Radius to ned of arc differs from radius to start

I am having issue running this code, but I do not see the problem.
Any help is appreciated. I am eliminating some of the in between code, but posting what I think is relevant. If anyone has suggestions or quesitons, I can post more or give specifics on something I may have overlooked.

G17 G20 G40 G49
G80 G90
...POCKET ROUTINE RUNS FINE (G54, G43)...
...I CALL FOR A TOOL CHG (MY MACHINE IS OLDER CONVERTED BP WITH QC30 SPINDLE, SO I DO A G53 MOVE TO MACHINE HOME TO CHANGE MY TOOL)...
G53 G0Z0
G53 G0X0Y0
(MSG,LOAD.375 ENDMILL FINISH)
T2 M06
S3500 M3
G54 X1.277067
G43 H2 Z.1
M8
G0
G1 Z-.708661 F16.
G17 G3 X.864665 Y-1.311024 I-.206201 J0.
G3 X1.277067 Y-1.311024 I.206201 J0.

THE FILE RUNS FINE UNTIL IT IS TO MOVE TO X1.277067 AND I GET THE ERROR MSG.

DO I NEED TO USE G90.1? IS MY G54 MOVE CAUSING THE ISSUE?
ANY HELP IS APPRECIATED. THANKS

Hi

Can you post the whole file?

I can't understand why the file even loads, that code fragment will always throw that error as far as I can see.
It should error when you try to load it initially.

regards

ArcEye wrote:

Hi

Can you post the whole file?

I can't understand why the file even loads, that code fragment will always throw that error as far as I can see.
It should error when you try to load it initially.

regards

This is why I always use the radius mode of G2/G3, but your problem seems to be much
more serious. The Y valuse doesn't change. The X value goes from .864665 to
1.277067, or a change of +0.412402 So, the center needs to be exactly between these
two coordinates, as you are making a 180 degree arc. Your I value of
.206201 looks right! But, note that the Y coordinates in the error message bear
no resemblence to the Y coords in the part of the program shown. I think
that is the clue.

Jon

I fixed the code by adding a Y move within the G54.
Also, I added a G0 preceding the G54.
When I added these 2 items, the file ran fine...No issues.
I think (I am new to what is required) I needed only 1 of these changes to make it run, but I have not had time to check.
If I had simply placed a G0 or G1 preceding the G54, I would have had movement and no error.
Is this assumption correct?
I plan on changing the code back, and adding changes in step wise fashion so I understand. I will likely post again for the next newb that posts something so minor.
What threw me was the error pointing to the G3 line. But, I think (If I am right) that the G54 without a move created the crazy r1 and r2. The 7.4... radius was representative of the G53 home position relative to the G54 G3. I will also verify this tomorrow,
(I am curious that my G54 X0Y0+the X machining coordinates were likely 7+ inches delta).

I apologize if my ignorance has wasted time and keystrokes, but I am trying to get a handle on how the code looks ahead and... So, help is appreciated however it comes.

I fixed the code by adding a Y move within the G54.

That would be my guess, Y was stuck at whatever machine zero equated to in G54 and was not referenced again until the arc moves

I would not have thought the G0 or G1 was necessary, the modality should remain across co-ordinate system changes until another G code of the same group was applied.

Like Jon I very often use radius arcs with G2 and G3, they are much easier to understand and less easy to mess up.
Even full circles just mean splitting into segments.

I still don't understand why the code did not error at the outset, when you loaded it, but I am not properly au fait with the interpreter and look ahead either.

regards

Perhaps...

G54
X1.277067
G43 H2 Z.1

Rick G