On Wednesday 02 April 2014 12:26:19 Peter Blodow did opine: > Am 02.04.2014 14:11, schrieb andy pugh: > > On 2 April 2014 12:54, Peter Blodow <p.blo...@dreki.de> wrote: > >> No problem with that, because reversing the drive is done with one > >> lever, no cranking, no re-engaging, no dial or indicator necessary. > >> All thread cutting is done this way over here in the shops, I've > >> never seen any other. > > > > I would imagine that it makes it hard to thread up to a shoulder, > > especially with a lathe with no spindle brake. > > (and older lathes with screw-on chucks can often not afford to have a > > brake) > > It can be challenging, good nerves necessary - I use a 5 Hz setting with > the VFD to slow things down in such rare instances. > > > Threading dials typically work better on Imperial lathes. With a > > metric lathe you need to keep changing the thread counter gear. > > I don't even know what threading dials are and where they are installed, > sorry. I usually cut a thread up to a given length, then stop and > retract the tool in one instance. It's a two hand job. I can put the > drive directly in reverse for a fraction of a second instead of a brake. > Don't like it, though. > > Peter
All this talk about doing manual threading, along with the pitfalls if ones reflexes aren't spot on, makes me want to see both the G76 and g33.1 (and possibly other spindle/zmotion locked canned cycles, get a little smarter. I have a tendency to start something like that at a slow enough rpms that I can visually follow the first cut or two, like when cutting threads up to a shoulder, making sure the backout at the end of the cut stroke is done both far enough out to clear the shoulder I am cutting threads up to, and has, at the first pass, enough clearance to be able to cut up to a thou or so from that shoulder on the last spring cut passes. Seeing that I have it set ok, then the tendency to want to speed it up and get a cleaner cut is very hard to resist. But if I do, the thread slides sideways and wrecks the threading op because of the lag between the index pulse and the actual lockup. If we have a true ABZ encoder, it seems to me that the zaccel to lock can be quantized during an air cutting run, and on the next pass, enable the lockup start that far ahead of the index, making adjustments in real time such that the desired position is effectively the index pulse. Further thinking along these lines says we can ask for and get, much more zaccel, since zvel isn't straining anything even at 2x the normal running speed, auto adjusting the z in the run-up space, say 3 turns of the spindle, so that the actual number of steps issued to the motor equals encoder pulses for 3 turns. There is of course more math to it than this simplification, but it does seem 100% do-able. Likewise, similar math could quantify the ballistics of the spindle stop, thereby making the desired depth of a peck tap cycle wrapped around a g33.1 to match up with the reverse timing so the actual stop/reverse will match what you fed to the G33.1 as the turnaround point, by initiating the stop so it coasts to the correct stop/reverse point. Yes, I am aware that a G33.1 does NOT have a stop cycle, unless its running under my my-lathe.hal file. There, the state change of the dir line initiates a dynamically braked stop, and only clocks the new direction thru to the output when the spindle is a very few degrees from truly stopped. :) I could even imagine a lookup table that we would have to fill in the blanks, so it might even hit pretty close on the first pass by looking up the precount from the index, but that would mean we'll have to measure it somehow to fill in the blanks. But it makes more sense to just let the routines measure it and track it themselves. See what happens when I let my imagination out to play without a chaperon? Scary, ain't it. :) 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) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
