeport. Your KMB1 was a big jump up from the Taig and I wouldn't consider the KMB1 a toy either. Nice machines and worth the effort and money to retrofit for home and prototype shops.
              Cheers,
              Gary

              Comment


              • #8
                I really enjoyed your story Bob.
                Gives a clear picture of the learning curve with machining.
                The struggle is real.
                I don't think I would have had the patience you had learning CNC.

                --Doozer
                DZER

                Comment


                • #9
                  My first CNC machine was called a Spindle Wizard knee mill retrofit. Got it for less than nothing. Bought it at a sealed bid auction with another one that was big ass Shizouka conversion, and a mint SB 10" lathe. Paid $7500 for the lot, Sold the Shizouka for $12K and a few years later the small mill for $800. I still have the SouthBend!

                  You had to see the Bandit controller on those mills. A really minimal keypad, one line of LED numbers for the codes, and pilot lights to indicate G, M, T etc. 500 lines max but had some very nice canned cycles. I even used the tool offset compensation all the time. Learned a ton.

                  Comment


                  • #10
                    Reading your story, most of the way through, it felt like you were bitching, or complaining. All I could
                    think of was I was hearing a story of a fantastic education. As the story neared the end, it was a pleasant
                    surprised to see that was exactly your point.

                    I'm jealous
                    John Titor, when are you.

                    Comment


                    • #11
                      That's what I love about this hobby/business. It's a new lesson every day. if you aren't learning you're doing it wrong.
                      Not CNC but my boss handed me a picture of a hopper/ tub dumper and wanted me to make one. Very little machining involved but a lot of welding and my first time using hydraulics. I learned a lot about the how and why of pivots gussets and braces.
                      'The resident welder said it would never work. Turned on the pump, loaded a 2000lb box of screws and pushed the lever. Worked as advertised.
                      Biggest take away from Bobs storey?
                      persistence.

                      Comment


                      • #12
                        The Taig is a good little mill, I run a MicroProto well actually 2. One at work and one at home. They are very versatile and like you I have learned a lot from rebuilding, servicing and improving.


                        Sent from my iPad using Tapatalk

                        Comment


                        • #13
                          Great read, Bob. Thank you for taking the time to post all of that.

                          Are you at liberty to say what those molds, with the 1+ million lines of code, are for and about how large the finished molds are? They can't be that large having been made on machines of that scale, they must be mind numbing-ly intricate.

                          Comment


                          • #14
                            Originally posted by alanganes View Post
                            Are you at liberty to say what those molds, with the 1+ million lines of code, are for and about how large the finished molds are? They can't be that large having been made on machines of that scale, they must be mind numbing-ly intricate.
                            Frankly that story sounds a bit fishy. I'm not going to take the bait. But some will, hook line and sinker....

                            Running and programming CNC machines for a living (couple haas mills, and a nakamura lathe) I've always been kinda curious to how the tiny hobby CNCs ACTUALLY perform. I know they'll be slower, etc, but in a small garage shop where time can be discounted and the cost of entry is much cheaper they might be fun to play with. One could always draw something up, and send it out to be cut, but it's not always about that.

                            What kind of feedrates were you seeing on the taig while surfacing? Making those molds on that machine your were certainly pushing the limits of what that machine was probably designed for. I commend your resolve, I certainly wouldn't have had the patience for that.

                            Comment


                            • #15
                              Originally posted by Dan Dubeau View Post
                              Frankly that story sounds a bit fishy. I'm not going to take the bait. But some will, hook line and sinker....

                              .....
                              Clearly the millions of lines were not written out longhand, and typed in..... I assume that the way they got to that size is by having a short routine which got copied many times into a larger routine, possibly with some one line modifications. Then that larger routine may have in turn been copied many times, again with some one line modifications.

                              Otherwise, well....... it would take something over one working year to input a million lines at 10 seconds per line.
                              Last edited by J Tiers; 11-10-2017, 09:22 PM.
                              CNC machines only go through the motions

                              Comment

                              Working...
                              X