[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]

gEDA-user: pcb SMT footprints



Friends -

Can I call your attention to the disturbingly inconsistent state of
affairs regarding simple passive footprints in pcb?  And BTW, my
conversations with other ECAD users suggests the problem is not
unique to this software.  All dimensions below are in mils (sorry).

AVX 0603 part dimensions:
   length    63 +/- 6
   width     32 +/- 6
   terminal  14 +/- 6

Panasonic 0603 part dimensions:
   length    63 +/- 4
   width     31 +/- 4
   terminal  12 +/- 8

not_vetted_ingo  smt0603.ele
   length       91
   width        30
   terminal     31
   gap          29

~geda  geda_0603  Standard SMT resistor, capacitor, etc
   length       72
   width        42
   terminal     24
   gap          24

~generic  smd_capacitor_603  chip_capacitor
   length       90
   width        40
   terminal     30
   gap          30

~generic  generic_smd_chip_603  smd chip 603
   length       60
   width        30
   terminal     14
   gap          32

~genericsmt  SMT_603_CAP_RES  SMT 603 type capicitor/resistor
   length       70
   width        39
   terminal     17
   gap          36

footgen
   length       87
   width        32
   terminal     28
   gap          31

Average (after removing extrema)
   length       80
   width        35
   terminal     25
   gap          30

The first five footprints (in order, from not_vetted_ingo to ~genericsmt,
top to bottom) are displayed in the attached PNG (if your mail system
mangled it, you can see this file at
  http://recycle.lbl.gov/~ldoolitt/caplist.png
).  These footprints are all included in pcb snapshot 20040903, and I
don't think anything relevant changed for 20050609.  I ignore, for now,
the equally dramatic differences in center position and silkscreen.
Of course, other sizes (0402, 0805) are similarly inconsistent.

While I know some simple stuff, like those 60 and 70 mil long footprints
would make it a bear to hand-solder a part that's 60 mils long.  I suspect
a lot of people here have more first-hand experience as to what makes a
good SMT footprint.  Can you share it with me?

More generally, does anybody have a plan for systematizing and vetting the
pcb footprint library?  The "newlib" part of PCB seems to go one direction,
and my qfp-ui and Darrell Harmon's footgen go quite another.  Can we keep
and expand both?  My goal would have the following:
  - on-line database of footprints, with tracking of modification history
  - clear separation of, but cross-index between, footprints and chip
    part numbers
  - on-screen "wizard" interface to edit and create footprints
       * better interface than qfp-ui
       * more geometries than footgen
       * simple user option to submit result to on-line database
  - 90% of the legacy crud removed from default pcb installs

Thanks for your attention!

    - Larry

Attachment: caplist.png
Description: PNG image

Attachment: signature.asc
Description: Digital signature