What is the proper procedure to mill a chamfer using V9? Here's the way I've been going about this, if someone has a better way I'd like to hear it. For simplicities sake lets say I have a piece of material thats been milled to .100" thick. Along one edge you need a .06 x 45 deg chamfer. In mastercam I'd offset the edge in .080. I'd do a regular contour on that line and give the depth as an absolute value of .120". This way I'd comp the dia of the tool to what its dia actually is using wear offsets, and set the length to the intersection of the angle and the outside dia. This gives me the best chance at a perfect chamfer from go regardless of how the chamfer tool has been ground at the tip. The only problem with this method is it can't be verified graphically since mastercam thinks the tip of the tool is .020 above the part. I screwed around with the chamfer 2d setup but its not making sense to me and the system help on it isn't very much help. Thanks. Hoping I don't get flamed too badly for this one :-)

I have a few Haas indexers, one of which is on an Okuma. If you don't need 4-axis simultaneous motion ( you're just positioning with the rotary ), then this works fine. The programs do command the rotary to index via an M-code. The only thing the rs232 port will do for you is allow you to upload the index positions rather than typing them in at the indexer. Either way it's something you only do once at setup. If you only have two positions why make things more complicated?

The biggest drawback to on machine measurement is using the machine that makes parts to do something other than make parts. Not many places can afford to tie up equipment like that.

The place I worked prior to where I am now, had two lagun's with vickers acromatic controls. Although the controls were pc based which I liked ( you can dnc via ethernet + they had about 4 meg of program space ). The machines themselves were junk. As the last person said you can't do heavy milling on them, the clips break off the carosel with great regularity which causes you to not be able to use certain positions or continually fix them, and the head sounds like a bucket of bolts. Even though the machines were only a couple of years old we stopped using them except in emergencies, and were thinking of scrapping them. Bottom line is I wouldn't recommend them. In fact I'd recommend avoiding them at all costs. Look for a nice matsuura or johnsford etc.

There seems to be some problems with the graphical setup sheet in nc utils V9. First of all I hardly ever crash this p4 2gig beast unless I try to create a setup sheet. Then it crashes about as much as it works. So I've gotten used to saving just before I attempt to make a setup sheet. I'm thinking that the code in this area of V9 must be weak. Further evidence of this is the fact that any manual entry ops end the setup sheet prematurely. Anybody else have problems with this part of MC?

My 2 cents. If I understand correctly you are milling at 20ipm with a ball end mill over some sculpted surface. One thing you can look at is something we used to refer to as data starvation. If you have a zillion little short xyz-xyz moves then depending upon how large the look ahead buffer is on your control the machine may physically arrive at the last location in the buffer before it's replenished. Therefore the machine has to keep momentarily wait for the control to catch up. It may be hard to see but usually there's a bunch of noise and vibration associated with that. Depending upon the shape of the surface and all the other factors it can produce surface anomolies such as you describe. Wait that was more like 4 cents worth. :-)

Thanks guys. I had figured out the part about importing the prm files. But as you say they don't have any geometry associated, so they are pretty much the feeds and speeds and maybe a couple other things of that nature. My problem is that I wasn't the person that made the v6 files. So its a little hard to tell exactly what geometry was used and how cuts were offset etc without an in depth analysis of the old programs and backplotting the old nci files ( which I do have ). Kind of a pain in the a$$ but if everything was easy anybody could do it :-)

I'm sorry to be told that :-) But thanks for answering my question. Hopefully I can leave most of the old jobs alone, and the ones that I have to make changes to, I'll just have to chip away at them as they rear thier ugly heads.

Whats the procedure to change over existing lathe jobs that were done in V6 to V9? I've converted my posts and can of course read the pre 9 ge3 files. But can I import tools and operations? Thanks.

It's too bad you hadn't programmed from the center and top of the table and left the post to trig out trunion heights. Now it sounds like your cheapest solution is to modify your posts with a distance value the represents only the difference from what your trunion height was originally. Although it would be quick and you could repost as nessesary, one of the drawbacks will be new jobs. You would then either have to continue to layout your geometry using the old figures forever or create a seperate post for new work either way kinda sucks since its becomming a cluster f#*k. Good luck.

I may be way off from what you're talking about but let me share this and see if it makes sense. I had similar things happen with arcs back in the day when I used to use Unigraphics and APT etc. What it boiled down to was this. Even though a circle looks like a circle from above and below the system you use has to store its cannonical description of your geometry in some sort of database which of course is not graphical. In the cannonical definition of a circle buried in your processor ( in this case MasterCam ) it would be described as a point xyz, a radius, and a vector ijk. This vector tells us what the angular orientation of the arc is for example if the arc were created in the xy plane the vector would be 0,0,1 pointing up in the z axis. People often create geometry from different views and so the direction vector of the arc centerline can be pointing say straight up in z in the case of an xy plane arc or straight down. Either way the arc looks the same represented graphically by your system. You might want to try flipping the arcs or recreating them from the same perspective. That way all thier direction vectors in the database will be the same and thus any translation you are doing from one system to another will be consistent. Like I said I may be way off base. Good luck.

Thanks Andrew. Tried that but it doesn't help. I'm talking about how after each side of the groove is finished and the tool overlaps at the bottom it then retracts at feedrate on a 45 for some predetermined value ( probably hardcoded in mc) like .030 or so in XZ, then it rapids out straight. I don't want this little 45 feed out move at the end of the overlap. I don't think we can control it and its bugging the s*$t out of me.

Jeff...backoff percentage is only available in the groove rough parameters. I'm not roughing in this case so roughing parameters are unselectable. We're talking about the finish passes only. Andrew thanks I tried that to no avail. I activated the ref point turned off approach, turned on retract, turned on X, set it to 1.0, turned off Z and set it to incremental. Path looks exactly the way it did with the 45 retracts at the end. Did I do something wrong?

Fellow V9 users, I love the groove lathe toolparth options, since they work pretty much the way I normally did grooves ( the plunging and finish contouring from both sides down to overlap the middle ). All I want to know is how to keep it from trying to retract from the bottom of the finish overlap at a 45 deg? You can control the entry of each side of the finish pass but I see no way of controlling the "lead out". I don't always have the room to retract at a 45 I'd rather it be perp. Thanks in advance.

Marc and whoever else it may concern, My biggest mistake was thinking that people were following direction. It was a case of "we know we used this kinda tap last time why isn't it working". It wasn't working because my setup man wasn't reading the process instructions, and was using the tap drill for a standard 10-32 tap. With a thread form tap ( roll tap ) the actual tap drill size is about .011 bigger in dia than the largest minor dia tap for a standard. Once I discovered this we never broke another tap. Bottom line is we tapped about 150 holes in Hast-X with a single tap at 500rpm and a fcode of 15.625!!!!. Rigid tapping, using a gh3 cheap xxxx bullxxxx uncoated roll tap and some crappy tapping fluid. Thanks for everybodys advice. Keep the chips flyin and the parts in the usa.

Anybody with actual experience tapping Hastelloy X? I need whatever knowledge you have on this. Here's our requirement... Material AMS5754K ( Hast X ) (1) Hole .1596-.1675 minor dia, depth .36-.40 .1900-32 UNJF-3B depth .25 min csk 90 deg to .21-.23 Tried a couple different things ie. floating and rigid tapping. Different tap styles incl roll tap. Different fluids. Feeds and speeds etc. Anybody have any suggestions? What is supposed to work on this? I have a couple hundred of these to make so hand tapping isn't very attractive. Thanks in advance for your help. V9 rules! :-)

What is "unacceptable" about the results that you get?

Thanks for posting the question about stock view. I didn't even know it was there until I read it. It's actually pretty cool if you define your stock boundaries ( which I do ). It's sort of like verify except you don't watch the tool path. Instead you get an instant depiction of your part at each operation defined as you step through them. If you've ever used verify you might like stock view more. It gives you what amounts to a shaded surface of rev for each operation which can be dynamically rotated for viewing ( something which is a pain in the butt in verify ). Thanks, I almost got through this day without learning something new.

Here's my two cents for what they're worth. When setting up dnc its always best to debug by punching back to the dnc computer rather than trying to read in at the control. The pc will accept just about any garbage you toss at it but cnc controls won't. Anything outside the range of characters the control will accept can cause some alarm or just sit there and flash lsk or something equally annoying. If you send back to the dnc computer and get a file that has a bunch of seemingly random characters, it would usually be an indication of baud. I think MC actually runs mcomm.exe and there is a corresponding mcomm.cfg file that holds the parameters. If you can't find a copy of your MC8 mcomm.cfg then go out to a machine and see what its set to for baud. Good luck!

If all you want is to know where the top or center of the tooling ball is, then I would do it in 2D. Just draw the cross section of your cone and create and arc tangent to 2 lines.

does anybody know if its possible to have the tool manager panel popup from an icon in the toolbar without the lameness of recording a macro to do it? There doesn't seem to be a command to get directly there...it looks like you have to go thru the job setup screen to get to the tool list.