1

Hey,

Im working in a PCB line process, and as you can see in the image the first saw is cutting well but in the second saw which is smaller, it literally rotate the part 90º and i’ve been trying to figure it out in the phyton script and nothing came to my mind of what it could be?

Does anyone know what could be making the part to rotate 90º?

jdjjddja.PNG
jdjjddja.PNG1483×851 193 KB

Thanks in advance!

2

I would guess that the RetainOffset is unchecked for the circular saw’s conveyor path. If I read the picture correctly, the delta robot places the cut parts with side facing to the conveyor’s direction, meaning that the parts get a 90 degree offset relative to the conveyor path. When RetainOffset is on, the parts move between container transitions naturally, but when it is off, it suddenly rotates (and moves) the parts aligned with the conveyor path. That option can be changed in the Modeling tab for the circular saw’s path behaviour.

3

Hey,

It works perfectly, now i have another issue, it happens the same but instead of rotating in the input it rotates in the output. Is there any checkbox that im missing?

Here the image:

fdssfdsfdsa.PNG
fdssfdsfdsa.PNG1190×659 104 KB

Thanks in advance!

4

And also the cut that the saw makes is wrong it should make 100mm and another 100mm in half in the Y axis not in the X axis.

Thanks!

5

I see now that the circular saw component does quite many assumptions on the product’s position. I managed to modify it so that it makes the cut along a different axis:

  1. Use product’s width parameter as PartLengthProperty
  2. Switch PathAxis in the saw’s path into Y-Positive (This affects which side length of the part is used to calculate the sensor hits.)
  3. Modify the script on two lines: new part’s position matrix copies the orientation from the old part, and the old part is shifted along Y-axis instead of X-axis
    image
    image858×400 36.4 KB

    Circular saw different axis.vcmx(411.9 KB)

That component is quite old and nowadays the majority of eCat uses a different approach for modeling this kind of processes. I will see if a similar machine built fully using a process routine would work, and we could then replace this machine with a modernized version.

1 Like