X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient: geda-user AT delorie DOT com
Date: Sun, 2 Aug 2015 20:37:44 +0200 (CEST)
From: Roland Lutz <rlutz AT hedmen DOT org>
To: geda-user AT delorie DOT com
Subject: Re: [geda-user] Net attributes
In-Reply-To: <55A7D999.5020101@ecosensory.com>
Message-ID: <alpine.DEB.2.11.1508021958580.11805@nimbus>
References: <CAM2RGhQ70Pex5aNeQ86vKHc7sKf_Vpws69__CPb2QKg6fJTeHg AT mail DOT gmail DOT com> <55A2E9B7 DOT 9040502 AT neurotica DOT com> <CAM2RGhSmFhz9=oRaaj2EDL79c5XVmeuwmX_RdLqMv0evEHCyTw AT mail DOT gmail DOT com> <CACwWb3DbmOZbtb9Hrp2JvvG7-toryXuequQA=YV7X6ss762zsw AT mail DOT gmail DOT com>
 <20150713131707 DOT GA782 AT recycle DOT lbl DOT gov> <CAM2RGhTT__nWCkA4y36rb24yT8xP=o1nR04H=DNqTJiu=naSGA AT mail DOT gmail DOT com> <55A4042E DOT 5060402 AT neurotica DOT com> <CAM2RGhS=5xq0_oN4e0M55Kor4bcnXNn3NfLvRZoi7Vw9Aq1ZXg AT mail DOT gmail DOT com> <55A41B30 DOT 50602 AT neurotica DOT com>
 <alpine DOT DEB DOT 2 DOT 11 DOT 1507151114560 DOT 2113 AT nimbus> <254F9AFE-1A3E-4D88-BABF-E6E0F87A56B1 AT icloud DOT com> <1436960577 DOT 1072 DOT 6 DOT camel AT ssalewski DOT de> <EE9F72D6-E17A-4F5C-8E9E-45F12C901C9F AT noqsi DOT com> <201507151820 DOT t6FIKYME001704 AT envy DOT delorie DOT com> <201507152007 DOT t6FK7lv8005229 AT envy DOT delorie DOT com>
 <24AD56C6-B7C2-4D7E-B69A-F68DBACCBFDC AT noqsi DOT com> <4DD9C29A-2105-45DC-9F17-82DA23433919 AT icloud DOT com> <alpine DOT DEB DOT 2 DOT 11 DOT 1507160907360 DOT 3383 AT nimbus> <55A7D999 DOT 5020101 AT ecosensory DOT com>
User-Agent: Alpine 2.11 (DEB 23 2013-08-11)
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; format=flowed; charset=US-ASCII
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

Ok, so trying to abstract from these cases, there seem to be several types 
of net attributes:

1. Net attributes applying to an entire net (e.g., "GND" as a whole). 
These can be represented by attaching an attribute to any net segment.

   - any examples?

2. Net attributes applying to a bottleneck (e.g., the part of "GND" in the 
area of stepper driver 1, as opposed to the rest of "GND").  These could 
in theory be represented by attaching an attribute to a net segment, too, 
but then the affinity of visually unconnected net parts with the same 
netname would be unclear.  This is the case for which I suggested the "net 
attribute directive" solution[0].

   - trying to keep traces separate (power star pattern)

3. Net attributes applying to two entire nets.

   - keeping two nets away by X distance

3. Net attributes applying to the part of a net connecting two sets of 
pins (e.g., "between R5-2 and T1-2/4")

   - keeping trace resistance below X ohms (by making them either shorter 
or fatter)
   - placing power decoupling caps near a pin
   - min radius spec on a trace edge
   - smoothness of board material, since the waveguide smoothness affects 
how much mismatch there is and so reflections and in a short transmission 
line in a system, standing wave patterns on the transmission line 
(differential pair)
   - definition and smoothness of the trace width so it does not have blips 
or sharp edges
   - thermal function: carry away the heat vs. conduct as little heat as 
possible (maybe this should be a pin attribute?)

4. Net attributes applying to a pair of such parts of nets.

   - minimizing signal run time differences

I doubt there is a consistent way to represent all of these in a 
schematic.  How do other EDA tools solve this problem?  It wouldn't be 
hard to allow the netlist backends access to attributes attached to a net, 
but how would you specify which part of the "GND" net needs to be short, 
or where the parts of the net should meet?

Roland


[0] http://www.delorie.com/archives/browse.cgi?p=geda-user/2015/07/16/04:06:38