Mail Archives: geda-user/2015/02/18/04:28:00
On Wed, 18 Feb 2015, Kai-Martin Knaak wrote:
>>> Drawbacks:
>>>
>>> - the same pin numbering and back-annotation problem that you
>>> have with connectors
>
> @Igor2: Would you mind to elaborate?
>
>
>>> - the star symbol/footprint will have explicit number of pins,
>>> you may end up unused pins or will have to replace with the next
>>> bigger if more nets are to be connected
>
> The pads of the star symbol can be 100% overlapping in the footprint.
> So a replacement won't break the layout.
If there's 100% overlap of all pads, and you connect a trace to it, you
connect it to all pads, not just the one you wanted to. This would require
another exception. The simplicity of my proposal 3 works only if you have
a section of the pad which does overlap and a section which doesn't and
the trace is connected to the non-overlapping section only. This means
find.c figures the trace is connected to that given pad, pad number N, and
this is what the netlist wants too, so all fine. The only exception to be
added is that it should stop checking whether this pad is galvanically
connected to other pads as well.
Another problem with 100% overlap is that once you connect two traces of
two different nets to it, the two traces may (under the pad, sort of
invisible to the user in some cases) also touch. This way find.c would
detect both traces are connected to the pad(s) but they are also shorted
together in the same point.
My proposal makes this more explicit because the non-overlapping parts of
the pads would be sticking out, like bristles of a hedgehog and there is
an 1:1, visible connection between a pad and nets. Just like with
connectors now.
Hence the pin-numbering/back-annotation problem I mentioned: when drawing
the schematics, it's hard to predict what pin number of the star should be
used for a net. When you do the PCB, you already see swapping two pads
makes routing easier, but then there's no easy mechanism (that I know
of) to get this swap back-annotated to the schematics.
Regards,
Igor2
- Raw text -