On Thursday 08 August 2013 19:36:42 Jan de Kruyf did opine:

> tangent of 1 degree is 0.0174550649282.
> 
> So if there was a 1 degree error on your milling machine and a 1 degree
> error on the cutting of the pcb, the offset from flipping would be
> 0.0349101298564 inch per inch boardwidth.
> 
But, said another way, it says there should be an angle to the offset as 
the row marches down the board when comparing a row drilled in the top, 
with the same row drilled using the bottom offset after flipping the board 
on the axis of the row of holes.  Parallax effect IOW.

Unforch, for that theory its a null, the offsets marching along the x axis 
on top, match exactly as near as I can optically check them.

> Just as an example. I am not saying this is the case at all, but at the
> same time I would be very suspicious.
> 
> I have made a similar error before in my life. I was doing acceptance
> tests on a big machining centre and
> the machine gave a few hundreds of a mm error compared to the calibrated
> bar on the table used for checking.
> 
> That is, until it dawned on me that it was summer and the workshop
> temperature was in the 90's and the scale was glass (Heidenhain) and the
> rod was steel. And a few quick sums on a piece of scrap paper relieved
> my anguish.
> 
> Good luck

This error, all along the x axis seems incurable.  I have cut up two more 
bits of lexan, and the bottom of boards holes are about 1/3rd of a hole to 
the left, and I am so frustrated because adjusting the fudge values near 
the top of the file, and the results it then cuts, make very little sense 
to me.

So I guess I make a new pallet tomorrow that will hold the board with the 
long length on the Y axis so I don't have to screw with the (R)otation at 
all in the G10L2P# statements.  Then I can adjust the fudges to put the 
pattern in the center of the board and it will Just Work(TM).

Question for whomever wrote the G10 L2 code:

Does it apply the arguments given in the order given on the command line, 
followed by the amount of rotation, or does it somehow apply the rotation 
and then the origin offsets?

Also, do I have to switch maps and run the machine to 0,0 before I issue 
the rotation command?  If so, then that is not being done but can be if it 
would help.

To say the docs I have been able to excavate so far are less than 
transparent about how they actually work, might be a bit strong but somehow 
I'm missing the one clue that makes this all fall into place.

So the best I can probably do is make a new pallet that doesn't need to 
have its map rotated.  The last time I did this, I was making the encoder 
board for the lathe which didn't need to be rotated, and I was able to walk 
it into perfect registration in about 3 test runs.  I must have 50 runs in 
trying to rotate the pattern, and get X axis registration and ATM I am 
further off than when I did the first run using SWAGS for all 3 of those 
values.

So if a set of rules that actually can make it work can't be developed 
tonight, tomorrow I make a new pallet.  That I know I can make work. :)

> j.
> 
> On Thu, Aug 8, 2013 at 7:29 PM, Troy Jacobson <[email protected]> wrote:
> > Gene,
> > I've been starting to think about how I want to mill double sided
> > pcbs, and have been trying to figure out how to avoid problems like
> > this.
> > 
> > Is the size of your PCB blank causing an offset when it is rotated? 
> > It seems to me that the point around which the coordinates are
> > rotated need to be in the exact center of the PCB, or at centered
> > between the two relevant edges depending on the type of rotation/flip
> > being done.  If you were to drill a hole at that place, it should be
> > half the board size away from the edges of the pocket.
> > 
> > When calculating the coordinates for a rotate or flip, are you using
> > the same point for all files associated with the board?

Those points are all supplied by pcb2gcode, cloned just this past week from 
its repo.

I have scanned the drill files, and they are identical except the bottom 
file has a - sign in front of the Y values.  IOW, they are perfectly 
mirrored about the original Y axis 0.0 point.  I'd have to assume at this 
point that the etch files are generated by the same rule set, so the exact 
same top to bottom offsets should work that work for the drill files.  The 
drill file runs in 4.5 minutes, the etch file is something in the 75 minute 
per side run time.  But I am not going to waste time with a test etch if I 
can't get the drill files to work.

Thanks Jan, I appreciate the suggestions.

[...]

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page: <http://coyoteden.dyndns-free.com:85/gene> is up!
My views 
<http://www.armchairpatriot.com/What%20Has%20America%20Become.shtml>
T-shirt:
        Life is *not* a Cabaret, and stop calling me chum!
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
         law-abiding citizens.

------------------------------------------------------------------------------
Get 100% visibility into Java/.NET code with AppDynamics Lite!
It's a free troubleshooting tool designed for production.
Get down to code-level detail for bottlenecks, with <2% overhead. 
Download for free and get started troubleshooting in minutes. 
http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to