Long Programs??

[crazy^millman]

Well here is the end of the NC code for one my current OP2 roughing programs. I ran a filter on most of the HST toolpaths. Had I not I think it could be almost double.

N11079124 X-26.3532 Y-6.2399 Z4.5677
N11079126 X-26.3569 Y-6.2414 Z4.5685
N11079128 X-26.3604 Y-6.2433 Z4.5694
N11079130 X-26.3638 Y-6.2454 Z4.5702
N11079132 X-26.3669 Y-6.2478 Z4.571
N11079134 X-26.3698 Y-6.2506 Z4.5719
N11079136 X-26.3724 Y-6.2535 Z4.5727
N11079138 G40 X-26.3386 Y-6.281
N11079140 X-26.3724
N11079142 G0 Z12.
N11079144 TRAFOOF
N11079146 M09
N11079148 M5
N11079150 SUPA G0 Z=0.000 D0
N11079152 SUPA G0 A=0.000 C=0.000
N11079154 M30
%

Yes that is 11 million for the N numbers counting by 2. over 5.5 million lines of code. What do others run into for NC code lengths?

[Guest]

Biggest  nc program I have ever has was 12 megs in size......don't remember what the line count was though

[crazy^millman]

19 minutes ago, JParis said:

Biggest  nc program I have ever has was 12 megs in size......don't remember what the line count was though

This one is 165mb in size.

[Guest]

14 minutes ago, C^Millman said:

This one is 165mb in size.

[Mick]

I had an NC file once that was 210mb, and that was without line numbers. It was a huge wax die, that almost filled the table of the machine, with lots of fine detail.

I pushed the button at 4pm on a Friday, and came in on Monday morning, and it was all done, with no issues. I sure felt relieved.

[jlw™]

Biggest one I've had was 68mb.  I do not know line count as I only output N# at safe tool changes for that machine.

[jeff]

You gonna single block that?

[Guest]

4 minutes ago, jeff said:

You gonna single block that?

Don't laugh, on t he 12 meg program I reference, they guy actually came to me and said he couldn't follow the code....that it was processing to fast and he wanted to single block it so he could watch it

Edited July 14, 2017 by Guest

[Grievous]

That's a big one...lol.. You're running that program just like that?...I mean I think u gonna put all that code in a sub and call it external, no?

[F10Brandon]

1 hour ago, JParis said:

Don't laugh, on t he 12 meg program I reference, they guy actually came to me and said he couldn't follow the code....that it was processing to fast and he wanted to single block it so he could watch it

I worked with a guy once that would single block every line. He would check over and over again with a gauge block any time a tool would come to the feed plane (0.1"). Slightly frustrating to watch considering it was weld clearance haha. 

[gcode]

45 meg is my personal best

 

[MIL-TFP-41]

Wow...how long did that take to post? I know on some of the larger programs Ive done, posting time was to the point where you thought mcam had stopped responding. (was in earlier versions of X...maybe X3 or X5??)

[jlw™]

I will usually wait until I go home to hit post, same with regen once I'm some what happy with the toolpath parameters and I have a very nice rig.

[gcode]

Anybody using batch posting for big files

[jlw™]

Never even played with it.  If it works similar to Epsrit... stay away from it.

[?Mark]

A million lines of code is my personal best/worst. Depending how you look at it

That's 28.3 MB (80 % 5axis)

[crazy^millman]

About 10 minutes to post that on my Laptop. It takes Vericut about 10 hours to verify that much code in the 8.0.3 version with everything set to max in it. Machine sim can do it and gives me a model in about 40 minutes. I have got about 99% confidence in Machine sim for this machine that what I see there will be what they see on the machine. I have that same level of confidence in Vericut.

[?Mark]

I'm still on the fence about the machine sim. Tried it and wasn't too happy, but maybe I didn't give it much of a chance since X9 or so.

Camplete and Vericut together give me almost 100% confidence...almost

[jeff]

Wouldn't that be a paltry 5.5 million lines of code since the sequence numbers go up by 2 per line? 

[crazy^millman]

20 minutes ago, Mark @ PPG said:

I'm still on the fence about the machine sim. Tried it and wasn't too happy, but maybe I didn't give it much of a chance since X9 or so.

Camplete and Vericut together give me almost 100% confidence...almost

Here is my issue with using anything outside of the CAM Software.  (Tangent swerve (Really wish CNC Software would have adopted Verisurf's machine simulation capabilities that works directly inside of the Mastercam without ever having to go into another software. Verisurf was doing this back in V7. ))

It is outside of the CAM Software. I want to check a clearance or a collision last thing I want to to is run an interface, or import an NCI or NC code to go check that. I have access to CAMPLETE and Vericut and both are great software, but neither give me 100% confidence. I am only 100% confident in anything once it has been run and proven out. Not one of them take material, tools, tool wear, and 500 other factors into account that running the part on the machine does. I might have a 100% confidence in some thing one day I am not ruling that out. I have programs that run for 80 to 600 hours and they never have the 1st problem. I think that is awesome, but sure enough on hour 81 or hour 601 there is a problem that nothing caught and until something catches those issues will I feel 100% confident. Need to remember I am not making a plate with 4 holes in it.

[Newbeeee™]

5 hours ago, C^Millman said:

 Need to remember I am not making a plate with 4 holes in it.

...and there lies your problem Ron - you're a masochist!

[mkd]

Years ago a tweaked a post to remove line numbers and the forth decimal place. Absolutly no need for that last decimal unless you're finishing and even then it's questionable on ~80% of the work out there.

 

[crazy^millman]

5 minutes ago, mkd said:

Years ago a tweaked a post to remove line numbers and the forth decimal place. Absolutly no need for that last decimal unless you're finishing and even then it's questionable on ~80% of the work out there.

 

I am good with that. Ask the customer to accept that and I will be good to go.

[White Feather]

I don't actually remember file size or how many lines of code... but once upon a time I did a lot of stuff for a 5 axis gantry which was approx. 60ft long in X and prolly 20 ft long in Y.. these were large panels for the SLS fuselage mostly 3 axis but with some fine detail work in 5 axis as well as 3d work with fine finish.

 

The larger of the files would take  an hour or better to post in Catia, and the vericut sims I would leave running overnight and into the next day.

 

I loved those parts.. post out a roughing program and let it run for weeks at a time.

[Leon82]

I had one 25 meg. Never loaded the whole thing though.

 

We only use n numbers on the tool change line