kicad-lib-committers team mailing list archive

[Vesa Solonen <vesa.solonen@xxxxxxxx>]

11/09/16, 05:01, Carl Poirier kirjoitti:

> Do any of you have comments about the changes?

Here are some comments an assembly house gave to me:

Silkscreen does not matter on board anyway. It's mostly a hindrance with
paste placement (creates a gap between paste stencil and board) and not
readable on a dense smd board. The only use is QC and hand placed
components, but that is not on board, but on a piece of paper. In that
case the best position for references is directly on the center of the
component.

Polarities and rotations. Kicad has a lot of components that are in
funny orientations. E.g. for polarized capacitors the polarity should be
in X-axis direction. These are something to check.

Finally the paste openings are hackish in the case of QFN. We should use
_separate_ paste pads that are not numbered and are just on paste layer.
The copper pad should be separate as should possible via pads for
cooling. The paste shrinking should also be lockable in the component,
but there is no zero and have to use a very small term to simulate that :)

-Vesa