X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f X-Recipient: geda-user AT delorie DOT com Date: Sat, 13 Jun 2015 05:52:06 +0200 (CEST) X-X-Sender: igor2 AT igor2priv To: geda-user AT delorie DOT com X-Debug: to=geda-user AT delorie DOT com from="gedau AT igor2 DOT repo DOT hu" From: gedau AT igor2 DOT repo DOT hu Subject: Re: [geda-user] Parts with internally connected pins In-Reply-To: Message-ID: References: User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Reply-To: geda-user AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: geda-user AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Fri, 12 Jun 2015, Kai-Martin Knaak wrote: > Donald Tillman wrote: > >> My current approach, giving the shield pins the same pin number in the >> footprint, isn't bad, the rats nest just complains about missing >> connections. And I can add the connections, it's not holding me up. > > In some cases, the redundant connection can be considered beneficial. E.g. > the multiple ground pins of a micro controller. Or the shield of a CAT6 > RJ45 connector. > > >> But I'm wondering if there's an approach that's more preferable, that's >> more in line with the internals. > > How about this: > 1) all internally connected pins receive the same pin number > 2) the designer can attach an ignore flag in the layout tool to the rat. > > This would make pcb ignore the respective rat when it assembles the > currently unmet rats nests. > This issue keeps popping up from time to time... My solution in my PCB fork is an "internal connections" list, which comes from pin attributes of the footprint. If a pin/pad is assigned to a group, pcb simply implies galvanic connection between the pin and any other pin in the same group. Details: http://repo.hu/projects/pcb-rnd/intconn.html Example: A 0 ohm SMD resistor has both pins in the same group so can be used as a bridge to remove a rat line. The resistor still needs to be added in gschem, but as a single-pin device connected anywhere to the network. This means if I move the resistor to bridge gaps at different section of the network (e.g. connect the pin of a different IC to the same network), I don't need to back-annotate the change to gschem. Together with the nonetlist feature, I can place such jumpers in pcb without having to add them in ghschem at all. Details: http://repo.hu/projects/pcb-rnd/nonetlist.html I've been using these for more than a year now and they work well. I kindly recommend implementing these features (and flagcomp, http://repo.hu/projects/pcb-rnd/flagcomp.html for supporting similar transitions) in mainline PCB. Regards, Igor2