Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Bill I on September 22, 2016, 01:45:55 PM
-
Using MACH 3 Ver. R3.043.062, USB Smooth stepper, PMDX 126,134 & Gecko 203V drives.
The motors have been tuned the problem is all cuts are 4 thousands smaller than they should be. The code checks out fine. Is there a setting in MACH that could be causing this problem?
Thanks
Bill I
-
4 thou?
In both directions?? Could be cutter wear, not a 100% accurate tool diameter?
-
The cutter used is new and yes the cut is smaller in both X and Y .
-
Have you done axis calibration using accurate measuring tools over a reasonable distance?
-
Yes axis calibration has been done.
-
What kind of machine, and what are you cutting? Climb or Conventional?
-
Custom built CNC Cutting MDF and plywood I cut in both Climb and Conventional.
This is a new controller I never had this problem with the previous controller.
-
Tool diameter set correctly? Is the tool not the size you think?
-
Using 1/4" cutter 1st thing I checked was tool diameter.
-
Just a query - 4 thou on MDF - thats 0.1mm !
I would not think accuracy on MDF is possible in that range TBH.
I would be inclined to change the tool diameter using the wear factor (if used) by 2thou and try again but I'm still unsure if it is really possible.
-
This is a new controller I never had this problem with the previous controller.
If I would cut a pocket .750 X 1.00 I would get a pocket .750 X 1.00 not smaller.
This happens in what ever material I cut including Acrylic and plastics.
A job I do for a company requires all measurements to be spot on.
x
-
When you say new "controller", what are you referring to?
New drives?
New Motion Controller?
New Software?
If you cut the same part twice, one Climb, and then one Conventional, are they the same size?
-
Hi ger21
Using MACH 3 Ver. R3.043.062, USB Smooth stepper, PMDX 126,134 & Gecko 203V drives.
NEMA 34 640 oz 5.5 amp motors 3-1 belt reduction. motors and belt reduction checked they are fine.
Cut in both clim and conventional all cuts are 4 thousands smaller.
-
What's different from the old setup that worked?
-
Hi ger21
Old controller did not support proxi. switches ,tool changer, spindle speed control also used old Gecko202 drives that had caused problems.
-
Hi Bill I,
when you calibrated the axes who did you go about it?. When I did mine I did the
calculation, the number of stepper pulses/rev times gear reduction time pitch. It worked
fine and have never had need to go back.
Imagine the situation where I had 6mm pitch ballscrews but are in fact 1/4" ie 6.35mm pitch.
The calculation would be persistently and consistantly wrong.
Could your problem be of the same type ie one factor in your steps/unit calculation be slightly
adrift of the actual?
Craig
-
I used the MACH Set steps per unit just like I did with the previous controller.
-
One is left to wonder about some backlash creeping in.
Against that, the X & Y axes both show it.
If you start at 0 and make cuts at 20, 40, 60 mm, where is the error?
If all 3 are short by the same amount wrt 0, then I would look at backlash.
If the gap between each (20 mm) step is short by the same amount, I would look at calibration.
Cheers
Roger
-
Hi rcaffin
Backlash was one of the things that was looked at when I worked with Gecko on this problem.
There is no backlash or bit deflection the resistors were changed and the belts were checked for tension and wear in the belt reduction.
Bill I
-
But what about the measurements I suggested? Measurements give cold hard facts.
Cheers
Roger
PS - where do the resistors come into this, and what are they?
-
As suggested check the steps per unit / calibrate the axes and use the new values for steps per unit.
That should be done "anytime" one changes or adjusts any electronic or mechanical part of the system.
RICH
-
Resistors are used to regulate the amount of amps to a motor from 0 to 7 amps in my case I needed to supply 5.5 to 6 amps.
The resister used was 270K.
Bill I
-
Hi Rich
That was the first thing that was done after installing the new controller and has been re calibrated several times since with no change in steps per inch.
-
How much run out on the spindle and how much backlash.
BTW, you have spindle runout and backlash it's just a matter of degree.
RICH
-
Doing a check out of the "SYSTEM" to find just what is causing an accuracy problem requires a methodical
approach. Yes you do the obvious to start. I like to put what i am looking for into perspective.
So the easy things are done, but, always with the thought that simple things are related to other factors.
MDF, wood, plastic is not exactly a true medium to check tool cutting, different materials, etc all come into play.
Heck, the tooling marks left in the surface can be worth 1 or 2 thou.
Electronics is hard to check as most will not have the equipment to do so. Also simple mechanical
items can be difficult to check. ie; Belt tension can have an affect on the axis motion, is usualy
small amount in terms of steps, same goes for a timing gear, but they are repeatable per revolution
and a bear to isolate components of the system. another simple example is a coupling, other than a solid
coupling,it can affect the motion also depending on applied torque, age, etc.
ALways remember this: What am I measuring and it it the same as the basis!
Just some things to consider before you even start to question what is wrong as compared to before!
RICH
-
MDF, wood, plastic is not exactly a true medium to check tool cutting, different materials, etc all come into play.
Heck, the tooling marks left in the surface can be worth 1 or 2 thou.
RICH
Exactly what i said earlier, 4thou on MDF, I would be happy to hit that on aluminium :)
Could be a slightly different ground cutter, is it new, has it been verified to be an exact dimension tool?
These are very tight tolerances for MDF, what measuring tool is used, vernier, mic, etc?
-
Just machine current material using different feeds, speeds, rough and finish and if not cutting correctly than adjust cutter diameter in CAD or MACH.
Some of the exotic woods machine like metal, some metals machine poorly, but if not what you want adjust accordingly.
I can get different results depending on how and what tool is used.
In fact, I can really screw something up if a re-ground is used and not accounted for.
RICH
BTW Dave,
I would be happy to hit that on aluminium
That's why files and sandpaper were invented and still used today! ;D :D
-
Of course :)
-
problem resolved
Thanks
-
well don't keep us in suspense......
-
Recalculated motor tuning manually cut better but still off discovered backlash was turned off.