On Thu, 2009-06-04 at 23:04 +1000, Erik Christiansen wrote: > The mill-drill which I'm starting to convert to CNC has 0.2mm (0.008") > backlash in the Z feed, making it a much better candidate for stepper > control than servo, IIUC. There is though, sufficient in-built upward > spring-loading for the quill to retract if the feed worm is disengaged. > I wonder if that would permit servo control, without hunting? > > My brand new rotary table has almost 0.1° backlash, so I guess it has to > be a stepper axis?
In my opinion, the biggest issue with backlash has nothing to do with the type of axis motor system. The problem is that your axis control will put your table or spindle into position, you will start or change your cut direction and the cutting forces will cause an uncontrolled movement within the backlash, which the motor controller will usually never see. Backlash compensation can improve your initial position, but the table or spindle can still have uncontrolled movement with in the backlash. With servos you can move the feedback closer to the cutter by having a linear scale on the table, but the backlash movement can be hard for the feedback system to deal with. Bottom line: steppers or servos, get rid of as much backlash as possible. (Has anyone used Turcite or other on leadscrews? http://www.moglice-turcite.com/ ) > It's a 970 kg machine, with a 1m long table, so even when I stump up for > ballscrews, the thought of trying to drive X and Y with steppers is > daunting, and I fear lost steps, so expect I'll have to go with servos > there. I'm not a stepper fan, but they work well (not fast) on my Shizuoka: http://www.wallacecompany.com/machine_shop/Shizuoka/00018-1a.jpg http://www.wallacecompany.com/machine_shop/Shizuoka/ Both steppers and servos will miss steps or trip a following error if overloaded. The key here is to size the axis motor system to fit the load. Trying to evaluate the load is the hard part, which I haven't figured out yet. It may be best to look at a similar machine that is proven to work well and copy it. One thing I have been thinking about recently is to make up an over-sized motor mounted so that I can measure torque, speed, voltage and current, mount it to the axis in question, run it through its paces, then use this data to determine how much smaller the real motor system can be. > I've found some servos with attached encoders here: > > http://www.oceancontrols.com.au/motors/servo/servo_motors.htm > > but they look rather wimpy. On the other hand, I'd be amazed if I was > outputting 150W when doing manual rapids, even with the acme leadscrew. > Add some HTD pulleys, to increase the torque, and take what rapids I > get, maybe? At least the motor compares favourably with the one Kirk > found recently: > > http://www.cadcamcadcam.com/motorservoglobepm-dc.aspx > > giving 3.5 times the torque at 2800 RPM. Now, if only the appreciating > A$ had brought the local price down. > > Erik If budget is the primary problem, I would try to make what you already have work better, or buy what whatever is least expensive and will still work well enough, steppers or servos. Look at what other people are throwing out and consider how parts might be used for a CNC. Determine what is the single most objectionable issue and chose the shortest path to making it better. I often try to make something perfect on the first pass and end up doing something different in the long run, so jury rigging to prove a design may be a good thing. my2c -- Kirk Wallace http://www.wallacecompany.com/machine_shop/ http://www.wallacecompany.com/E45/index.html California, USA ------------------------------------------------------------------------------ OpenSolaris 2009.06 is a cutting edge operating system for enterprises looking to deploy the next generation of Solaris that includes the latest innovations from Sun and the OpenSource community. Download a copy and enjoy capabilities such as Networking, Storage and Virtualization. Go to: http://p.sf.net/sfu/opensolaris-get _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
