Igor Chudov wrote: > I have a possible job to do to drill and tap 200 holes. The more I > think, the more it seems that I would be served well if I get rigid > tapping to work. > Having done exactly that - a fixture plate with 288 holes - I am sure you are right. Either a tapping head or rigid tapping. Otherwise, very sore arms and back. > My question is, how does rigid tapping handle spindle reversal? > When the depth specified in the command is reached, the spindle speed is reversed. If your spindle direction is by relays only, then it is a pretty abrupt reversal. if you have spindle speed set by analog output to the VFD, then it can be run through the HAL component lowpass (how I did it because I understood the component) or limit (recommended by others). You set the filtering rate-of-change so that your Z axis servo can follow the reversal without excessive following error. > How does EMC know so well how speeds acts when the spindle is stopped, > then reversed? > It is constantly watching the spindle position. "Speed", in fact, isn't really monitored, it is position. The spindle's encoder counter is zeroed at the index mark when entering spindle-sync'ed motion, and then the encoder is a count of rotation. 1.00 equals the first full turn, 2.00 is the second full turn, etc. The Z axis is slaved to the rotation times the thread pitch. As the spindle stops and reverses, the Z just stays in sync with the encoder, all the way in, and all the way back out.
There are a couple things that have to be right for this to work. One is that the trajectory planner must be run at the same period as the servo thread. Otherwise, the position interpolation of the T.P. introduces a delay which causes Z to lag one way going in and the opposite way coming back out, dragging the tap. So, you would have something like this in the .ini file : # Servo task period, in nanoseconds - will be rounded to an integer multiple # of BASE_PERIOD SERVO_PERIOD = 1000000 # Trajectory Planner task period, in nanoseconds - will be rounded to an # integer multiple of SERVO_PERIOD TRAJ_PERIOD = 1000000 Your spindle encoder needs to have minimal backlash in however you attach it to the spindle. You have to check with air cuts and Halscope to make sure the Z following error is within a reasonable tolerance during the spindle reversal, at the thread pitch and spindle RPMs in question. Once you have this figured out, you can probably come up with a table that threads finer than X TPI can be cut at 2000 RPM, threads down to Y TPI can be done at 1000 RPM, and the threads coarser than Z TPI must be done at 500 RPM or below. If your following error limits are set tight, then exceeding that rule would cause a following error on Z, requiring you to back the tap out manually - not a lot of fun! The tradeoff on the VFD acceleration is that the tap will go deeper than the commanded depth. The slower the acceleration of the spindle, the more the tap will coast in before the reversal happens. When tapping thick material or blind holes, the amount of this overrun can be a concern. Jon ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
