Hi Charles,

Your response described exactly what I was thinking!

We are accustomed to the planes surrounding the
critical traces, if for no other reason than to create
a metal enclosure for them.

To move them outside would create better plane to
plane capacitance, the traces would have a more
unified chunk of "return" metal to glom onto, but
now we start pouring ground fills on the surface
layers to simulate a slightly larger "metal box" for
everything. I can''t calculate this quite the same way
because the outside planes are all "swiss-cheesy"
and chopped up by pads/vias/fanouts, so I'm not
even sure if the traces are now stripline or microstrip
(???)
ok, I'm exaggerating, and my question is mostly
theoretical since we are running and slower speeds
than some of you walkin the boundary tightrope, but
I like to keep tabs on what yer doin out there...

Jack
.


On 1/4/07, Charles Gervasi <[log in to unmask]> wrote:
>
> Jack,
>
> I've always gotten away with just assuming there are enough decoupling
> caps throughout the board to consider power planes as identical to DC
> ground.
>
> Your message got me thinking decoupling caps may not be the reason this
> works.  In a standard six-layer "sandwiched routing layer" stackup, it's
> easy for a trace on a vertical layer to switch to a horizontal layer
> through a via not near a decoupling cap.
> TOP
> GND
> Vertical
> Horizontal
> PWR
> BOTTOM
> Perhaps it helps in this scenario that a) some of the return currents
> can travel on the plane not immediately next to the trace and b) there
> is some capacitive coupling between planes.
>
> You could solve this potential problem by putting the two planes in the
> center, especially if GND and PWR are close enough (3 or 4 mils) to
> provide good capacitive coupling.
> TOP
> Vertical
> GND
> PWR
> Horizontal
> BOTTOM.
> But now there's no stripline (internal) traces AND you have to think
> about Top-Vertical and Bottom-Horizontal crosstalk.
>
> I don't know a scientific way to determine which is better.  It seems
> like we always end up weighing the various tradeoffs, and the values of
> the various tradeoffs are reckoned by more handwaving than science.
>
> Handwaving Through the Fog,
>
> CJ
>

---------------------------------------------------------------------------------
DesignerCouncil Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF DesignerCouncil.
To temporarily stop/(restart) delivery of DesignerCouncil send: SET DesignerCouncil NOMAIL/(MAIL)
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-615-7100 ext.2815
---------------------------------------------------------------------------------