On Saturday 06 July 2013 18:29:25 Jon Elson did opine: > Chris Morley wrote: > > I am not an expert, just interested. I don't follow your reasoning. > > Jerk limiting is about having the TP ask for movement that is possible > > for the machine to actually produce. > > infinite jerk is impossible for a machine to produce movement for. > > While we can ignore it in relatively slow and small machines, I can > > not see why you would want to turn it off in some cases. > > G33 (lathe threading) assumes the spindle is mostly maintaining constant > velocity. G33.1 (rigid tapping) assumes the spindle reverses fairly > quickly at a certain point. The Z axis must follow the spindle quite > closely, or it will break small taps and muck up the thread on larger > ones. > > > It seems if you have the TP request infinite jerk, then you are must > > realize that your are asking the machine to NOT follow the TP command > > for a small instant. > > I don't see how G33.1 is any different then other machine movements. > > Other than spindle synched moves, ALL axes are under TP command, and > that should always keep them synched so they are all at the correct > coordinated position. With a spindle-synched move, the tool must > follow the spindle, which often is NOT a servo axis, and is just > generally obeying a velocity command. When the G33.1 gets > to the point of reversing the spindle, it can reverse fairly quickly, > depending on the particular machine setup, and the Z BETTER > keep up with however fast it reverses! Having any interpolation, > jerk limiting, etc. between the spindle encoder and the Z axis would > apply strain to the tap, and be very undesirable. > > Jon
I think that is another way of saying what I did Jon, certainly no argument. In my case, since the reversal of the G33.1 is commanded directly by the canned code, with no intervening stop, I had to synthesize the stop in hal, and wait for it to stop before issuing the changed direction. The actual stop is a separate function in the .hal file in that it sequences loading relays & resistors across the motor to try and bring it to a halt dynamically, and that was done with a wcomp module watching the encoders velocity output. If stopping from 1500 spindle revs, it gets a 16 ohm 40 watt load, until its down to a more reasonable speed, at which point 2 more of those resistors but paralleled for 4 ohms gets switched in at around 450 revs. At 150 revs it just throws a short on it. Now I can write a peck loop wrapping up the G33.1, that can drive a 10-32 tap half an inch into a prepared hole, backing out to clear chips, and do it in perhaps 45 to 60 seconds. Each direction change, at 300 revs, takes a bit less than 3 seconds for the stop, and accelerate to the same speed in the other direction. Listening to the z stepper growl seems to say that it is totally and absolutely locked. I have killed the motor power in mid cycle, and rolled the spindle by hand, with the z drive following it perfectly, as I expected. ;-) I am a bit proud that I was able to carve up a .hal file that seems to have convinced LinuxCNC that my lathes spindle is a servo motor. The way its setup, a stop, or any change in the motion.spindle-rev line results in a full disabling PID & PWM stop before things are re-enabled for the new direction only after the full stop has been detected. I kill it all, so even the PID doesn't get "wound up" waiting for the catchup. And, give credit where its due, without rockhopper to visualize it, it would have taken far longer to do. Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> is up! My views <http://www.armchairpatriot.com/What%20Has%20America%20Become.shtml> You work very hard. Don't try to think as well. A pen in the hand of this president is far more dangerous than 200 million guns in the hands of law-abiding citizens. ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
