[PCB_FORUM] Re: Removing unused pad rings on inner layers DURING design...

  • From: "Gerry Meier" <gerry.meier@xxxxxxxxxxxxxx>
  • To: <icu-pcb-forum@xxxxxxxxxxxxx>
  • Date: Thu, 31 Mar 2005 13:06:41 -0800

Thanks Jim - I only allow myself to be verbally thrashed once a day. 
Tomorrow is a new day

Gerry


 Gerry Meier 
Sr. PCB Designer 
Freedom CAD Services, Inc. 
Voice: (603) 864-1300 x1350 
Alt. Voice: (386) 753-0048 
Email: gerry.meier@xxxxxxxxxxxxxx 
visit our website at<http://www.freedomcad.com 


-----Original Message-----
From: J Wages [mailto:jwages@xxxxxxxxx] 
Sent: Thursday, March 31, 2005 3:56 PM
To: icu-pcb-forum@xxxxxxxxxxxxx
Subject: [PCB_FORUM] Re: Removing unused pad rings on inner layers
DURING design...

Do you really need a 24 mil via? I've routed several 1mm on a standard
63 mil thick board and use 19 mil via pads with a 6 mil finished drill. 

Jim S. Wages / SR. PCB Layout Designer:  
Cary, NC - (919) 466-1596

-----Original Message-----
From: Austin Franklin [mailto:austin@xxxxxxxxxxxx]
Sent: Thursday, March 31, 2005 3:39 PM
To: icu-pcb-forum@xxxxxxxxxxxxx
Subject: [PCB_FORUM] Re: Removing unused pad rings on inner layers
DURING design...

Hi Gerry,

>  Oleg,

I think you mean Austin, as that is my response TO Oleg...

> Not sure what your motivation here is.

My motivation is to find if there is a solution available in the tool,
as opposed to having to perform a "work-around" so I can route a board
using standard design constraints using a tool that should allow me to
do what I am trying to do.  Apparently, the answer is no, the only
solutions are work-arounds.

It's simple arithmetic.  1mm BGA, 24 mil via pad, 10 mil drill, 4/4.5/4
mil differential signals for 100 ohm impedance on inner layers for a 12
layer board .062" thick.  With the pad ring on inner layers, that leaves
39.37
-
12.5 - 24 = 2.87/2 for trace to pad clearance.  That does not work.
Also,
not routing differential pairs together you lose the differential
coupling for the distance it takes to get out of a BGA...which in some
instances can be significant.  With the unused pad rings removed during
the design process, this problem is  alleviated.

> By properly defining your vias,
> line widths and spacings you should be able to route two lines through

> even 1mm BGAs (unless you boards have high aspect ratios).

I'm not sure what you mean by "properly".  What I have defined for
design parameters is perfectly "proper".  So there is no confusion, I am
specifically speaking of routing between the dispersion VIAS (not the
BGA
pads) in a 1mm BGA field on the inner layers.

The trace width/spacing for differential pairs is determined by the
fabricator, with, of course, my input.  I can not go below 4 mil trace
width, and this gives me a 4.5 mil differential gap, so that is what is
"proper" for the technology board I need to use.  What is proper for the
via, again, for the technology I have to use, is 24 mil pad, 10 mil
drill.
So, again, simple arithmetic will not allow 4/4.5/4 AND proper pad to
trace clearance between a 1mm BGA dispersion field with a 24 mil via
ring.  A ring/drill/trace any smaller and that is outside the design
requirements.
Besides, as I've stated, it IS doable, and I have done it with no
problem, but it requires additional work on my part to create and use
the custom padstacks.

> Attempting to
> use the drill hole as you clearance can be dangerous for high speed 
> nets. You could actually get the line close enough to the hole that it

> is not referencing the plane. With dielectric thickness going down I 
> find my line widths are actually getting too narrow.

If you mean impedance controlled, this is not an issue.  The lines are
still directly above/below the coupling plane.  It's simply the trace to
pad clearance that causes a DRC violation...

> In really tight areas where you have trouble getting one line through
(<
> 1mm BGA etc.) I use a constraint area to define a smaller line to via 
> spacing but also increase the line to line spacing in the constraint 
> area to Not Allow a route past pins with connections on that layer. 
> This will allow a route past unconnected pins and when the vias are 
> removed you have your additional spacing.

That would "work", but as I've suggested, simply not having the pad
there when it isn't used, is a very easy solution to this, and something
that, IMO, the tool should just do.  If it can do it post layout, why
not during layout?

Your suggested methodology, and my solution to use custom padstacks both
are "work-arounds" IMO.  The tool should just be able to do this, and
not require any work-arounds.

Regards,

Austin


-----------------------------------------------------------
To subscribe/unsubscribe: 
        Send a message to icu-pcb-forum-request@xxxxxxxxxxxxx
        with a subject of subscribe or unsubscribe

To view the archives of this list please login at
//www.freelists.org. Our list name is icu-pcb-forum or go to
//www.freelists.org/archives/icu-pcb-forum/

Problems or Questions:
        Send an email to icu-pcb-forum-admins@xxxxxxxxxxxxx

Want to post a job listing ?  DON'T DO IT HERE!  
Better yet, join our jobs listing forum.

SUBSCRIBE:  icu-jobs-forum-subscribe@xxxxxxxxxx
POST:       icu-jobs-forum@xxxxxxxxxx
-----------------------------------------------------------

-----------------------------------------------------------
To subscribe/unsubscribe: 
        Send a message to icu-pcb-forum-request@xxxxxxxxxxxxx
        with a subject of subscribe or unsubscribe

To view the archives of this list please login at
//www.freelists.org. Our list name is icu-pcb-forum or go to
//www.freelists.org/archives/icu-pcb-forum/

Problems or Questions:
        Send an email to icu-pcb-forum-admins@xxxxxxxxxxxxx

Want to post a job listing ?  DON'T DO IT HERE!  
Better yet, join our jobs listing forum.

SUBSCRIBE:  icu-jobs-forum-subscribe@xxxxxxxxxx
POST:       icu-jobs-forum@xxxxxxxxxx
-----------------------------------------------------------
-----------------------------------------------------------
To subscribe/unsubscribe: 
        Send a message to icu-pcb-forum-request@xxxxxxxxxxxxx
        with a subject of subscribe or unsubscribe

To view the archives of this list please login at
//www.freelists.org. Our list name is icu-pcb-forum
or go to //www.freelists.org/archives/icu-pcb-forum/

Problems or Questions:
        Send an email to icu-pcb-forum-admins@xxxxxxxxxxxxx

Want to post a job listing ?  DON'T DO IT HERE!  
Better yet, join our jobs listing forum.

SUBSCRIBE:  icu-jobs-forum-subscribe@xxxxxxxxxx
POST:       icu-jobs-forum@xxxxxxxxxx
-----------------------------------------------------------

Other related posts: