I agree! Balanced z-axis constructions are such an important factor to bow
and twist, that I've even suggested to designers to duplicate the innerlayer
(for a 3-layer mlb, as designed) to achieve a balanced 4-layer construction
(with 2 duplicate innerlayers, as fabricated). This has worked quite well.
Fred Paul
PCB Staff Engineer
FLUKE Corporation
425 356 5734
-----Original Message-----
From: Charles Barker [mailto:[log in to unmask]]
Sent: Saturday, August 28, 1999 1:53 PM
To: [log in to unmask]
Subject: Re: [TN] In Circuit Test
Hi Paul,
Regarding the even number of layers question, as I understand it, if you
want a
board to have minimal warp adn twist, it must have balanced construction in
the
Z-axis. For instance, a four layer board is typically signal, plane, plane
and
then signal. This gives relatively the same amount of stress on opposite
sides
of the board so as to to counter balance each other. Thus, the board will
lay
flat both before and after assembly processing.
If you build with an uneven number of layers, then you have to play games
with
the dielectric thicknesses to try to achieve the same effect. In an earlier
life, we were given a three layer board to build. They were always warped to
the
limit before we processed them and were worse after reflow. The engineer was
so
adamant about not messing with the layering that he wouldn't let us change
it.
We finally went from a .062" thick board with the middle layer evenly spaced
from top and bottom, to a .093" thick board with the inner layer at about
.063/.031. The inner layer was a plane layer btw.
I believe one might also achieve a better electrical balance by using a
physically balanced layering scheme.
Does this answer the question as you posed it?
Regards,
Charlie B.
Paul Klasek <[log in to unmask]> on 08/27/99 08:18:31 PM
Please respond to "TechNet E-Mail Forum." <[log in to unmask]>; Please respond
to
Paul Klasek <[log in to unmask]>
To: [log in to unmask]
cc: (bcc: Charles Barker/US/I-O INC)
Subject: Re: [TN] In Circuit Test
Tempted to agree (with your last line) ; one day an otherwise nice chap from
Honeylands will visit you with a shotgun for clogging his files !
Go to chat up chicks for a change and let earl have his yarn for gods sake .
My sanity on Saturdays is indeed questionable :
do any of you folks automate big lines (like telephones ; pc's ' etc. for a
private exchange please ?
paul
-----Original Message-----
From: [log in to unmask] [mailto:[log in to unmask]]
Sent: Saturday, 28 August 1999 10:33
To: [log in to unmask]; [log in to unmask]
Subject: Re: [TN] In Circuit Test
In a message dated 8/27/99 6:26:31 PM Central Daylight Time,
[log in to unmask] writes:
> interesting , how thick is 52 ? (just curious) .
> Earl we have the layers cost dilemma here :
> What would be ratios of yields and costs from simple double sided FR4 ;
> to 4 layers, 6, 8, >>> etc. ?
>
> And why even's only ; no odd layers ever ?
>
> thanks paul
Hey Paul!
What you doing sending a email on your Saturday? Workin' a week-end or
something? I'm just answering you just to test what little I've learned
about
PCB fabs ( I need to learn a lot more!) So all you teachers out there,
correct me if I'm wrong.
All layers start out as a double-sided copper-coated laminate, thus the even
number. The technology is advancing all the time (what's new?) but the
smallest core I've heard of is .014". So at that dimension, a 52-layer would
work out to .728" at the minimum...but there has to be some resin between
the
layers, so I would think that it would be some VERY THICK BOARDS. (I've
never
seen a 52-layer board! 18-layers is the most I've ever seen!)
The problem that I see with that many layers (and why Earl's boards are big
bucks!) is the registration between all these layers when making the fab! We
think we have problems putting some components on boards like this and
reflowing them, I'd hate to have the job of putting this FAB together!
I've gone to the dogpile and looked around, found some stuff about resin
coated copper, but it was listed as a "product in development" so maybe you
might see some odd numbered layers soon...boy, I'm I dumb about this stuff!
-Steve Gregory-
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following
text
in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following
text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
|
