Mind offically blown

[jlw™]

I've done that! Actually, I still have two tape punches in my garage. Would you like some sample tape punched?

 

Don't want anything to do with it! Last job hd 4 monarchs that ran tape and we were upgrading them to some kind of drip feed that took SD cards... well worth the investment!

[Mick]

Don't want anything to do with it! Last job hd 4 monarchs that ran tape and we were upgrading them to some kind of drip feed that took SD cards... well worth the investment!

 

Damn, and I was so looking forward to getting one out and punching out some tape... not.

 

[civiceg]

The only issues I have seen with R values is incorrect control definitions in Mastercam. I will take the smaller program files any day. No problem here interpolating holes within +/- .0002. 

[civiceg]

Dynamic tool paths too long.......filter the code. I had a million line tool path today...filtered it and it came out less than 20k lines.

 

edit: that was not the only trick to reduce the file size. 

[ahaslam]

the problem is the 3 place R values are larger than the X Y move.  That shifts the center of each arc.  If you measure the bore in the Y direction it is probably to size but smaller across the X direction.

 

That was something I thought as well. Hole is round with in .002"

[pete_hull]

R is the ebonics of G-Code, it's expeditious but lazy, you will eventually get burned by continued use, as I did many years ago and immediately stopped using it. 

[CJep]

R is the ebonics of G-Code,

 

[Matthew Hajicek - Singularity]

Three place decimal and "R" arcs should be just fine.  For hand-coding a garage shop wood router.  Programming in a professional environment should be of a professional level, using as much precision as your equipment supports.  When you're peel milling hardened steel or titanium with a .020 stepover and a few hundred inches a minute feed, an arc error of a couple thou with snap your tool off.

[civiceg]

How can you get an arc error of a few thousandths, with an arc tolerance of .001 or less in the controller parameter? 

[Colin Gilchrist]

I firmly believe that the most dangerous mindset in manufacturing is the "we've always done it this way" mentality. The other is the "we cater to what the operators "like", or are "used to". It is these attitudes that create an environment where "hand-editing" at the control is allowed. This is a terrible precedent, I don't care "how good" your machinists are.

 

The only constant in manufacturing is change. As I've seen on those image macros; you can't use yesterdays tools, with today's processes, and be in business tomorrow...

 

There should only be one "source" for any NC Program, and that should be the operations inside the CAM file, properly processed by a correctly edited post processor.

 

I can't tell you the number of shops I've dealt with over the years that "preferred R values", because it made it easy to read the code on the controller. If you can't calculate a radius from the start point, end point, and center point callout, then you have no business calling yourself a machinist. Any button pusher can read an "R" value, but that doesn't mean they are good at their job. Especially when using R values can lead to the machine "swinging an arc the wrong way around".

[civiceg]

My argument was for file size. The only time I have seen a clover made out of a hole is because a "programmer" had his arc setting wrong. This has been the case in IJK and R value usage. 

 

File size is very big deal to the shop I work at, with no data servers, 2Mb onboard max, and the amount of hand holding it takes to put files on cards. 

[Colin Gilchrist]

My argument was for file size. The only time I have seen a clover made out of a hole is because a "programmer" had his arc setting wrong. This has been the case in IJK and R value usage. 

 

File size is very big deal to the shop I work at, with no data servers, 2Mb onboard max, and the amount of hand holding it takes to put files on cards. 

 

Is there are reason you aren't networking the machines to use DNC besides lack of knowledge or cost? I can't figure out why so many shops out there let machine memory become an issue. I setup my first DNC link to a machine in the late 90's. Unless you have machines that are total dinosaurs, you should be able to take a basically "scrap" computer, and hook it up as a DNC server.

 

What I've done many times in the past is install Mastercam V9 on the old computer, and use the text editor to run the DNC. (You can launch the editor without needing a HASP.) It only takes a cable with three wires, and two DB plugs. (Typically a DB9 for the COM port on the computer, and a DB25 for the machine itself...)

[civiceg]

90 percent of the machines are 5 years old or less. All but 4 have Ethernet. The issue is IT and not granting access. 

 

They are all networked with RS232 on Predator. Because of how long the cables are, a 500 kB file can take 30 minutes to download to some of the machines, if the control has the memory for it. If not it goes on a card or rs232/usb DNC at the machine. The fix is obvious, both predator and the machines support Ethernet. I even tested a machine via my laptop and FTP to make sure there were no issues.

 

Anything I can do to make less code I do. 

[CJep]

90 percent of the machines are 5 years old or less. All but 4 have Ethernet. The issue is IT and not granting access. 

 

They are all networked with RS232 on Predator. Because of how long the cables are, a 500 kB file can take 30 minutes to download to some of the machines, if the control has the memory for it. If not it goes on a card or rs232/usb DNC at the machine. The fix is obvious, both predator and the machines support Ethernet. I even tested a machine via my laptop and FTP to make sure there were no issues.

 

Anything I can do to make less code I do. 

 

Sounds like IT has you drillin' holes in the bottom of the boat to let the water out,..