[SI-LIST] Re: Cu Thickness
- From: "Salkow, Steven" <steven.salkow@xxxxxxxx>
- To: olaney@xxxxxxxx, si-list@xxxxxxxxxxxxx
- Date: Thu, 19 Jul 2007 11:31:03 -0700
Mr Lanely is right on in his advice, especially about visiting the
fabricator! No two fabricators are exactly the same hence what may be
costly one place may be no charge some place else.
The process of adding copper is commonly referred to as Thieving
For more information see this article on my web site:
http://www.bychoice.com/Thieving_and_Copper_Balancing_in_Printed_Circuit
_Board_Design.pdf
At Xerox, PCB were made with up to 15 mils of inner copper on multiple
heavy inner power planes. Lockheed has made several mixed signal boards
that included multiple inner power planes of 2 ounce copper.
Steven Salkow
Lockheed IS&GS
3130 Zanker Rd, San Jose
Ca. 95134
W:(408) 473-4058
steven.salkow@xxxxxxxx
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of olaney@xxxxxxxx
Sent: Thursday, July 19, 2007 10:29 AM
To: si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Cu Thickness
Just so we all understand each other, I'll review the assumptions that
should apply here.
*Boards can be made with mixed copper thicknesses. This is no big deal.
*Only the power and ground planes "need" 2 oz copper, though this might
be overkill.
*Even if 2 oz cu is used for power distribution, the signal layers
should
still be sized for fine pitch, 1 oz or less (unless we are talking about
some heavy duty power signals, in which case the fine pitch concept does
not apply anyway).
*If you do design with mixed layer thicknesses, warpage is a problem if
the stackup is asymmetric.
*If the thicker layers are the two middle layers or the two outer
layers,
as an example, then the board is symmetric and the board is not more
prone to warp than a board without mixed thicknesses. The tendency to
warp is not a cliff edge effect, which is to say that a small amount of
deviation from perfectly symmetric means only a small amount of warp
tendency, a lot means a lot.
*Multilayer boards can have an even or odd number of layers, but the
actual board is built by laminating thin, double sided boards together
with prepreg in between. An odd layer count simply makes one of these
single sided. Go for an even layer number and the "extra" layer is
almost free.
*As a matter of wisdom, the stackup should be designed so that the
thicker layers are paired on the same double sided component. Layer
pairs with the same cu thickness on both sides is ordinary. Ask for a
different weight on each side and the fabricator sees dollar signs
floating in the air.
*Also be aware that fabricators might add copper to layers in open areas
to keep the board the same nominal height over the entire surface.
These
floating pieces of copper can be a constant impedance or stray coupling
issue, so be sure to specify where this should not be done.
One thing I recommend for any engineer involved in multilayer design is
a
tour of a PWB fabricator. This is usually just a matter of a phone call
and a few well spent hours. See the steps with your own eyes, ask about
all the points just made and any others you can think of, and design
your
boards to fit the process. Ask what space and trace their particular
process supports without a big kink in the cost curve. There are
process
variations that might apply. It's that or wonder why the price of
different designs keeps jumping around, and problems like yield,
warpage,
or worse keep happening. OK, experts, time to add your 2 cents.
Orin Laney
On Thu, 19 Jul 2007 06:33:35 -0400 (EDT) Stuart Brorson <sdb@xxxxxxxxxx>
writes:
> Andrew,
>
> Thanks for your answer. Your points make sense. The point is that
> 2
> oz Cu is inappropriate for fine-pitch boards, particularly for large
> boards where the final feature size might vary over the
> board's surface due to non-uniform etch rates. Right?
>
> Is there a rule of thumb, like trace/space = 10/10 => OK for 2oz Cu,
> but 5/5 => not OK? A similar rule for 1oz Cu?
>
> Cheers,
>
> Stuart
>
>
>
> On Wed, 18 Jul 2007, Andrew W. Riley III wrote:
>
> > Stuart,
> >
> > I do not see board warpage being a factor unless there are extreme
> > differences in copper layers of the PCB, or the fab house is not
> reputable.
> >
> > Coming from the layout side, I have only had complaints from fab
> houses WRT
> > tolerances and extended reliability at 2oz routing layers. At
> that time it
> > was the customer's request and I was able to talk them down with
> facts
> > supplied or easily researched and sometimes a conference call that
> included
> > an engineer specialized in signal integrity with reliable
> simulation tools.
> >
> > In an attempt to make this reply short, my experience has been
> that
> > increasing copper weight was not a favorable consideration in
> those designs.
> > Having the backup of an experienced and proven signal integrity
> expert on
> > call was a factor.
> >
> > Sometimes you gotta bite the bullet and let the powers that be
> know it's not
> > gonna be free.
> >
> > Cheers!
> > Drew
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
