Help! Something is horribly wrong!

[harolda]

What control software are you useing? Try g80 at the begining of all
programs to cancel any offsets.g90 allso at the begining

[touser]

ger21, if you are referring to the top left of the screen where all the g-code is loaded then no, i dont see a G91 anywhere. Also i just tried adding G80 and G90 directly after it to the code and it worked perfectly again, but i had it going at a feedrate of 10 and it was painfully slow, so i upped the feedrate to 30 on the second try and the machine freaked out and didnt work right at all, so i returned the feedrate back to 10 and it went back to the same problem as the original post. Its very strange.

[MrBean]

Do you have software backlash compensation turned on? I had a load of trouble when running with software comp. Stuff cutting in the wrong places. Especially when using Feed rate override.

Just a thought.

Regards Terry.....

[ger21]

Originally Posted by touser ger21, if you are referring to the top left of the screen where all the g-code is loaded then no, i dont see a G91 anywhere.

No, across the top of the screen in the center. To the right of the "NEW" button.

[touser]

Well, the problem appeared to be the computer, (an athlon xp 2400+). I switched the software over to my laptop (pentium 4 2.8) and everything suddenly workes (almost) perfectly. It drew my squares perfectly the first time at a feed of 30ipm as seen in the pic below:

However, when i tried to draw a simple face, i got this every time. I drew over the same face several times repositioning the start point manually every time which is why the lines are off in some areas. The outside circle is not only skrewed up on top, but it also isn't centered correctly with the rest of the face, it is slightly high.

My question now again, is the g-code faulty or is my machine/controller to blame? Thank you all again for the help i really appreciate it!

G-code for the face:
M3
G80
G90
G00 Z0.05
X2.
G01 Z0. F5
G02 X-2. R2. F28
X2. R2.
G00 Z0.05
X1.3 Y0.8
G01 Z0. F5
G03 X0.3 R0.5 F28
X1.3 R0.5
G00 Z0.05
X-0.3
G01 Z0. F5
G03 X-1.3 R0.5 F28
X-0.3 R0.5
G00 Z0.05
X-0.4323 Y1.7031
G01 Z0. F5
X-0.1719 Y1.2969 F28
G00 Z0.05
X0.4115 Y1.7083
G01 Z0. F5
X0.1927 Y1.3021 F28
G00 Z0.05
X0.7314 Y-0.911
G01 Z0. F5
G03 X0.7121 Y-0.6432 R0.3457 F28
X0.6095 Y-0.5086 R0.4708
X0.4487 Y-0.3993 R0.6886
X0.2462 Y-0.3277 R0.933
X-0.1761 Y-0.314 R1.1057
X-0.2482 Y-0.3281 R1.0984
X-0.4479 Y-0.3989 R0.9338
X-0.6092 Y-0.5083 R0.6894
X-0.7119 Y-0.6427 R0.4712
X-0.7318 Y-0.9099 R0.3458
X-0.7113 Y-0.9585 R0.3492
X-0.6095 Y-1.0914 R0.4708
X-0.4487 Y-1.2007 R0.6886
X-0.2462 Y-1.2723 R0.933
X0.1761 Y-1.286 R1.1057
X0.2482 Y-1.2719 R1.0984
X0.4479 Y-1.2011 R0.9338
X0.6092 Y-1.0917 R0.6894
X0.7108 Y-0.9594 R0.4712
X0.7314 Y-0.911 R0.3511
G00 Z0.05
M5

[ger21]

It seems fine in Mach2 on my system. Are you losing steps, maybe?

[touser]

Ger21, i'm not sure if i'm losing steps or not. It repeated the same thing exactly the same way every time which made me belive that maybe it was the code sense if i was losing steps wouldnt it be more irratic?

[Bloy2004]

...just my thought...is there or are there any background programs running on a time schedule that might be reacting with Mach2 ?
This is so repeatable and lasts only for a short distance and then after the "shift" all seems to work again only in an offset position.

..But why does it occur at the same time every run?....I ask.

My test here works fine...

[ger21]

There might be some binding on the machine in that area, although if there were you would probably see it at the bottom of the circle too. Maybe not, though.

[lerman]

It is NOT just losing steps. First note that the circle seems to finish at the proper place in the X axis. Look at the top where it goes around the circle. (I'm assuming that the large circle is drawn in a clockwise direction.) At the top, there is a sudden downward dip. That would require that steps be added in the negative Y direction.

Describing what appears to be happening in more detail:

Starting at about 3 o'clock, a clockwise circle is drawn. At around 11 o'clock, there are some extraneous negative pulses on the Y axis. Then there is a period of no pulses on the Y axis. Then some normal operation (notice the smooth curve). There is another period of what appear to be extraneous negative Y pulses. Then there is normal operation until the circle completes. The two periods of negative Y pulses are approximately equal to the offset in Y between the starting and ending points. The X axis seems to be functioning normally.

Why don't you try switching the X and Y motor connections and see if you get the same figure?

(If the circle had been drawn in a counter clockwise direction -- the above would be wrong. But one could do a similar analysis.)

Ken

[touser]

Thank you all for the help, i really appreciate it! I went down to the machine today expecting the worst. Put a piece of paper in and with no changes sense yesterday, it worked perfectly the first try. So i ran it again expecting it to fail, perfect the second time. The only thing that has changed sense yesterday is a reboot on the computer and the mill cycled on and off. Now i'm really confused .

[ger21]

If it does it again, try increasing the debounce setting. You might be getting noise, possibly.