Re: gEDA-user: orthogonal lines

["Harold D. Skank" <skank54@xxxxxxxxx>]

Harold Skank here.

Samual is correct in the "original" sense.  The old light-board and
crepe tape on mylar sheet techniques used this technique precisely for
that reason.  A clever layout technician could create a two-sided board
of great complexity, primarily routing traces vertically on one side,
and horizontally on the second side.  Complexities of circuit topology
might still force you to use jumpers to solve some problems.

With the advent of multi-layer routers, the problem changed somewhat.
With the introduction of static power and ground planes, a designer has
the option of introducing these to "shield" trace layers from one
another.  One of the difficulties for me 'vis a vis' PCB is the 8-layer
limitation.  Using a single power layer and a single ground layer leaves
6 circuit layers, and 2 circuit layers have to reside adjacent to one
another.

Now, the nature of noise coupling:  it can be either capacitive coupling
or inductive coupling.  In either case using orthoginal traces reduces
the problem, as the coupling between any 2 traces is a "area" effect.

I once saw a backplane design using differential traces routed on
adjacent layers with no shield layer between.  In this case they were
taking advantage of the magnetic coupling between these traces both to
reduce cross-talk to adjacent parallel traces, and to control the
impedance of the resulting route.  I never found out whose router could
produce such a design.

	Harold Skank

On Fri, 2005-07-29 at 22:00 -0700, Samuel A. Falvo II wrote:
> On 7/29/05, Olgierd Eysymontt <olgierd.eysymontt@xxxxxxxxxxx> wrote:
> > > Actually, that's a side-effect of the way the previous generation of
> > > auto-routers worked.  PCB uses a newer technology, so does not need to
> > > have dedicated direction layers.
> 
> I always thought it was because it was easier to route the board that
> way, whether you're auto-routing or hand-routing.  Really, I thought
> this practice came about due precisely to the limitation of having
> only two sides to work with.  :)
> 
> --
> Samuel A. Falvo II