[Robert M]

Happens few time where I need for a part repetitive (nesting few) calling dual side machining (both sides)

Wondering if anyone have a "hack" to/by nesting (that is with what software ?) the nesting can be flipped on a given axe ( or by index pins) to then, re Zero to machine that other side ??

Thks

[deadlykitten]

Hy pls show some image or video, so to see what you mean. Maybe i can help you / kindly

[Robert M]

Ya...thks.

Note... re-read myself = Not obviously (poorly) explained !!
Usually, went nesting a part ( to the best I can find out from what I have / know & searched), nested part can only be machined on one plane ( 1 face)
If that part ( nested to a few ) needs to be machined on the other side ( while still in "group" nest bulk), seam no way top program that other revers side while still grouped )


Photo 3 shows only a multi repeated same part ( as nested would be) on a single "1-stock set up" but imagine those 10 units part of a bigger "single rectangle raw materiel / plate " to be machined as individual (restrained/hold to the bigger plate say by tabs)....

Then, flipping side (plane), repositioning X=0.0 and Y = 0.0 on that flip side, to then machine the back / flipped side while still grouped (nested)

Hopping this is a "better" explanation on what I'm seeking ??

Thks

[deadlykitten]

hy robert it's related to software capability

for repetitive, i use a different aproach : program is for a single part, then wcs is shifted, to suit each individual location

in other words, a simple program, is begin called by another one, that changes wcs's as needed

let's supose that you did not had this problem ... in the end, regardless of program construction, machine would move the same, so, which aproach is better ? repetition inside the program, or simple program + wcs shift ? i recomand the last one for long operations, setups that requires custom intervention, have a higher frequency of restarts, or nesting patern is a bit random ( as place and/or orientation )

what you need is someone to point you to how to multi-face operation inside a nesting group for your software, and what i propose is a different end program construction

so, until you fix software usage, simply edit your program:
// wcs 1 / main
// cut
// shift wcs
// cut
// shift wcs
// cut
*and so on

[CNCMAN172]

The most common way to do this is to add four tooling pins outside of your design that are symmetrical side to sign and top to bottom. You machine one side of the nested design, then you flip the workpiece and put it back on locating it with those tooling pins and then machine the other side. Make a small test design to test if you are flipping things correctly so you build confidence then do the real job

[joeavaerage]

HI,
I make double sided PCB's by isolation routing daily.

I use two pins through the PCB 5mm outside the boundary so when you flip it over the top and bottom sides line up.


The last op is the milling around the periphery of the PCB with then leaves to alignment holes in scrap blank material.

You could use a similar idea.

Craig

[deadlykitten]

please joe, will you share an image or something ? so to better understand what you are doing ? thank you

---

regardless of cam, i can take the final program and split it , making it work with multiple wcs, like machining one part at a time, or machining all parts progresively with same tool ( thus atc sequences will be as when machining a single part ) / kindly

[joeavaerage]

Hi,
am at home now but have to do a board first thing tomorrow morning. will take a couple of pics. It's not rocket science.

When I design the board even if it's not rectangular, there are the critical outer dimensions which form a rectangle and bounds the PCB.
The top surface is the first to be milled, and the lower left-hand corner of the bounding box is WCS (0,0,0)

I drill two 1.5mm holes at (0,-5,0 ) and at (62,-5,0) where 62mm is the width of the PCB, into which I install two 1.5mm dia. pins.

Once the top surface is isolation routed then I flip the board installing it on the same pins. The (0,0,0) point of the board of the bottom surface
is now the lower right-hand corner. Once must jog or MDI to (62,0,0) and re-zero the X axis, ie the WCS (0,0,0) is the lower right-hand corner
of the bounding box. Now you can proceed to isolation route the lower surface.

Once the bottom surface is finished then you can mill the periphery of the PCB and the PCB blank which contains the two alignment pins is effectively
separated and discarded.

Were it your intent to replicate the toolpath that you employed on the top side that could be done by running the same code but 'reflecting it in the Y axis'
with a G51 Y -1 at the top of the code. In my case the design software generates two distinct toolpaths, the top layer, and the bottom layer traces.

It might sound like a complicated procedure, but once you've done it a few times it becomes routine. To make double sided PCB's with surface mount components
it's necessary to get te top and bottom layers registered within 0.05mm or better, I routinely get 0.02mm or better.

Craig

[deadlykitten]

oh, so is about flipping parts and miroring wcs ... so to say

i'l go get a coffee ... keep it up guys

[joeavaerage]

Hi,
finished the PCB.
First pic shows the top side with alignment pins installed and at WCS0,0,0 at the lower left-hand corner, then the completed top side. The remaining pics show the PCB flipped. Notice how the right-hand lower corner is
now WCS 0,0,0. Finally the complete board viewed from the bottom side drilled and m9illed ready to use. Note how the alignment pins are outside the PCB periphery and therefore discarded.

Craig

PS, it looks like the order of the pics is screwed up... maybe they are listed in alphabetical order rather than in the loading sequence?

[deadlykitten]

hy joe that looks so nice : i see a dancing crocodile, and some stickmans doing disco moves ; i made a picture, so to clear doubts

i also noticed reorder when uploading, after a while i simply ignore it, because whoever is interested can figure it out ... during time, i noticed that images > 1000x1000 px are rescaled, maybe that has something to do with order ? sometimes i rename the images as 01 02 03, or abc, thus their names are sorted ( thus filename to matter over upload order ), but i no longer follow if this affects order or not