Tyler, What model of Fanuc is acheiving this "CNC type accuracy". I'm planning a meeting with Fanuc in Mississauga and will be taking our FARO arm to test out some of their machines. I would like to have a better series of model specific questions to ask them... Specifically about raw computer to arm accuracy - before any filtering or adjustment is made. How did your meeting go at Fanuc, was the robot as accurate as they claimed?

I've used Mastercam in conjunction with Robotmaster for a few years to program a 6-axis Motoman HP50 waterjet cutter. The interface of the two programs is great and I'm usually only limited by my imagination when creating paths... And the whole system continues to improve. BUT Speaking to the accuracy of the system: We're really limited to what the robots can achieve. Even with an aftermarket filter, over a meter cubed envelop we're lucky if we see +/-2mm in a single configuration. Start working on a lot of different planes and the +/- grows and grows. The strong point of the robot is its repeatability... So, basically you need to cut safe, CMM the part, alter your CAD to suit and do a final cut. That's how we achieve the tolerances that we need. Once you have to program set up, the robot's repeatability won't let you down and a production part run can be made.

I've noticed this same problem... Makes for much less efficient paths when you're roughing out material that doesn't necessarily need a climb or conventional specific path.

This may not be what you're looking for but... From the op manager open the "Properties" tree... Enter the "Files" Section. From there go to the "Operation Defaults" section and click the exclamation point. This will open the "Edit Operation Defaults" window. In there you have another "Properties" tree... From there, go into the "Tools" section and change your settings. We had some luck with changing our defaults this way... hope that helps. -Jake

We use a 6-axis Motoman robot to waterjet cut. The programming is done via Mastercam and we have the Robotmaster plugin to aid in posting and simulate toolpaths. As far as Robotmaster is concerned, we could not achieve the production that we do without it. We have a drill tool station that the robot can pickup and use, and with the correct attachments you could have many tools, depending on rigidity and design. Our Robot isn't nearly rigid or accurate enough to machine metal, but there are other robots out there that work with robotmaster. For our purposes, holding the waterjet paser, the robot works great. I would suggest contacting Robotmaster about your application to see if they can help you.

If you check your nethasp.ini check to see if the NH_USE_BROADCAST = Disabled. We had some big problems at our facility with Mastercam opening and just brousing for the first license that it found... From the sounds of it you only have one license on your network, so this may not be a problem. I had a similar problem with people checking out my 5axis license that was fixed by adding the "no" command line switch to their shortcuts. I think that the "/noL" switch will stop a copy from grabbing a lathe seat.

I tried the above formula (5/4)+1 and it would not work, that said, at this point division won't work at all... Not even 1/1, but division did work in the past. I tried (5*4)+1 , (5*4)+(2*2) ... and that worked fine. Again though, as soon as a decimal is introduced the cell produces an error. I'll try a division formula again after a restart and see if it makes a difference.

I've been messing around with some post files that we have here at our facility with varying degrees of success. I wish I understood better how the files worked, or even where to get more information about writing posts. How would I go about educating myself on the topic. What is the scripting language? Are there places that can teach you how to write? Any information would be appreciated.

It took a while today for this error to occur, but it did. I was able to try the following equations: 1.1+1 , 1+1.1 , 1+1+1+1.1 , 1.1+1+1+1+1 and so on... I tried this with all simple functions (+,-,/,*) and every time I got the same error. I tried mixing up the number and signs in the equation. Every example would work without the decimal, but as soon as a decimal was added anywhere in the cell I would get the error.

Just an update on my issue... The .dll reg helps, but after a certain point I have to re-reg the file. I'm going to send the issue to QA and see what they say.

Roger, thanks a million... Registering the DLL fixed the issue right up. -mada, I'm also using 64bit windows, you should try Roger's fix of registering the MCEval.dll. Before running this reg I wasn't even able to calcutlate 1.1+1 in the fields.

Is it just me, or are all the "Math Capable" boxes terrible at math... It seems that they can do simple math, but as soon as a decimal is introduced it spits out an error -1.0E38 to +1.0E38. Has the math functionality gotten worse, or am I doing something wrong.

More to this for my case... I just found Miscellaneous Values switch that sets my initial primary axis bias to a positive or a negative. It looks like this in my post: # mi4 - Start initial primary rotary axis bias # +/-999 represents start as close to limit as possible # 0 represents calculate without using bias # Any other value represents an angle in degrees to attempt # to position near. # -999 = Low, 0 = Off/Default, 999 = Hi, Value = Angle bias I don't see where the post calls this value for output, but it works.

The machine config settings in my machine def seem to have no effect on the output of my NC files. In my post there is a Nutating bias switch between + and - that (I think) should tell the post to use the -90 or 270 coordinate... It doesn't change a thing for my output, but may help you.

I've been trying out the "Convert to 5 axis" paths with some success on the screen... I was having problems with my terrible Onsrud post though. I think that I have it sorted now and am eager to try some new programs on the machine. -Thanks for all the help