Announcement

Collapse
No announcement yet.

CNC, Stepper Motors, and Cheap Mill Drill (HF) retrofit

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

  • CNC, Stepper Motors, and Cheap Mill Drill (HF) retrofit

    I have a couple three questions:

    1) How do you figure out the minimum ounce inches of stepper motor you need to drive x,y,z on a cheap mill? Do you measure the torque needed to twist the screw then multiply by 2 ?

    2) How do you, or how does CNC account for play in the lead screw ?

    3) Anything else I need to be aware of in a CNC retrofit ?

  • #2
    I look to see what other folks have done and listen to and read their complaints. I also tend to weight everything I read with the fact that everybody says their stuff is awesome, and if somebody says they had a problem there is always some guy who implies it must be their own fault since he didn't have that problem. Both have something to teach.

    Remember that steppers tend to have the best torque at lowest speeds. All of the steppers on the machines I am running are direct drive, but a lot of guys use pulley reduction with timing belts to optimize their motors to be in the best torque range.

    I've got three little Syil Speedmasters. 2 have linear rails and one is a dovetail and gibs machine. They use what looks to be steppers in the 800oz/in range. They all three have ball screws, and a gas spring supporting the head to reduce the weight that has to be moved by the motor. I've got rapids set at 150ipm and acceleration set at 25iss in X/Y and 120/20 for Z. All three machines seem to handle that ok. I use them in semi production for custom parts. I guess that means they run a full 2-4 days a week depending on jobs.

    Since I do not know WHICH HF mill drill you have I'd consider this to be a compromise. If you have the mini mill the head, table, and saddle are much lighter than the Syil Speedmaster. If you have an RF31 style machine the table and saddle are a bit heavier, and you would power the quill. Not the head. If you plan on using the acme screws in the machine you will have more friction, so the answer is probably not bigger motors, but to compromise at a lower rapid and acceleration in your controller setup. Lots of guys running small mills quite happily at lower feeds and speeds.

    Now for my oddballs. I've got three other CNC machines.

    A Taig mill with 351oz in motors that holds pretty well at 65/20 on all axis if I keep the machine adjusted very loose, but if I adjust everything to hold under .001 tolerance I max out with reliability at around 35ipm and 15iss. It is also direct drive. It has 20 TPI precision ground V-lead screws in brass feed nuts. They now make a ball screw version of the Taig which I have zero experience with.

    My next machine is a very light weight el cheapo Chinese ball screw and linear round rail gantry router. It came with much lighter motors, but I threw 351oz/in motors and a different control on it. I tested it at over 600ipm and upto 35iss. Sounds awesome and under no load testing it seemed to work fine, but because its light and flimsy it tends to bind slighty from flexing under cutting loads. Over the course of a long job it will slowly loose position. (I sometimes run jobs that take a full day to cut. In the past I ran jobs as long as 30 hours continuous.) I dropped it down to 200/20 and over long jobs it still loses a little position, but for what I do on that machine its "good enough." I mostly using it for wood cutting, but some hard woods I've cut on it are as hard as aluminum or harder from the sound of the cutter.

    My last machine currently in service is an old Hurco KMB1 that I gutted out and retrofit to a more modern PC based control. It has DC servo motors so it doesn't really apply to your question. Now I am just bragging. LOL. It does have a table and saddle that are pretty heavy. Probably over 1000 lbs. The whole machine weighs in at 2 tons. The X on this machine is direct drive. Y & Z are belt drive, but they are 1:1 ratio. I tested (with large following error allowed) at 450ipm and 100iss. No that is not a typo. When I dialed it in at what I considered a decent safe feed and accell I had it at 300ipm and 30iss with a less than 50 count following error in worst case situations. Usually less than 10 according to the diagnostic software. I ran it for a year that way when I lost one servo driver. It happened to be Y which carries the heaviest load. After replacing the drive I dropped everything down to 200 and 25. Its has a slow spindle (3600RPM max) intended for slow heavy cutting. I rarely run it anywhere near as fast as it can run due to the types of cuts I do with it. It has servo motors that look to be about the size on an electric starter motor on a truck. I never have looked at the actual torque rating for them since that type of motor is limited by the voltage and current limit you set. I run at 78VDC with a 35 amp per drive limit. I guess that means they have a peake limit of around 2.7Kw. That would be over 3.5 HP. The reality is if I ran them anywhere near that hard for more than a second under a hard reversal I'd burn things up. I would bet in the real world they can only run continuously at 1/4 of that load. Even then it probably only hits that load when I slam into a a part hard enough to break a 1/2" end mill.

    Anyway, the nice thing about servos is they self correct, and if they do loose a tiny bit they catch up. Your typical stepper system is what is called open loop. You send some many pulses to the driver and it moves the stepper so many steps (or micro steps), but there is no way to KNOW that its really about where you want it until you measure your finished part. There are closed loop stepper systems out there. I have a couple... still in boxes for future builds. No experience to share there. Sorry.

    If I was building a mini mill I'd look for solid reliable results with modest feed rates and low acceleration figures.

    I suggest you go over to CNC Zone and read a bunch of old build threads. Its a huge forum and there are a fair number of wankers over there, but there are a lot of good guys too. Many who are very advanced expert users who started out just like you trying to figure out the basics of retrofitting a cheap machine to get ok results.
    Last edited by Bob La Londe; 09-25-2017, 12:02 PM.
    *** 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
      Thanks for the info - I will check out that site.

      Comment


      • #4
        I looked at this some time ago for a project that is unfortunately, still in progress. Here are some conclusions that I came up with.

        1. Yes, you can measure the torque needed to move the axis, but this is not simple. Things like the force needed for moving the cutter, the tightness of the gibs, and other factors can easily and quickly come into play. Bob talks about how tight his machines are. If you want accurate cuts, then the gibs must be tight. Just put an indicator on the table with the gibs loose and push it one way and another. You will see 0.005" or more of movement.

        2. How is play or backlash in the lead screw handled? The most popular answer is "ball screws", which have very low or zero backlash. I would say that more than 99% of all CNC machines use ball screws. And most of the ones that use standard, Acme lead screws will also have a full servo control with scales to actually read the positions.

        If you can't convert to ball screws, then the CNC software will have to approach each cut in the same manner that you would while milling manually. Always start the cut from the same direction so that you are using the same side of the thread on the lead screw. We call this, "Taking up the backlash." I do not know which, if any CNC packages do this on an automatic basis. But you can do it if you manually go over the G-code and modify it for cutting in the same direction at all times. But, when moving to the start of a cut, you must also go past that start point and then proceed to it from that same direction. This also is just like you would do it on a manual cut. Raise cutter - back up to before the start point - go forward to start point - lower cutter - then do the cut.

        I mentioned a full servo system above. A servo system uses a scale or other device to measure the actual position of the table and then it moves it to the desired position by comparing that measured position to the desired one. This is called a closed loop system and it can be done with any type of screw. But many servo systems would still have ball screws for their other advantages. Servos are expensive so they are on high end machines and that makes ball screws a more natural choice with servos.

        Most inexpensive CNC systems are, what are called, open loop systems. The software issues a command to move to a new position. The stepper motors are responsible for making the proper number of steps for that move and that is it. The software has no way of knowing if that move was successful or not. This is why we worry about "missed steps" when using steppers. If the stepper motor does not actually move the table for each and every step, then the rest of the program is off by that missed step and all features cut after that point will be one step off. The answer to this is to oversize the stepper motors and be certain that the operational parameters are set so that steps are never missed. Generally, this means slower speeds.

        3. Anything else? Well, plenty. You need to worry about almost every detail.

        One way to approach this would be to find and examine a commercial CNC kit for your machine. Check out the size of the steppers. Does it use ball screws? What software? Read the reviews. Etc.



        Originally posted by 01-7700 View Post
        I have a couple three questions:

        1) How do you figure out the minimum ounce inches of stepper motor you need to drive x,y,z on a cheap mill? Do you measure the torque needed to twist the screw then multiply by 2 ?

        2) How do you, or how does CNC account for play in the lead screw ?

        3) Anything else I need to be aware of in a CNC retrofit ?
        Paul A.

        Make it fit.
        You can't win and there is a penalty for trying!

        Comment


        • #5
          great advice thanks

          Comment


          • #6
            If you have 0.010 backlash in X and 0.010 backlash in Y, then you probably have around 0.020 of accuracy under CNC control which may or may not be acceptable. In my opinion, you may not want to bother putting X/Y CNC control on any machine that does not, or can not be modified to use ball screws for zero backlash. If your X/Y axis has backlash then it's easy to compensate when you're manually milling. For example, if you're cutting in Y you can manually lock the X axis and mill away in Y knowing the table isn't going to be moving in X. With CNC, it will go ahead and mill in Y and assume the X axis is locked but X backlash will allow the table to jerk around in X within the tables backlash range. You don't want the X axis to jerk around while milling in Y which obviously isn't good.

            Comment


            • #7
              While ballscrews are the "go to" anti backlash solution, there are other ways. You can make or buy a spring loaded leadscrew nut that eliminates the backlash, but at the expense of eventual wear of the nut or leadscrew. For a home shop that may not be an issue. Then there are the "evannuts" molded in place plastic nuts. Probably not great for running 24x7 x365, but good enough for the home users.

              I really hoped someone knew the proper way to measure the needed stepper strength for an oddball use. The idea of copying someone who copied someone who copied someone.... just does not instill a sense of confidence in the information. I have a mill built in Taiwan in 1978 that would benefit from a power feed on the X axis, but no idea what size is really needed to drive it.

              Dan
              At the end of the project, there is a profound difference between spare parts and extra parts.

              Comment


              • #8
                Go ahead and check the torque needed to rapid the axis at max workpiece load (if work sits on the moving table). Include acceleration.

                Then check the torque needed when making a cut with an average cutter (sharpness) of max size on the same axis in the worst target material.

                Take the larger of the two, add at least a 50% to 100% margin, and go with that.
                1601

                Keep eye on ball.
                Hashim Khan

                Comment


                • #9
                  There is software to help get you in the ballpark, Kollmorgen etc, it is aimed at their products, but it gives you an idea what is needed.
                  On the question of backlash, many higher end systems use a programmed in parameter value, this has to be measured with dial gauge etc.
                  For example, if an axis reaches the end of a stroke and then reverses, the system know how many revolutions is needed to take up the backlash, before the actual axis move occurs.
                  I have always used closed loop, i.e. servo's, although many are using steppers, I prefer the added performance servo's offer.
                  Max.

                  Comment


                  • #10
                    Thanks Max
                    At the end of the project, there is a profound difference between spare parts and extra parts.

                    Comment


                    • #11
                      One other thing I forgot to mention. If you are talking about the mini mill drill (their smallest one) Little Machine Shop has one that is virtually the same except it has a fixed column instead of a swivel column. For CNC I believe this is a better machine as you have one less item to deal with. For manual milling being able to tilt the column is a mixed blessing, but some golks find the benefit.
                      *** 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


                      • #12
                        Spring loaded screw/nut? Ah, yes. That brings it's own set of problems. If the forces are greater than the spring force, you get backlash anyway. You just don't know when it is going to happen.

                        Nothing's perfect.



                        Originally posted by danlb View Post
                        While ballscrews are the "go to" anti backlash solution, there are other ways. You can make or buy a spring loaded leadscrew nut that eliminates the backlash, but at the expense of eventual wear of the nut or leadscrew. For a home shop that may not be an issue. Then there are the "evannuts" molded in place plastic nuts. Probably not great for running 24x7 x365, but good enough for the home users.

                        I really hoped someone knew the proper way to measure the needed stepper strength for an oddball use. The idea of copying someone who copied someone who copied someone.... just does not instill a sense of confidence in the information. I have a mill built in Taiwan in 1978 that would benefit from a power feed on the X axis, but no idea what size is really needed to drive it.

                        Dan
                        Paul A.

                        Make it fit.
                        You can't win and there is a penalty for trying!

                        Comment


                        • #13
                          I'm using Evanuts on my CNC router. They've been working fine for several years now. You may try finding the thread with Evan's PTFE nuts on his mill.

                          p.s. Don't listen to all those people that say HF mills aren't good for CNC. They don't have as much money as you do.

                          Comment


                          • #14
                            Something else to think about regarding CNC retrofits....

                            If you can add CNC ability to a machine but still keep the ability for full manual control without the CNC gear interfering then there shouldn't be any harm or reason not to add CNC control to any machine. Depending on backlash and other factors, the machine may or may not provide the tolerances you need when in CNC mode, but you'll always be able to switch to manual mode so it's not a total loss...

                            If however, once you convert to CNC and you can't continue to use the same machine in manual mode, then you really need to think about being able to maintain the same tolerances you had when in manual mode because manual mode is no longer an option and your now stuck with sloppy CNC tolerances...

                            Comment


                            • #15
                              Originally posted by elf View Post
                              I'm using Evanuts on my CNC router. They've been working fine for several years now. You may try finding the thread with Evan's PTFE nuts on his mill.

                              p.s. Don't listen to all those people that say HF mills aren't good for CNC. They don't have as much money as you do.
                              I think I found that thread but it may have been attacked by the photobucket virus.

                              Comment

                              Working...
                              X