Re: [CAD_CAM_EDM_DRO] EMC Lathe mode.

Brian Pitt wrote:

> this aint as easy as it might look at first
>
> the G-Code interpreter would need to be redone in some areas for a
> turning center ,many of the codes do different things and most lathes have
> some extra canned cycles for rough turning and facing that will take several
> passes and leave some stock for the finish pass
> often the parts finished profile is given in a subroutine and the control
> will run that section of code twice ,once to rough it down and again as
> a constant contouring finish pass (another canned cycle)

There are lots of modifications added by some builders, but, in fact, the
standard G-codes were developed so, that they would be universal.
The canned cycles are mostly different, but canned cycles are not a
standardized feature of G-code, and are meant to be used in a machine-
dependent manner.

>
> tool nose radius comp is done slightly differently and the
> standard 8 TNR vectors would need to be added to the tool table ,
> maybe in place of some of the extra work offsets (lathes only use one)
> also many lathes do not use M03 to change tools,you first move to
> a preset tool change position (X positive limit) and just use T020202
> (tool 2 , pocket 2 , offset 2) calling that tool with a 2nd offset would be
> T020222

M03 is turn spindle clockwise. I don't know of any machine that uses
M03 for tool change. The T020202 you mention is also a non-standard
use of the tool word. A D word following the T word provides the correct
offsets for the selected tool.

> the default cutting plane for arcs will need to be set to XZ rather than
> the mills default of XY (naturally)
>
> the spindle encoder should have a fairly high resolution both for threading
> and as the C axis ,the encoder is often driven from a small timing belt pulley
> separate from the spindle drive pulleys
> in fact most mills have a spindle encoder for rigid tapping and tool indexing
> during tool changes
> so you should really think of the spindle as a full blown rotary axis on ANY
> machine tool not just the lathe
> oh yea, even with a high resolution encoder at several thousand RPM the
> encoder counts will only be coming in at a few hundred Khz ,this shouldn't be
> a problem for a reasonably fast computer to keep up with

The computer does not need to take any action on individual encoder counts.
An encoder counter device does that task. The computer only inquires
periodically what the count is, and computes where the Z (and possibly X
on tapered threads) should be.

One change is that the spindle is a 'recycling' axis, while other axes are
not. What I mean is that every so many counts, the spindle is back at
the same position. Some systems arrange for the spindle encoder counter to
reset back to zero every time the index mark comes up. The STG board,
and all others that use the LS7166 and LS7266 encoder counters can't
do this. So, the software needs to compensate for this effect, that the spindle
counter just keeps counting revolution after revolution.

>
> constant surface speed was already mentioned
>
> and I don't remember if EMC has a feed per rev setting or just feed per min

This would be the G33 word.

Jon