To web or not to web?  No facts here either but some interesting
	perspectives on the problem.

	I've seen a move away from webs with one manufacturer, another one
	go into "hiding" on the subject, and a case where the webs are man-
	datory.  (we support several distinctly different DFM specs)

	Against:
	This relatively recent update (<1 yr) changed to gang sol-
        dermask voids on pin fields of less than .031" pitch.  One
        driver for this was trouble with solder balls trapped between
        land and soldermask directly out of HASL process.  This occured
        when the tug-of-war between dam thickness, land spacing and toler-
        ances put the LPI at less than .003" away from the land.
 
	Dunno:
        Another spec calls for zero soldermask-oversize, to keep the data-
	base independent of soldermask technology (moving target?). One
        caveat: when using this spec, we don't have a good way of know-
        ing where silkscreen is permissible, since we really don't know
	what the final soldermask will look like.  I dislike "not needing
	to worry".
 
        The above are both reflow-assembly based cases.  The following was
	actually pertinent to reflowing pin-through-hole.

	For:
	An preliminary case where soldermask was depended upon for gasket-
	ing the solder stencil, intending larger paste deposit than the
	pad itself.  The thin web kept definition pad-to-pad and kept the
	wicking predictable.

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

-----------------------------Reply Seperator----------------------------
Jim Williams wrote:
>      Some designers make an effort to maintain a web of soldermask between
>      the solder pads on SMT devices, where pad spacing permits. When the
>      spacing of fine pitch devices preclude the web, it is omitted. I have
>      heard that the webs are used to reduce solder bridging.
> 
>      Thinking through this issue, I find webs being used where pad spacing
>      is large enough so as to make solder bridging unlikely, while webs are
>      omitted when the pad spacing is close enough to make solder bridging
>      most likely.
> 
>      I am interested in any factual data that supports the reason for, or
>      the value of, soldermask webs.