Stick the tool out farther.... ? Sounds like you need to adjust your tool parameters and your holder parameters. Adjust the length of the tool in the tool parameters, and adjust how far it sticks out from the face of the toolholder in the holder parameters.

If all you're doing is simple movements like that, then you can do it with the point toolpath. You can G1 to a point, then G0 back to the first point, or to any other point you have drawn. You don't need to modify the post for that.

Stop screwing up your operations so you have to regen them all the time. I kid, I kid. But seriously, stop it.

Everybody has great suggestions and it's true speed comes with time. The best thing to do is practice, practice, practice. Soon enough you'll have your favorite toolpaths and ones you avoid and it all becomes second nature. The things I find most critical for my efficiency are levels, WCS View Manager, and colors. Levels for keeping things organized and quickly hiding/ unhiding geometry, solids, stock, etc View Manager for moving around the different planes Colors for highlighting specific parts of geometry that I know I'm going to be working with a lot. One thing that I seem to do differently is I have almost every toolbar possible (that I use) on my screen at all times. I don't find it cluttered and knowing I'm only a quick mouse move away from what I need makes things fast. Oh, yeah, and A TRACKBALL.

Hey, sign me up. I'm always down for learning anything.

We're ITAR certified as well-- you can share the ITAR part as long as you omit any information about what the part goes to or who you're making it for (and as long as it's not visually obvious, like say, a wing spar.)

I know this might sound silly and I'm just saying it to state the obvious, but have you made sure the lead in / lead out parameters are correct? When I face in my subspindle, it loves to lead "out" right into the face of my part, and I can't change the op defaults, because that changes it for the main spindle. Just one of the many ways I think Lathe is broken. Is this something you're seeing on your screen in Mastercam when it plots the toolpath, or something that looks fine in Mastercam and is all wonky when you post it? If it's only when you post it, there are switches and definitions in the .pst that could be set as default, and it's only when you change things in Mastercam that it overrides it.

I have a future project to do the same thing. I've written macros for damn near every machine in our shop to do things from helical boring, angle boring, hobbing, threadmilling, 3.5 axis orientation, etc... Now i need to get Mastercam to fill out the macro call line. My current project is having Mastercam output that line, while I still program Mastercam like normal so I can run an accurate simulation. Getting it to do one hole is easy; having it be modal (G66/G67) is giving me a problem.

We use a lot of Iscar Pentacut holders on our Okuma LU3000EX lathes; those come in 3/4" shank. Good tools, lots of insert widths, threaders, etc.

I'll take one when you're done!

Are you using the same machine definition for both style machines? If so, are you physically drawing the mcx file for the right-chuck machine differently than the left-chuck machine? On our machines, I don't have to draw things any different if they are operating on different chucks. For our twin-spindle machines, it's much different. If I'm doing right-chuck work and forget to change my axis combination to right, all kinds of hell breaks loose and the tools love to plow through the stock. I would recommend you create a different machine definition for the right spindle machine. That way you can set the machine home points and reference points independently. If you have mill-turn, you could set you machine def as a twin spindle lathe, and then just pick which spindle you are using when you program. Once you have you machine def set up correctly, there should be no difference in operation defaults between the left or right spindle.

If you guys are interested in plant tours, here's my shop I don't know what joker made this video, but the good stuff starts at 3:15. Of course it's a little out of date; we've added two new LU3000EX and two new swiss machines since this video.

"Making Due With What You Have" I agree with Helmut--rigging up that saw seems like more hassle than it's worth if you can get away with emilling it.

We're running a job now doing something similar: hinge pieces for armored vehicles, each hinge half has to have two bronze bearings pressed in. The machine swings a tool to the door and the operator slides a bearing on the tool, and then it presses it in. Afterwards it reams and mills them to size. Of course, this is an HMC so no worry about the bearing sliding off the tool.

::smacks forehead:: I'm an idiot sometimes. The problem with elliptical toolpaths is that they aren't true arcs (in a machining sense) to begin with. You can try to force them into arcs, but the more you force, the less resolution you'll have on your part and it will end up wavy and choppy. Is your machine capable of following NURBS toolpaths?

I was going to suggest removing the solid history as well. If you're on X7, you don't need to use the chook as it's a standard feature in the solid creation menu.

Intersecting a cone with an plane on an angle (or tipping a cone on its side) is the very definition of an ellipse and parabola. Have you tried surface high speed > waterline? If that doesn't work, you can always bust out the good ol' calculator. If you're really feeling fancy, you could try adding the equation to chooks > fplot and have it draw the splines for you, which you could then toolpath. Thats how I draw my P3 polygons. The problem comes with the angle of the cone- when you tip it sideways, does it produce an undercut? If so, then you'll never get a full ellipse, as the back half is being blocked by the top of the cone. Whats an RAH?

They're not broken, they're just finicky. Check surfaces and containment are two different things, they don't do the same exact thing. Changing you containment changes the zone your tool can move within; changing the check just means the tool will stay further away. If you increase the amount left on check, you will get huge fingers because the check zone will be violating your tool path, so the tool will retract to get over it. Try making the check surface smaller (change it to zero just to visualize this.) I bet you'll see the fingers disappear. The reason you have the retract amount is pretty simple: your tool is trying to follow a very linear path- back to front, move over a step, move front to back, etc. When you have geometry that isn't elementary (such as a boss with fillets) there will be parts that intrude into the tool path. Since Mastercam is trying to keep the stepover that you put it, when it encounters this intrusion, the only thing it thinks is "Oh, I've got to go around this." And the only way it can do that is to go up and over. If it tried to move left or right, it would violate your stepover amount (aka, it's not dynamic.) As you add material to leave on check, it essentially grows the size of the boss, meaning it will make even more violations into your toolpath, thus more retract amounts. Another way to visualize this is to make your stepover smaller: this will reduce the area of intrusions, and you'll likely see the retracts disappear. Containment boundaries are exactly as they sound, like fences around your tool. Increasing the size of the boundary lets your tool move in more area; this means it will let more of the boss into the boundary, thus making more violations, thus more moves. Making the fence smaller exludes more of the boss, so less violations, less retracts. If you make your boundary smaller but increase stock left on check, you're defeating yourself and not making any difference. Unfortunately sometimes the violations are things we can't even see. Imperfections in the solid or a mismatched fillet that is so small only a computer would think it's a problem, damaged solids, slight skewing, etc. This seems especially prevalent when you use surfaces, and not solids. I feel the legacy toolpaths are more prone to this kind of thing. Using the newer high speed toolpaths shouldn't make a huge difference in program size. You can adjust your arc filter tolerance to reduce the size of the NC by huge amounts without necessarily sacrificing surface quality. Sometimes just turning the filter on will reduce the file size by half. If you want to, you can upload or email your file so others can play around with it and show you good tricks to overcoming obstacles like this.

What version of Mastercam are you using? I personally tend to avoid the legacy style tool paths and would use the newer surface high speed toolpaths. I feel you have much more control over things like that. For that you could use surface high speed > raster and get basically the same path, only you would be able to select your containment areas and control the retraction factors much more finely. Anyways, usually when I see that, it's a matter of the gap size being too small, and the tool retracting. Try setting your gap size to something large, like 1.000 just to see what happens. Also, if you have that boss selected as a check surface, the toolpath is looking forward and seeing it intersect, so it pulls up and over to avoid it. Try reducing the amount left on check (so the tool is allowed to get closer), or tweak your containment boundary offset so the tool stays farther away.

Go to settings > key mapping > WCS and set a key combination to your desired WCS. For example, I have Top as CTRL+1, Front is CTRL+2, etc. This will change the WCS without changing the view or the T/C plane, you can just use the default combo for those. When you open key mapping you'll see all the different options that are available.

Unfortunately, you don't handle it with the post. I was a bit upset when the reseller touted all the awesomeness of syncing spindles and turrets with Code Manager, then said I can only do that if I buy MillTurn. So I sync spindles with a little bit of forethought while I'm programing using change at point > insert code when I need to have turrets working together, for things like pinch-turning. It sucks because you can't see it in verify or simulation. However nine times out of ten you don't need a point-to-point sync, and I just insert the sync codes using Manual Entry before the toolpath. When we're setting up a new job (which we do all the time being a job shop) I take my laptop on the floor and watch the operator set up. When I need to make a change I do it right there and repost using wifi directly to the machine.

We're big fans here, too. We have 13 lathes and a HMC; with two MacTurns, a Multus, and two new LU-3000 lathes (annoyingly, one has a P200, and the other has a P300). We'll have another big bore by the end of May. Unfortunately, at least in my area, the post is provided by a third party and it's been less than favorable, to put it delicately. I've had to virtually rewrite the .pst to make it work correctly. Since it's a third party I can't afford the time it takes to send a change request to my reseller, then wait for them to contact the writer, then wait for the reply yadda yadda yadda. So if you need any help with a LU-3000 C/Y post, I've got plenty of experience.

I've been trying to replicate this error but no success. What is your part geometry?

I just ran through several of my programs and changed the rougher from a 452 to a 431; Mastercam had no problems with the change. What version are you running? Is this a new problem?

I tried your program and it ran smooth and fast in my verify. What are the specs of your computer? Off topic, it sucks that you have to machine that part out of a solid block. I'm glad my shop has three lasers.