Question about soft limits

Are the soft limits defined in the ini file relative to the machines G53 absolute 0,0,0
which I assume is established when the axis are homed.

Excellent timing. I was just wondering about this very same question earlier, today while setting them on a Mazak and thinking about my Emco project.

Yes, The soft limits are defined in the ini file. They are coordinants from your home position.
When you issue a G53 x0 y0 z0.... the machine will move to home regardless of any G54 etc or tool offsets.

Even if you don't have home switches you can have a defined parking place.
Issue G53 x0 y0 z0 before you shut down. The next time you startup just turn power on and home each axis. Nothing will move but the current position will be 'Home'.

joekline9 wrote:

Yes, The soft limits are defined in the ini file. They are coordinants from your home position.
When you issue a G53 x0 y0 z0.... the machine will move to home regardless of any G54 etc or tool offsets.
Even if you don't have home switches you can have a defined parking place.
Issue G53 x0 y0 z0 before you shut down. The next time you startup just turn power on and home each axis. Nothing will move but the current position will be 'Home'.

Hmmmm... that might be what Gary was referring to but isn't what I was referring to... That's just G53 0,0 to me. Possibly I am using the wrong vernacular, here. By soft limit I am referring to a soft envelope of operation for the machine that can keep it ... ... let's say... from running into the chuck or jaws while operating. So the machine even knows where the steps of the chuck jaws are while running and will alarm out if a move is commanded that will cause it to hit. Or when using the tail and a center, it knows those are present and not to hit them. Those kinds of soft limits. In the case of a machining center, it might be an area occupied by a fourth axis. So, one might be working on the part, and the part is 5" long... One could set a soft limit of G54 X5.200 ( assuming that the part end is X0 ) (( which could also be translated into a G53 position )) and the machine would not move past that and into the chuck or fixture.

Zahnrad wrote:

By soft limit I am referring to a soft envelope of operation for the machine that can keep it ... ... let's say... from running into the chuck or jaws while operating.

No, as far as I know, the soft limits set a limit on axis travel only, and make no
adjustment for tooling. So, it can be set to keep the lathe carriage from striking
the end of the bed, but it can't prevent tools from hitting the chuck, for example.
It doesn't adjust for the Z length of tooling, so if you set it to keep a lathe
cutter from hitting the chuck, it would not prevent a long boring bar from
hitting the chuck. This is a problem, as you might be boring a piece that
is back from the face of the jaws.

I think some fancy turning centers have complicated safety zones that can
prevent such crashes.

Jon

jmelson wrote:

It doesn't adjust for the Z length of tooling, so if you set it to keep a lathe
cutter from hitting the chuck, it would not prevent a long boring bar from
hitting the chuck. This is a problem, as you might be boring a piece that
is back from the face of the jaws.

Exactly.

It pops up more frequently than one might expect for exactly the scenario you describe as well as others. Think of having a workpiece that requires boring deeply ( possibly past the jaws ). One needs to keep the boring bar long enough to reach down in there. But, now it is possible that while simply turning the OD, that the OD turning tool's location in the turret combined with turning the OD almost all the way to the jaws will make the extended boring bar ( that is rotated to another place rotationally on the turret ) possible to run into the chuck or bulkhead just behind the chuck.

I actually did not expect LCNC to deal with this but was curious if it did, could, or would. Having the soft envelope is nice just the same.

andypugh wrote:

I think that it might be possible to add this feature, though I don't know how practically useful it is.

It is relatively simple conceptually to determine if the tool-tip is inside or outside a volume, for example one defined by an STL file.

But boring deep in the jaws shows a problem with this simple concept, as a boring tool is allowed to do that, but not a turning tool.

This has been discussed a few times, but the difficulty of performing these interference
calculations in real time on every move is probably still a lot of CPU activity.

But, it certainly could be done, at least in a fairly simplistic way, with some
2D boxes to avoid. But, it isn't foolproof, as putting the wrong tool on the
turret or entering the wrong tool length would defeat the protection.

Jon

jmelson wrote:

This has been discussed a few times, but the difficulty of performing these interference
calculations in real time on every move is probably still a lot of CPU activity.

But, it certainly could be done, at least in a fairly simplistic way, with some
2D boxes to avoid. But, it isn't foolproof, as putting the wrong tool on the
turret or entering the wrong tool length would defeat the protection.

Jon

To be sure, I am VERY ignorant about what's required, but I have to believe it should be more simple than first imagined as industrial controls that do this are notoriously LOW on CPU power, speed, and memory. I can tell you that it does indeed keep track by tool definition ( outside, inside, RH, LH, tip radius, cutting angle, relief angle, etc... ) and shape, as well as tool tip position.