Re: Soldermask Design Rules

Subscribe to Thread

I am looking for opinions on soldermask between the lands o... - Jun 13, 2000 by

Todd: You say: "... solder masking between the lands of ... - Jun 13, 2000 by davef

Dave, you are right you would expect the soldermask to be h... - Jun 14, 2000 by

Todd, two things: 1 On Doe I feature that one could be requ... - Jun 15, 2000 by davef

Hi Todd, it might be a philosophical question if soldermas... - Jun 16, 2000 by Wolfgang Busko

Todd Murphy

#3830

Soldermask Design Rules | 13 June, 2000

I am looking for opinions on soldermask between the lands of fine pitch components =< 25 mil. My personal opinion is that soldermasking between the lands of fine pitch components can cause stenciling problems and or manufacturing problems. If the soldermask height is greater than the height of the plating on the SMT lands then you may have a stencil gasketing problem. When this condition exists you will also have a greater amount of paste deposited to the pad than anticipated. This in turn can cause bridging. I guess you could look at this like a pond overflowing its levee.

reply »

davef

#3831

Re: Soldermask Design Rules | 13 June, 2000

Todd: You say: "... solder masking between the lands of fine pitch components can cause stenciling problems and or manufacturing problems."

I say: Fine pitch components without solder masking between the lands can cause stenciling problems and or manufacturing problems.

You say: "� If the solder mask height is greater than the height of the plating on the SMT lands then you may have a stencil gasketing problem."

I say: If the webbing between the fine pitch component land is higher than the plating on the land, why wouldn�t the mask be higher than all pads? ... regardless of pitch and regardless if there is webbing between pads or not.

Furthermore on a board with pads that are higher than the mask, without webbing, the stencil will not gasket well, partitions between the fine pitch pad openings on the stencil will bow downward under squeegee pressure, and deposit reduced amounts of solder.

Yano, opinions about solder resist-dams vary from "absolutely necessary" to "has no advantage at all". Board fabricators normally start complaining when lead pitches are at 20 mils or below.

I prefer webbing. Beyond the earlier points, boards with webbing between pads are: � Easier to "clean-up" after rework � Prevent "dross bridges" during hand soldering and rework.

What I really believe (no proof mind you) is that the benefit of solder resist-dams depends on the situation, it�s a function of: � Solder mask � Pad lay-out � Solder paste � Some combination of the above

So, that makes me want to do a DOE on new product lay-outs to determine the proper approach.

reply »

Todd Murphy

#3832

Re: Soldermask Design Rules | 14 June, 2000

Dave, you are right you would expect the soldermask to be higher than all SMT pads on the pcb if it is higher on the fine pitch components, but the other areas of the board do not concern me as much as the fine pitch areas. The pads on all the other components are generally much larger than the component lead itself when comparing it to the fine pitch lead to pad ratio. Thus this allows these components to be supseptable to addtional solder paste deposits. When designing a stencil, aperatures should be reduced such that they are smaller than the pads to insure proper gasketing and prevent squeeze out which is my main concern on the fine pitch component. I think that the distance between the pads on a component <25 mil is so small that stencil bowing can not occur between the leads when using a metal squegee. This might occur on 50mil or greater components. I appreciate your inputs and would like to hear about the results of the DOE if you get a chance to perform it.

reply »

davef

#3833

Re: Soldermask Design Rules | 15 June, 2000

Todd, two things: 1 On Doe I feature that one could be required for each combination. 2 I still don�t understand the problem. So, let�s talk through it:

So, this should not be a problem where the solder mask stand-off is less than the pad, where typically:

Liquid Photo Imageable (LPI) solder mask thickness => 0.0006" LT Total pad height = Thickness of ( Copper pad + OSP ) = 0.0014" + 0.00000157" = ~0.0014"

(You must be using an organic solderability preservative (OSP) or Electroless Nickel immersion gold (ENIG) because 0.003" tin/lead or white tin would increase the pad height further. Right?)

So is the problem that you�re posing is: Since copper thickness doesn�t vary much and OSP/ENIG thickness is not of consequence, when LPI thickness starts moving past 0.0014" (2X typical) thickness, you start having printing problems. Is that it?

No, that can�t be it, because this has nothing to do with webbing. I totally out of it!!!! Arrrrgh, someone else help this guy ( and me).

OK, OK. One more try. Todd you aren�t using dry film solder mask with fine pitch components are you?

reply »

Wolfgang Busko

#3834

Re: Soldermask Design Rules | 16 June, 2000

Hi Todd,

it might be a philosophical question if soldermask between finepitch pads is prefered or not. That should depend on the capabilities of a particular site and their experience. We actually prefer soldermask down to 0,4mm pitch and don�t have any problems with it nor do our CMs have any problems or complaints about it. We do have the feeling that it�s helpful in some way especially during rework and haven�t noticed any difficulties yet. The main influences on the amount of paste are stencil thickness and aperture design and with the figures Dave mentioned you can estimate how great the influence would be compared to other print and stencil parameters. IMO the decision to leave out the webbing between finepitch pads had a lot to do with the capabilities of boardhouses and that it nowadays isn�t a problem any more because the modern PCB-structures require more accuracy and the boardhouses adapted their abilities to it. In literature I read that there�s no noticable difference in the occurance of failures and up to now no one came and said this or that is better. So I trust my own experience and will go on with it until serious proof of the contrary.

Best thing would be you check with your assembly and do some DOE and let us all know what the results are.

This is just my oppinion and I might be awfully wrong

Wolfgang

reply »