arc calculation error

[mmitpw]

Need some help with a post.

 

I am getting small differences in calculated arcs that are erroring out at the machine.

 

For example the following moves should have the same arc:

 

N436 G3 X1.5036 Y-1.7636 R.0434 (from X1.4284 Y-1.8067)

N438 X1.4284 Y-1.8067 R.0433

 

The .0001 difference gives an overflow error.

Actual arc is .043338

so .0434 is fine but not the .0433 arc.

 

I am not sure if I should be changing a tolerance setting in the post processor

or if the problem is in the MasterCam toolpath.

 

Using MasterCam X3, generic mpfan.pst, Tiger IV controller.

 

Thanks in advance for any insight.

Jeff

[MotorCityMinion]

I'm not post editing guy nor am I familiar with the Tiger IV control. Perhaps try this:

 

The MPFAN post can handle this, not sure if your control supports IJK output...

 

Settings, Machine Definition, Control definition, Arc, set the XY, xz, yz planes to Delta start to center. This should produce IJK's for your output.

 

Play with the break at quadrants, 180., or Allow 360 arcs.

 

For more detailed info, search the forum as the instuctions above have been covered repeatedly in greater detail than I provided.

[mmitpw]

I tried IJK output, same overflow error.

Control prefers R output.

 

I assume it's some type of rounding error.

The arc should round down to .0433, but this makes the arc is too short.

If I add .0001 to the arc value, error goes away until it encounters another.

There are multiple situtations in the post where this occurs.

 

Truncating the arc value won't work, I need to round up for all values.

Arc needs to be > than 1/2 of point to point move.

 

Thanks for your thoughts.

Jeff

[Allan]

The issue is most likely in your Control def.

Check the NC precision set it to .0001; i've seen this behaviour when the value is set tighter than your output i.e. .00001

 

Allan

[mmitpw]

NC precision already set to .0001

 

What are the rounding rules in general?

I assume .04335 is rounded up to .0434

where .04334 is rounded down to .0433.

 

I need to a force round up for all cases.

Is this even possible?

 

Or do I need to edit the code each time to add .0001?

 

Thanks,

Jeff

[kh012]

I have run into this prob. some times with the tool size. say a 3mm em (.1181).

it will try and round up and down the .00005 rad. value of the .0001.

i change the tool size to an even # (.1180) and re generate. this usually cleans up

the prg.

 

HTH Ken

[mmitpw]

Ok, I'll try that.

We are using a 4mm endmill.

Although, I believe we've had this same problem with Inch tooling as well.

Thanks.

[mmitpw]

Sorry to say that didn't help either.

Tried going up to .1580 dia. and down to .1560 dia.

Still have the slight error in arc calculations.

Adding a tenth to the arc fixes the error,

so I know it's a rounding problem.

Any other thoughts?