By Mark Thompson | Published on: October 03, 2016
Our Customers are dealing with tighter and tighter pitch surface mount devices which require tighter and tighter CONTROL at Solder-mask to maintain the very small webs of mask left between the surface mounts themselves.
Traditionally, dealing with this at a design and layout stage meant reducing the trace widths to be able to rout between the mounts or pins.
At a Fabrication level this meant we have less mask clearance to be able to maintain webs of mask material between the surface mount clearances. This lead to people “gang relieving” the surface mounts leaving NO mask webs between them at the layout stage. Unfortunately this approach was used at a time when most parts were HASL ( Hot Air Solder Leveled ). The more malleable Tin Lead tended to “wick” or short between these mounts without that mask barrier to prevent the solder wicking.
Additionally at the Assembly level even if the Tin Lead did not wick between mounts at the Fab stage there was a good chance it would when assembled.
Lets talk “PITCH”.
Many times we are approached by customers talking about a specific “Pitch” between mounts Like .4mm or .5mm. Understand that without a design to physically LOOK AT via Image data it is difficult for a fabricator to properly understand what PITCH means to them.
The “Pitch” is the distance between the CENTER of a given entity to the CENTER of an adjacent entity. This can mean different things based on pad/mount size.
Let me give you an example:
.5mm “Pitch” means .0197 between the centers of two BGA pads or surface mounts. This actually seems like a very reasonable distance given today’s Circuit board geometries. However,
What if the designed surface mount pad width is .015 wide? This would mean a .5mm “Pitch” would leave only .0047 edge to edge between the mounts. This gives you little or NO room to rout a trace between mounts at that width and pitch.
Clearly as the “pitch” between parts decreases based on today’s smaller and smaller chip footprints the associated Surface mount or BGA pad widths need to decrease as well to be able to rout even a .003 or .004 trace between them. And what if the need dictates differential pairs routed between mounts?
At .003 traces with .0035 spaces as a Diff pair and .5mm pitch between parts, How skinny would my Surface mount pad have to be? Let’s do the math.
.0035 space Trace to mount plus .003 wide Trace plus .0035 space trace to trace then another .003 trace and lastly another .0035 space, that’s.0035 +.003+.0035+.003+.0035 or a total of .0165 mils.
At .5mm pitch between mount this would result in a surface mount width of about .003!! (.0197-.5mm minus .0165 = .0032)
This is not a practical number, So most folks avoid having to rout diff pairs between mounts or BGA pads in favor of single ended terminations. ( routing the diff pairs INTERNALLY where they are not affected by surface mount pitch.) This allows for a much wider surface mount or BGA pad size.
Now Let’s talk about “Pitch” vs. Webs or Dam’s of mask between mounts.
Even if the surface finish is NOT HASL, many times a small web or dam of mask material is desired between mounts.
Typical mask registration numbers are +/-.003.
What if the design was the same .5mm pitch between mounts and the mount width is .010.
In order to maintain a web .004 wide of mask material between SMT clearances the maximum clearance pads size for the .010 wide Mounts would be approx. .0027 per side, and whereas this is less than the normal desired Min expansion of .003 per side the .004 web can be maintained.
Lastly, Let’s talk about surface finish choice vs. mask geometries.
Let’s say you have a part that based on SMT pitch you cannot maintain both a .003 mask expansion and a .004 web between clearances. Would this be a good candidate for a HASL finish?
No, As mentioned above if no webs can be maintained between mask clearances this would allow the surface mounts to “wick” or short. This is where a thinner surface finish would be a good choice.
If the same tight pitch device that does not allow for a web of mask material to be between SMT clearances you may consider a surface finish like Imm Au or Imm Ag, Both are VERY thin depositions, (2-5um)and tend NOT to wick or migrate between mounts causing shorts.
Additionally, If you have a large SMT device with many pins and Co-planarity at Assembly is a potential issue. You may not want to choose HASL as a surface finish due to the much thicker deposition of HASL.
( again, typically 1 to 2 MILS as opposed to 2-5 um for an Imm Au or Imm Ag finish)
Any “Non Wetting” or areas where the HASL finish does not stick to the surface mount or pad as HASL could mean the difference in co-planarity of 1-2 mils, this difference at ASSEMBLY could mean some pins on the surface mount part may not make proper connection.
Copyright © 2013 Prototron Circuits, Inc. All Rights Reserved.