Hello Guest it is November 27, 2020, 03:14:42 PM

Author Topic: What Steps/ Units Settings for Z-axis (Quill) of CNC Retrofitted Radial Machine?  (Read 21903 times)

0 Members and 1 Guest are viewing this topic.

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
hi ,

Backlashs for Axes:
1- X-axis: 0.005708 inch  =  0.145mm
Y-axis: 0.013385 inch   =  0.334mm
Z-axis: 0.0535433        =  1.36mm
Backlash speed of max%: 50%
shuttle Acceleration:0.08
Mill : climb mill

Calibrated Step/units(mm) :  calibration was taken at 290mm length with needle gauge and gauge blocks
   X-axes:  6093.32125758 
   Y-axis :  6093.32125758 
  Z-axes:   15177.80388
Kernel Speed: 25000Hz


I am running the steppers in 1/2 step mode.. Do we have any problem with the steps i mentioned above?


Today i checked all the things and they found tight. With the same settings as yesterday, i just plot 100mm Diameter circle and 150mm Square. The backlash for all the Axes was enabled. The circle meets all the points and were found round within 0.01mm. The Square was perfect within the 0.01mm. and the start and end point meets exactly...

This tells me that i have no motors as well as backlash calibrations issues in no-load conditions.

PS:
The circle was drawn first with Lead pencil and i measured it. but when drawing the square the pencil stopped marking. So i install the Pen, this work perfect for Square drawing but torn the page while drawing the circle.

Any thought?

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
Forget to write that the SQUARE is actually rotated square (45 degree) so that both X/Y motors collectively make it. So making the square both X/Y motors engaged all the times.

Offline RICH

*
  • *
  •  7,415 7,415
    • View Profile
KHALID,

Plotting a square and circle are good ways of checking. In fact there is an old standard for actaully machining a similar piece
with a raised hub for the circle and tapered sides out to the square for confirming machine accuracy.  Attached is a picture of my adjustable
spring loaded tracing pen which i use since it comes in handy sometimes. ( no tearing of the paper)

Just a comment on the Z backlash ( .053" ). You may find that Z will push up or pull down when machining
depending on how you have it counterweighted. There is no "locking" of the rack other than weight which may
not be a problem in softer / easier machining materials. But on harder materials it could say drop into the material
some and create a problem.

RICH

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
Hi RICH.. First i am thankful to you for showing us a nice tool you made..:) As i have shown you the 45 degree square and the Circle..I have perfect all the things..but today when i tested on Aluminum i found some trouble as you can see in picture..
It is the picture of  a simple 5-vane open impeller just to check the machine behaviour on the Aluminum... As you have seen in the picture, the simulation perfectly shows the same resultss of my Gcode.. I am annoyed only because i am going out of my centre.. As you have seen the picture one of the Vane just at the edge of the stock... All the dimensions are perfect..and this is very strange for me..if the motor looses steps then the Vane thickness and curvature must be different but in this case it is the same thickness.. I stopped the machine when the cutter strike the clamp.. If the part was completed then may be i had some clue why this happen?

The part dmension is 110mm, the stock used was Aluminum 121mm dia.. The pocket was done in 4-passess upto 8mm depth with  7 IPM and 800 RPM using new 8mm 2-flute HSS endMill..
Any thought?

PS:
SHOULD I SAY, THAT THE TOOL IS  GOING OUT OF CENTRE? ... A little confusing to me... Seems that  stock is offset to the Geometry in the CAM software.. But when i saw in the CAM the geometry is centred in the STOCK...... i am confused because thee impeller  machining and dimension is right after removal of the part.. It seems that i have some offset of the complete part in the stock.. Help me i am in a very perplexing situation... Anything in MACH3 that can cause this???

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
I am posting here the screen shot of the cncsimulator, the simulation and the actual part machined matches but it is out-of-way or it seems like that.. any suggestion.. I am attaching the gcode file too...

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
Here is the Simulation results and Actual - side by side... The simulator is new to me so i can't color map the depth...

Offline RICH

*
  • *
  •  7,415 7,415
    • View Profile
khalid,
Post the Gcode file.

RICH

Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
Some More informations...
The machining/pocketing of part was carried out with TOOL rotating CCW and Table rotating CW as shown in Picture..
Flood coolant was ON with 3-High Pressure high flow jet Nozzles... for the above conditions :

I plunge the Tool using Ramping technique for the Aluminum Impeller..
Ramp ANgle= 10mm
Ramp Height= 12mm
Ramp distance= 10mm
The same ramp with 5IPM and the cutting speed was 8 IPM, using HSco 2Flute 8mm end Mill...


Again , i don't know why my Z-axis goes down in the work piece... In CAM i gave 2mm DOC/Pass for the above Aluminum part and the first Pass i measured was about 3.5mm.. there is some clearances in the Quill i think that once adjusted for the first pass and then for the next consequent passes it gives me exact DOC...


The backlash compensation was done in Mach3 and in the Circle and rotated square i am getting 0.01mm tolerances and no sign of backlash.. I afraid that this is done in no-l;oad condition... But if Mach3 compensate the backlash then the problem should not appear in actual work piece...
One other thing i want to share here, is that i installed the dial guage on the tooltip.. and gave lateral movement to the quill by hand i was getting 0.5mm flex... so total flex will be 1mm that will appear in workpiece..
This is the reason I am currently making drawings of a spindle... and will change the complete Z-axes...
Do i have to rework on my X/Y cross slides?.. These slides are boxway type and very rigid.. I feel no vibration in them.. Just to check the Tramming tomorrow...


Offline khalid

*
  •  213 213
    • View Profile
    • My Wood carving
Following is the Header of my impeller Gcode:
G90
G70
G40
g20
g64


I enabled the G64 (CV Mode is ON)...
Should this CV mode be on when we have backlash compensation ON?...

Regards

PS:
I used Quantum a lot in past without backlash compensation On and it gave me satisfactory results.. I am using QUANTUM MACH3 Version R2.43..
I have following settings to AntiJersk compensation as can be seen in Picture attached