Re: gEDA-user: latest PCB w/gsch2pcb deletes parts?

[Dan McMahill <dan@xxxxxxxxxxxx>]

And just to add to this, there is no single "IPC one".  In fact the 
~geda library contains no less than 9 (!) IPC 0603 footprints.  These are

CAPC1608L
CAPC1608N
CAPC1608M

INDC1608L
INDC1608N
INDC1608M

RESC1608L
RESC1608N
RESC1608M

which are for capacitors, inductors, and resistors respectively.  the 
L/N/M are for least/nominal/most.  In addition, there is still just 0603 
in ~geda.  0603 needed repair because the one which was there was not 
large enough.  That one has been updated to something representitive of 
the "N" version.

As for other versions of 0603 footprints, they're there because no one 
has made footprint library maintainence a high priority.

-Dan

Steve Meier wrote:
> The IPC footprints have been critisized for being overly large. I
> believe this is true especially for the smaller size components. My
> understanding is that the IPC wants to leave some minimum pad space in
> front and behind (toe and heal) the devices contact. If you then compare
> the IPC recomended foot prints to a device manufacturors recomended foot
> print you will see a consierable size difference. I suspect this might
> very from manufacturor to manufacturor. This is one cause of multiple
> footprints. A second cause is that different assembly techniques may
> suggest variations in the width of the pads.
>
> Is this all really necessary? Propably not. But which one to select?
>
> Steve Meier
>
>
> Matt Ettus wrote:
>
> > I was able to fix this problem by going back to an older PCB CVS.
> >
> > One question -- why are there multiple 0603 footprints?  Why wouldn't
> > you want to use the IPC one?  All these versions makes for a lot of
> > confusion.
> >
> > Matt
> >
> > On 12/6/05, Stuart Brorson <sdb@xxxxxxxxxx> wrote:
> >
> >
> > > > > I recently upgraded to the latest PCB CVS.  Now when I run gsch2pcb on
> > > > > an already existing pcb file many of my components get deleted.  Also
> > > > > new 0603 caps no longer have silkscreen around them.  Is it possible
> > > > > that changes to PCB caused this?  I didn't change gsch2pcb.
> > > > >       
> > > >
> > > > Yes, latest pcb from cvs does not have silk around the 0603 footprint.
> > > > IPC-7351 seems to indicate no room for silk on 0603.  The newer 0603 (or
> > > > preferably the IPC-7351 compliant names for 0603 -- CAPC1608N for
> > > > example) in the ~geda  library should be much better from a soldering
> > > > point of view too.  The previous one was no good.
> > > >     
> > >
> > > A tangential point to this:  If you use newlib footprints, John
> > > Luciani's caps have partial silk at the ends of the parts which help
> > > determine the component body size during placement.  And (shameless
> > > plug) some time ago I wrote a perl utility called smtgen which
> > > generates footprints for two terminal passives if you give it the
> > > physical parameters like length, width, pad dimensions, etc.  You give
> > > it all parameters on the command line, and it writes footprint to
> > > STDOUT.  It draws a full rectangle around the part on the silkscreen
> > > layer.  I put it on my website for interested parties to use:
> > >
> > > http://www.brorson.com/gEDA/
> > >
> > > The resulting footprints may or may not be IPC standard (I haven't
> > > paid attention to recommendations for dealing with the silkscreen),
> > > but I have used them and they work.
> > >
> > > Have fun,
> > >
> > > Stuart
> > >