\0
TO: I4235700 IBMMAIL new address for ipc technet 25.6.96
FROM: DSTEWART EX2 D.Stewart - Product Development Manager.
DATE: 22 October 1996
SUBJECT: re-microblind vias laser v plasma
REFERENCE: design
********************************************************************
From: [log in to unmask]
Date: Tue, 22 Oct 1996 08:10:22 EDT
To: [log in to unmask]
Subject: UNDELIVERABLE MAIL
=
FROM: DSTEWART EX2 D.Stewart - Product Development =
Manager.
=
DATE: 22 October 1996
SUBJECT: RE: PLUGGED VIAS (AND LASER VIAS AND PLASMA VIAS)
=
Just a couple of corrections re-plasma drilling of microblind vias-
Most circuit shops do NOT have a plasma machine, only those who use
teflon boards or other exotics which do not like/react with
permanganate for desmear. Even if they do have a plasma machine, it
is unlikely that the distribution of the etching process is
sufficiently tight to guarantee all vias etching completely, and
anyway, they are likely to infringe Dyconex's patent or sail close
to the wind, unless they have forked out =A3600K for a license.
=
Microblind vias by plasma ARE cheap on the basis that they all etch
simultaneously, but we decided to follow the laser route for the
following reasons:
=
1. Plasma etch is isotropic ie. it etches sideways as well as down,
so what starts life as a 0.1mm (4 mil) opening in the copper acting
as the mask, ends up as a 0.3mm (12 mil) opening by the time the
plasma has etched down through 3 mil of dielectric to expose 2 mil
of the copper on the layer below, and then the copper has been
etched back to remove the overhang. (Based on experience not
theory) At 12 mil we could mechanically drill this!
Using laser the cost goes DOWN as the hole size gets smaller, and
the positional accuracy is superlative - +/-0.5 mil, so the
manufacturing tolerance for the laser drilled blind via is reduced
to around 8mil total - so now you have a 0.1mm via in a 0.3mm pad,
compared to the plasma 0.3mm via in a 0.5mm pad. Loads more routing
space.
=
2. The dielectric MUST be a homogeneous organic substrate for
plasma etch, whereas laser can ablate through copper, and glass
fibre reinforcement, so materials do not HAVE to change, although
there are advantages to using the same materials as plasma.
=
3. Board manufacturing materials are priced in relation to their
global volume use, hence FR4 is still the cheapest, aramid (once
its impregnated by a laminator) can be anything from 3 to 10 times
more expensive, PTFE is extortionant, and polyimide foil is
probably 3 to 6 times more expensive. They are also more difficult/
impossible to bond by most fabricators, so the option of FR4 looks
more likely to be accepted.
=
=
4. Finally, there is the Dyconex license fee, which is a
particularly galling stone to swallow, as Exacta have examples of
their own plasma etched via products made back in the 70's, but
that's not sour grapes, just a comment.
=
=
=
=
=
=
=
=
=
=
=
=
=
As far as volume manufacture goes, all these new technologies are
in the early phases of volume ramp, Plasma etch machines can take 6
panels at a time if you buy the biggest machines, laser drills only
take 1 panel at a time, but obviously this will change as time
moves on.
=
For a final note, we are also working on the SLC type processes
(now known as Sequential Build Up -SBU or HDI- High Density
Interconnect) with some success. These are limited currently in
terms of the minimum via size for exposure, and dielectric
seperation, but the investment needed for most circuit shops is
minimal in comparison to laser or plasma, and the volume production
is not limited in the same way as the alternatives. There are
always advantages in different ways to make boards, so
fundamentally, the design rules should allow fabricators their own
choice as to which process they use to produce the board.
=
D.Stewart
Product Development Manager
Exacta Circuits
Scotland
=
The views expressed above are mine and do not neccessarily reflect
those of my employer.
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
---- End of mail text
Additional SMTP headers from original mail item follow:
Received: from simon.ipc.org by ibmmail.COM (IBM VM SMTP V2R3) with TCP;
Tue, 22 Oct 96 09:10:46 EDT
Received: from ipc.org by simon.ipc.org via SMTP (940816.SGI.8.6.9/940406.SGI)
id HAA20833; Tue, 22 Oct 1996 07:58:33 -0700
Resent-Date: Tue, 22 Oct 1996 07:58:33 -0700
Received: by ipc.org (Smail3.1.28.1 #2)
id m0vFg2o-0000RLC; Tue, 22 Oct 96 07:34 CDT
Resent-Sender: [log in to unmask]
Old-Return-Path: <[log in to unmask]>
Message-Id: <[log in to unmask]>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Resent-Message-ID: <"3ipQU.0.2Q7.hzBRo"@ipc>
Resent-From: [log in to unmask]
X-Mailing-List: <[log in to unmask]> archive/latest/7058
X-Loop: [log in to unmask]
Precedence: list
Resent-Sender: [log in to unmask]
***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to: *
* [log in to unmask] with <subject: unsubscribe> and no text. *
***************************************************************************
|