TECHNET Archives

1996

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Jeff Seeger" <simon.ipc.org!bort.mv.net!rapidcad!jseeger>
Date:
Wed, 26 Jun 96 16:14:30 EDT
Content-Type:
text/plain
Parts/Attachments:
text/plain (40 lines)

	Gary Willard wrote:

>     My main question is, are there any major concerns or advantages in 
>     having the vias connecting directly into the plane without relief, I 
>     vaguely recollect concern expressed over reliability of these 
>     connections if extreme thermal cycling is undertaken.


	There are definite electrical advantages to direct plane contact,
	we've done this for years.  The inductance of the supply/return
	loop drops enough to notice (important for high-speed switching
	along with proper placement of vias and decoupling), as does re-
	sistance (important for high loading).

	Plane performance on dense boards improves as the planes are 
	noticably less "broken up".  Also plane splits are a little easier
	to accomplish without leaving bad etch geometries behind.

	I agree that if you're doing your power/gnd routing well, there
	is a price to pay in rework.  I don't think the thermal relief
	image makes any difference to via reliabilty over thermal cycling,
	since the usual thermal geometry looks like a normal pad at the 
	junction to the PTH barrel.

	Our SOP is to call them "direct hits" with a small flash (so every
	one knows that we know).  For some fabricators, who require therm-
	al relief, we point to the appropriate thermal geometry for sub-
	stitution.

	I wish I knew why they do that.  I was told once it was due to
	"pink ring" but gather that's old news.  Fabricators?

	Regards,
        Jeff Seeger                             Applied CAD Knowledge Inc
        Chief Technical Officer                      Tyngsboro, MA  01879
        [log in to unmask]                               508 649 9800



ATOM RSS1 RSS2