I saw this problem not long ago and didn't think much of it.

[Müřlıń®]

That sucks to hear about your crash. I would be annoyed, and angry too.

 

I modified my posts to force "dog leg" moves, ie, XY on one line, and then Z on the next.

 

 

+1.....my post has been that way since ver 9 as well. I do not use ref points and always use full retract.

 

Mastercam has all the tools needed to make a good part. But things dont always work the way they should.

 

Sorry to hear about your spindle Husk but it looks like you broke the golden rule and posted an UNverified NCI.

 

I never send code to the machine unless #1: I have observed the NCI backplot.

#2 I have run collision check on the NCI. and #3: I have run graphical verify on the NCI.

 

 

If one does not want to follow these rules then they need to dig deep for another 30k and get a good verify and HOPE it is bug free set up properly.

 

I just got reminded reciently in ProE Manufacturing how important my verify procedure is on a simple contour with depth cuts.

 

The last slice looped back into the wp and crashed my 400$ tool.

 

I had become so comfortable with my system and had processed so many extremely difficult operations 100% error free I just sent the code out...how could something with 6 little contour cuts screw up so badly!!!

 

No damaged spindle but it could have been....I got off lucky...

 

You do not need 3rd party verify...Mastercam comes with all the tools needed to verify a good part. except they took out collision check in 5 which irritated me severly....

 

Pucker factor is always high when upgrading to the next release of Mastercam....that is the part I dislike the most about this software.

[Watcher]

Mastercam comes with all the tools needed to verify a good part.

 

:rofl: :rofl:

 

Husker, I do support 35 NC Programmers around the world with Pro/NC and other tools, like Vericut, CrashGuard Studio, CIMCO, TopSolid, blablabla...

 

I understand the pressure we have in the shop to get parts out of the door and I understand that you expect your CAM system should not melt your spindle because of a bug from the last decade that is being overlooked... Day after day the programming work is getting more complex, so it's hard enough without a nasty bug like that.

 

That being said, IMHO, you did not break any "Golden rule" here. We all know where the fault is.

 

JM2C

[huskermcdoogle]

:rofl: :rofl:

 

Husker, I do support 35 NC Programmers around the world with Pro/NC and other tools, like Vericut, CrashGuard Studio, CIMCO, TopSolid, blablabla...

 

I understand the pressure we have in the shop to get parts out of the door and I understand that you expect your CAM system should not melt your spindle because of a bug from the last decade that is being overlooked... Day after day the programming work is getting more complex, so it's hard enough without a nasty bug like that.

 

That being said, IMHO, you did not break any "Golden rule" here. We all know where the fault is.

 

JM2C

Thanks for the support Looking like just north of 40 grand to get it fixed with a new spindle(+warranty) so we can put a rebuilt spindle on the shelf.

 

I keep having to tell myself that the more I trust Mastercam to keep doing what it is designed to do, the more times I roll the dice. Usually, I get lucky if there is an issue, and happen to catch it. But this time I didn't... This is something that had I handed it off to a less experience operator, that they probably would have caught it. Me being the programmer is the least cautious when proving things out. If it looks like it is going to the right place, I check my Z to go and let it fly... But I know the flow of the program. I know what it is supposed to look like. I don't necessarily watch the code.

 

yada yada yada. I just want the machine to be all fixed.

 

Husker

[robk]

You do not need 3rd party verify...Mastercam comes with all the tools needed to verify a good part.

 

Murlin, are you serious?????

[Müřlıń®]

Murlin, are you serious?????

 

I could always find every gouge with the tools available.

 

Backplot alone would have shown this crash. It would have jumped out and bit you. Everyone looking for the easy button.....there is none.

This is a known bug. Decades old.

should it be fixed? Of course I think so. I would like all bugs fixed.

is that going to happen? Nope....

 

Now I never said that the verification tools within Mastercam are easy to use or the best. They are adequate. For 90% the of the users they are just fine. Now for the small percentage of high end users that machine on 50K workpieces, a gcode verify is in order....I think that goes without saying.

[Matthew Hajicek - Singularity]

Backplot alone would have shown this crash.

 

I think the problem was that he did backplot and verify and it looked good. Then he reopened the file later and posted and it was bad. Should you have to re-check everything every time you open a file?

[huskermcdoogle]

I think the problem was that he did backplot and verify and it looked good. Then he reopened the file later and posted and it was bad. Should you have to re-check everything every time you open a file?

 

 

This is exactly what happened.... Not to mention this portion of the file was untouched from a previous revision...

[butlermere]

I'm relatively new to Mastercam, 3 yrs on lathe, and after using it for a while am of the opinion that Mastercam is ultra cautious when it comes to forcing a regen whenever you do 'any' change(even changing back to an original entry).

The changes done by the post is another area to bear in mind. So an independant simulation, Vericut perhaps, is the closest safety check available short of running line by line on the machine!!

Just my 2cents worth.

[Müřlıń®]

I think the problem was that he did backplot and verify and it looked good. Then he reopened the file later and posted and it was bad. Should you have to re-check everything every time you open a file?

 

Of course not. But he was working with ref points.

a function that had already been proven not to stay put in previous versions.

 

Now I understand the imported ops is nuked and he must rebuild just like he did when he imported from 4.

 

l feel bad to anyone who has a loss like that.

[Namliek_J@TheeCircle™]

I make it a point to recheck everything that gets reposted, doesnt matter if I have done so 100 times. If you would have simply sent the old code out, then that is one thing. But to open it up in a different version, repost, and send it out without any type of checking...... well you see where that got you. Bet you dont ever make that mistake again.

 

Sucks that you may have to replace a spindle, but taking a couple extra minutes to check it out first would have been a lot cheaper.

 

And as I always says, VERICUT everything if you can.

[chip]

I'm kind of old skoll, I use referance points all the time... in lathe too. I don't really set my stock if I don't have to. I always retract and enter with a point.

 

Chip teh V8.1.1.1 HAHA! - where's cop bike guy I have some prayer requests!-

[Matthew Hajicek - Singularity]

Of course not. But he was working with ref points.

a function that had already been proven not to stay put in previous versions.

 

Is this publicly documented somewhere? Is there a warning that pops up telling you that it might not work? If not how is someone supposed to know before getting burned?

[Tyler Robertson]

Is this publicly documented somewhere? Is there a warning that pops up telling you that it might not work? If not how is someone supposed to know before getting burned?

 

yeah I read the forum a fair bit and I didn't know about this

[K2csq7]

One thing I do on the regular (yup, on the regular). Is inside cimco..... "file compare". Simple, easy to use, (free!!!!!) very effective. I have learned over the years.... (all 7 of them) that you will get bit when you least expect it.

ANY time I make changes to an existing program.... reposting from a newer version, for a different machine, adding/removing an operation, or even changing a feedrate or spindle speed....

 

One thing we can all learn from this is that Thee Circle is great.

Sorry about your spindle husker.

[Müřlıń®]

Is this publicly documented somewhere? Is there a warning that pops up telling you that it might not work? If not how is someone supposed to know before getting burned?

 

I'm sure it is somewhere if you could actually get the search to work and search the old arcives you could find something.

 

Go over to CNCZone and do a small search

 

All I know is that were buggy so I stopped using them back around ver 8 or 9. But I don't think I was having the same problem.

All I remember was that I wasted a couple helimills and I think it was on the XYZ move on retract. So I fixed my post and stopped using ref points. Back then I would drill a hole and then plunge that hole to rough a cavity. You could bury those helimills and really move some metal @ 5 IPM. But I used lots of HSS serrated roughers back then too. So I cant remember all the particulars...I slept a couple times since then.

 

But after that I took shallower cuts and fed faster with helical entry like everyone else was doing. I never knew what was wrong. I didnt care, I didnt want to know how the software worked, I just wanted it to do what I told it to do.

 

I think husk has just about nailed down the problem, at a very high cost. And that sux someone had to pay so much to find out.

 

Now Matthew I don't know why you have been picking apart my words. If you think it was CNC software fault that this crash happened, then that is your opinion. I just do not share that seniment.

 

Sometimes I take for granted not everyone has been using this software since ver 5. and they may not have ran across certain things over the years like some of us have.

 

I do think this is a good thread that hopefully 5 years from now won't get lost in archive hell....

[huskermcdoogle]

I'm sure it is somewhere if you could actually get the search to work and search the old arcives you could find something.

 

Go over to CNCZone and do a small search

 

All I know is that were buggy so I stopped using them back around ver 8 or 9. But I don't think I was having the same problem.

All I remember was that I wasted a couple helimills and I think it was on the XYZ move on retract. So I fixed my post and stopped using ref points. Back then I would drill a hole and then plunge that hole to rough a cavity. You could bury those helimills and really move some metal @ 5 IPM. But I used lots of HSS serrated roughers back then too. So I cant remember all the particulars...I slept a couple times since then.

 

But after that I took shallower cuts and fed faster with helical entry like everyone else was doing. I never knew what was wrong. I didnt care, I didnt want to know how the software worked, I just wanted it to do what I told it to do.

 

I think husk has just about nailed down the problem, at a very high cost. And that sux someone had to pay so much to find out.

 

Now Matthew I don't know why you have been picking apart my words. If you think it was CNC software fault that this crash happened, then that is your opinion. I just do not share that seniment.

 

Sometimes I take for granted not everyone has been using this software since ver 5. and they may not have ran across certain things over the years like some of us have.

 

I do think this is a good thread that hopefully 5 years from now won't get lost in archive hell....

 

Quality words Murlin, Quality words.

[Altek]

Just my opinion, I don't think it should be so easy to give Mastercam/CNC Software a pass on this.

 

I haven't been using the software since V5, but I have been a user since V8. Conservatively I've mentioned this bug persisting for 8+ years, but really if I looked up release dates its probably been more like 10.

 

So, for 10 years I'm supposed to just ignore a feature that exists in the software that I actually want to, and should be able to use? Sure I know about the bug, I know what CAN happen, I know how to look for it... but what we users don't know anymore is what gets fixed & when. Without CNC Software publishing a bug/fix list (I'm aware they recently buried one in the latest release if you dig for it, but that's a brand new development) the only options you have are: turn your back on that feature in the software FOREVER, or HOPE that its fixed & use it (again, like you should be able to).

 

I don't disagree with anyone's method of dealing with the problem - to each their own. To me modifying the post is a bad option because I believe outside of the start & end of any tool the motion you see on-screen is exactly what should post in your code. What I disagree with is the principle that a potentially dangerous bug (yes dangerous, it can cause a machine crash - it produces motion different than the values you have set) that has been reported for many, many (many many many) years can so easily be dismissed.

 

For myself, in my experience - I write production programs, aside from making a good part, the other thing I need is efficiency. I can't accept not being able to use reference points. Sometimes what I have running in the machine is scrutinized down to minutes & seconds, as such I cant have a program full of huge, slow (some of our equipment is certainly not the fastest) retract moves. Reference points are a tool I have available in the software that allows me to tailor my programs & minimize wasteful motion. I don't have much choice but to use them to get the motion I want - they're a feature in the software (a basic one at that) to me they should just flat out work. Its hugely disappointing that I have to struggle with this bug constantly, its sad that its experienced by others as well & it has existed this long.

 

-Jason

[Müřlıń®]

Just my opinion, I don't think it should be so easy to give Mastercam/CNC Software a pass on this.

 

I haven't been using the software since V5, but I have been a user since V8. Conservatively I've mentioned this bug persisting for 8+ years, but really if I looked up release dates its probably been more like 10.

 

So, for 10 years I'm supposed to just ignore a feature that exists in the software that I actually want to, and should be able to use? Sure I know about the bug, I know what CAN happen, I know how to look for it... but what we users don't know anymore is what gets fixed & when. Without CNC Software publishing a bug/fix list (I'm aware they recently buried one in the latest release if you dig for it, but that's a brand new development) the only options you have are: turn your back on that feature in the software FOREVER, or HOPE that its fixed & use it (again, like you should be able to).

 

I don't disagree with anyone's method of dealing with the problem - to each their own. To me modifying the post is a bad option because I believe outside of the start & end of any tool the motion you see on-screen is exactly what should post in your code. What I disagree with is the principle that a potentially dangerous bug (yes dangerous, it can cause a machine crash - it produces motion different than the values you have set) that has been reported for many, many (many many many) years can so easily be dismissed.

 

For myself, in my experience - I write production programs, aside from making a good part, the other thing I need is efficiency. I can't accept not being able to use reference points. Sometimes what I have running in the machine is scrutinized down to minutes & seconds, as such I cant have a program full of huge, slow (some of our equipment is certainly not the fastest) retract moves. Reference points are a tool I have available in the software that allows me to tailor my programs & minimize wasteful motion. I don't have much choice but to use them to get the motion I want - they're a feature in the software (a basic one at that) to me they should just flat out work. Its hugely disappointing that I have to struggle with this bug constantly, its sad that its experienced by others as well & it has existed this long.

 

-Jason

 

 

Dude I totally agree about ref points. They are a great tool all I said is that I dont use them.

 

It is because of the type of work that I do. I do not run parts, I am job shop/prototype all the way.

recreating the world every program. I could care less if a program takes an extra hour to run. I just want it to be 100%.

I don't have the skill, time or energy, even the want to, to try to debug mastercam. I like everyone else just go for the easy

button.

 

And at this point is time is full retract, no ref points, absolute retract, kill all doglegs....

If I ran production in todays competitive market, I would be forced to use them to shave every second off cycle times.

 

But you can be damn sure I know they are bugged and I would take great caution proving out the first article.

 

If I thought about it enough I could probably come up with a script that would check them....

It might be pretty tricky but you could compile an exe that prompts you for point locations and examines nc code, then echos out the fitst moves of the program from the ref points to screen so you could double check them.

 

I am sure by now this subject has drawn alot of attention.

[chip]

I'm sure it is somewhere if you could actually get the search to work and search the old arcives you could find something.

 

Go over to CNCZone and do a small search

 

All I know is that were buggy so I stopped using them back around ver 8 or 9. But I don't think I was having the same problem.

All I remember was that I wasted a couple helimills and I think it was on the XYZ move on retract. So I fixed my post and stopped using ref points. Back then I would drill a hole and then plunge that hole to rough a cavity. You could bury those helimills and really move some metal @ 5 IPM. But I used lots of HSS serrated roughers back then too. So I cant remember all the particulars...I slept a couple times since then.

 

But after that I took shallower cuts and fed faster with helical entry like everyone else was doing. I never knew what was wrong. I didnt care, I didnt want to know how the software worked, I just wanted it to do what I told it to do.

 

I think husk has just about nailed down the problem, at a very high cost. And that sux someone had to pay so much to find out.

 

Now Matthew I don't know why you have been picking apart my words. If you think it was CNC software fault that this crash happened, then that is your opinion. I just do not share that seniment.

 

Sometimes I take for granted not everyone has been using this software since ver 5. and they may not have ran across certain things over the years like some of us have.

 

I do think this is a good thread that hopefully 5 years from now won't get lost in archive hell....

 

Im not that old but I've been around since the late 90s. I started on V6. Right about when V8 was being perfected. No regenerated surfaces. Diamond chain... I try to keep up with you guys. I like X4 and I don't get to use X5 much. But my last batch of parts (last month) had all kind of 'glitches'. Simple pocket didn't recognize an island, I re created an arc then it goes..... And actually I had entry point issues too, my entry point wasn't recognized. That's been the worst for me in like 10 years.

 

4 axis points are kind of funny. 5axis drill sometimes gives me three axis moves. I always program with the point drawn in 3d space and use Inc. Makes for fun setting up. 5 axis contours sometimes gives me 3 moves too.

 

Gotta admit... lathe points have been consistant. I like using points for boring bars. Both entry and exit.

 

I'll tell you, I know two "buddies" who have updates but haven't they must be too busy I haven't asked.

 

Tough break really!

[htm]

Mastercam comes with all the tools needed to verify a good part.

 

LOL, good one! Its 2012 and Mastercam verifier still looks like crap!

[Watcher]

LOL, good one! Its 2012 and Mastercam verifier still looks like crap!

 

Quality words htm, Quality words.

 

All jokes aside: Today I went to a TopSolid seminar showing the what's new in the 2012 version... man.. their verification is simply adorable... this year they added the ability to define fixture sets for each operation... let's say you have a 3 axis VMC and a plate where you have some pockets and you have also to contour it... let's suppose it's a coffee tray...

 

You clamp the a square billet with two paralell clamps in the diagonals... machine the pockets and make some holes in the center so you can hold the part with some screws in the 2nd op by removing the clamps in the perifery and machine the O.D. / external contour once the part is clamped by the screws...

 

Well, for the pockets (1st op), select the fixture elements (3D Clamps/Bottom plate) you will use in this step... it will automatically avoid collision with them (You can set a safety distance too) and leave the workpiece untouched in these areas... go to their MachineWorks simulator and the clamps will be shown there along with the workpiece...

 

Create another operation, select the screws and the bottom plate as fixtures and unselect the clamps... machine the O.D. / External...

 

Now go to the MachineWorks simulator for this operation alone... it will show the workpiece in the state it was left by the 1st operation plus the screws...(The area where the clamps were put is un-machined)

 

Now simulate the two distinct operations in a row... MachineWorks simulator show the material in the original state, the clamps... blablabla... as soon as it finishes the 1st operation it shows the second one without the clamps and with the screws + the bottom plate... no air cuts in the workpiece because it knows the raw material condition... no manual setups to set the new fixtures for the 2nd op within the simulator...

 

And now do all this at once in the machine simulation environment...

 

To finish if you want, export the workpiece model in the state it was left after the two operations as a STEP model, Parasolid, whatever... it's not STL or triangulated/tesselated data... (It can be if you want so)

 

Awesome...

 

And yet is not a complete simulation tool like Vericut...

 

Bottom line: MachineWorks have great technology... and they sell it in packages... 2+2=4 ? (BTW: The 3D cut stock model export is TopSolid technology - MachineWorks can only handle STLs at this moment)