At 03:08 AM 5/7/2012, you wrote:

>On May 6, 2012, at 4:58 AM, Electron wrote:
>
> > what is the typical/standard size for a circular pad surrounding=20
> a hole of X mils, to get maximum reliability?
>
>Well, "obviously" the width of the ring of copper around the hole=20
>should be at least as wide as the minimum track width of the PCB=20
>process being used.  So if your have an 8mil design requirement for=20
>track width, you'd get padDiameter =3D holeDiameter + 16mil.  Add on=20
>tolerance factors for drill position

All this (and the difference between drilled and finished hole sizes)=20
is usually covered
by a specification from the board mfr called 'minimum annular ring'=20
or 'minimum ring'. It may
vary by layer and by copper thickness. For example (double and=20
multilayer boards)

Inner:
0.5oz: 5mil
1oz: 6mil
2oz: 8mil
3oz: 12mil
4oz: 15mil
5oz: 18mil
6oz: 20mil

Outer:
1/ 3oz -0.5oz: 4mil
1oz: 5mil
2oz: 7mil
3oz: 10mil
4oz: 16mil
5oz: 18mil
6oz: 20mil

To get the minimum pad diameter, take the nominal finished hole size=20
and add twice
the minimum annular ring.

Sometimes vias are different from pads (a different drilling process=20
may be used).

The above is for plated-through holes, where the main concern is that=20
the hole does
not completely break through the side of the pad. You may choose to=20
use larger pads
for physically stressed parts (switches, encoders, pots, connectors)=20
or heavy parts
(transformers, buzzers etc.) to improve reliability.

Single-sided boards generally demand much bigger pads than the=20
minimum the process would
allow, because the adhesion under the copper is all that's holding=20
the pads on. Usually you
would use large pads for heavy components, mechanically stressed=20
components, or
components with thick leads. It's not uncommon to use pads as big as=20
0.25" or larger
for some parts. Deciding how big is best is a bit of an art- if in=20
doubt it's best
to go bigger and add a jumper or two if necessary. Cheaper materials=20
tend to require
bigger pads than good materials for the same strength. Seems like=20
lousy processing
can also weaken the adhesion. If in doubt, try to follow consumer=20
designs subjected
to similar abuse. When the pad lifts and breaks right where the trace=20
meets it, it
can become a nasty intermittent fault that is hard to see.

--sp


--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.