Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: AndyMS on June 07, 2015, 06:35:10 PM
-
The effect in the photo occurs at 300 ipm and lessens as I slow the feed rate. I have to slow the feed rate down to 50 ipm before I can get acceptable, clean corners, but that's just way too slow.
Sorry if this has already been discussed at length. I just wasn't sure what keywords to search for.
-
Looks like a mechanical issue maybe? Not sure if I've ever seen anything like that?
-
I don't know about your machine but could it be tool deflection and just before the corner, the machine slow down and the tool have less of it.
Jeff
-
Backlash in the axis'.
Running slower, there isn't enough influence from the rotation of the tool to move the table.
The heavier cuts will displace the amount of backlash.
Possible ?
Russ
-
Hmmm looking again, that notion would require a left hand CCW cutter in the direction shown.
Might be more like the servo tuning issues we've seen earlier.
Are you using servos ?
Russ
-
Can you tell WHICH is the correct cut as just after the turn or after it moves over a step.
SO the FASTER it runs the worse it gets ?
Stepper or servo system ?
Are your velocity and accelleration settings the SAME for both axis's ?
Have you set ANY CV settings ?
Try running the program in exact stop mode.
(;-) TP
-
I'm using steppers.
The cut was CW.
The faster the cut, the worse it gets.
The velocity and acceleration are the sane for both axes.
I have not set any CV settings.
-
Looks like tool deflection; what is the diameter of the bit and how much sticks out of the chuck? To confirm, Stick a marker in the chuck and use your mill like a printer and see if you see the same problem. The small Sharpies work well.
-
It is odd that it happens AFTER each corner and at teh same point in the next LINE, and it is happening in both axis. That looks mechanic and looks like flex OR axis shift(slipping coupler, belt stretch) in both axis. The main indicator is it gets worse with higher speeds . That is when cutting forces grow higher.
Try the sequence in Exact stop mode (G61) OR reverse the direction of the cutting and see IF the shape of the problem changes.
Does the toolpath LOOK correct in Mach3 ?
(;-) TP
(;-) TP
-
Look at the GCode too you should only see one axis vector changes; not multiple axis vector course corrections like you would in a circle.
If the GCode looks ok, look for play (backlash in the screws/couplings/pulleys/motor mounts) as well.
-
Do you have backlash comp enabled? If so, turn it off and retest.
-
BR549 - The correct cut is the line before the shift goes inside the boundary before the turn. The shift occurs just before the turn going CW, not after.
Chaoticone - I'll have to check on the backlash comp
-
does the DRO show the axis moving on the axis that shouldn't be moving?
-
That is NOT what your picture shows .The one with the red arrows. It shows the tool turns the corner (with no rounding as it should) first THEN shifts outward. Then goes to the next corner and repeats the same motion in the opposite axis.
OOPS Sorry that was not your picture (;-)
(;-) TP
-
OK can you post the gcode you used ?
(;-) TP
-
Just looking at it BR549 it looks like tool compensation is being turned of on the corners.
I think you might be right in thinking its in the GCode.
Wes
-
BUT that cannot be as the feedrate is effecting it. It is still pointing to mechanical issues. IF it was in the Gcode it would do it all the time. I just want to double check the Gcode and test it here.
I am with Gerry I have NEVER seen this problem from Mach3 as it defies the motion logic in mach3.
(;-) TP
-
I went home and ran a program under no load, and the issue was still present. I even ran the code in Mach3 with the machine off and ob :)served the shift in the DROs.
I then started studying the gcode. I discovered that these offsets were in the gcode, and each corner approach involves over 30 lines of code. So, it appears the machine is just doing what the code is telling it to do, but why is the gcode doing this?
-
what cad program are you useing
-
:o ;D (;-) TP
-
true