Re: gEDA-user: Part Rotation in XY-data

[Bert Timmerman <bert.timmerman@xxxxxxxxx>]

Hi John, Duncan and all,

On Sun, 2008-11-09 at 08:23 -0500, John Luciani wrote:
> On Sun, Nov 9, 2008 at 6:02 AM, Duncan Drennan <duncan.drennan@xxxxxxxxx> wrote:
> 
> > I expected the parts to be rotated relative to their library
> > references, but the SOIC part has an extra 90º rotation. Rotation in
> > pcb increase in a clockwise direction. I've attached a test .pcb file
> > and the resulting xy coordinates.
> 
> The calculation does not compare the placed part to the library.
> With the centroid at (0,0) zero degrees of rotation is when pin one
> is in the upper left quadrant.
> 
> The angles should increase with counter-clockwise rotation. It sounds
> like there is a bug.
> 
> > I am guessing that rotation (like centroid data) is calculated by some
> > algorithm, what is the basic thought behind the calculation?
> 
> This is from the PCB manual ---
> 
> Computation of Centroid and Rotation
> 
> The center of each element is found by averaging the (X,Y) coordinates
> for the center of each pin and pad in the element. For example if an
> element has 2 pins, 1 at (1,0) and another at (1,4) then the centroid
> will be at (1,2).
> 
> The calculation of rotation is a bit more complex. Currently a
> rotation is not stored for each element but rather the rotated element
> is stored. In other words if the element from the library has a pin at
> (0,0) and (0,2) and it has been rotated by 90 degrees, then the .pcb
> file will store (0,0) and (2,0) for the pin locations with no
> indication that they have been rotated from the original.
> 
> In the event that the element has only 1 pin, then the rotation is set
> to zero.  If the element has only one pad (as opposed to a
> through-hole pin), then the rotation of the pad is used.
> 
> When the element has multiple pins, the location of pin #1 is placed
> in the coordinate system which has the centroid of the part at
> (0,0). Then which quadrant pin #1 falls in determines the
> rotation. Zero degrees of rotation is defined as pin #1 being in the
> upper left quadrant. Increasing angles correspond to counterclockwise
> rotation so a rotation of 90 degrees places pin #1 in the lower left
> quadrant. Currently, the only allowed rotations are 0, 90, 180, and
> 270 degrees.
> 
> If pin #1 happens to be at the centroid of the part, then pin #2 is
> examined to see which quadrant it is located in. The same rules apply
> for the definitions of rotation. In other words, when pin #1 is at the
> centroid of the part and pin #2 is in the upper left quadrant, the
> rotation is declared to be zero degrees.
> 
> 
> (* jcl *)
> 

Maybe this issue was on the list before, I can't remember :)

Anyway, if we start to rotate elements with arbitrary angles 
(angle is !0, !90, !270, !360), output from pcb's XY-exporter is likely
to give some serious problems with pick-and-place machines.

Example:

Pin #1 is at the Mark = 0,0
Pin #2 is somewhere in the upper left quadrant
Rotate the element 0.5 degrees around pin #1

0.0 degrees != 0.5 degrees --> Jackpot !!

Just my EUR 0.02

Kind regards,

Bert Timmerman.



_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user