Steve,

We, at GenRad, have just started using a double sided refolw with selective wave 
solder with great success. We had the priviledge of stipualting the layout of the 
PCB so it was suitable for the process we had in mind.

This involves the same process you use (I assume) but instead of wave soldering 
the whole of the underside of the PCB we use a board carrier with appertures 
exposing only the through hole components to the solder wave, negating the need 
for gluing the SM components. Using this method also closes the process window on 
the wave solder process as you only have to solder through hole devices.

If your PCB is well down the design route then this may be of no use whatsoever 
but with teamwork and gentle(!) persuation we have managed some major advances in 
DFM.

Andy Richardson
Manufacturing Engineer
[log in to unmask]
GenRad UK

From:	MX%"[log in to unmask]"    21-SEP-1995 01:08
To:	MX%"[log in to unmask]"
CC:	
Subj:	Solder balls....

From: [log in to unmask]
Date: Wed, 20 Sep 95 11:37:27


  Good Day!
  
      I've got a question that is kinda' peculiar that maybe someone can 
  help me with. We're building products that are double-sided, and due to a 
  number of reasons (design mostly) we print solder paste on the bottom and 
  dispense glue as well. The reason we do that is to ensure that we get 
  good solder fillets on the locations that do not wave well either because 
  of orientation to the wave direction, or that the locations are shadowed 
  by other components. 
  
      Our customers supplier quality engineer has a concern with solder 
  balls that are embedded in the glue dots beneath the components...this is 
  mostly prevalent on 0805 components that I feel is occurring because of 
  the hot slump that occurs during reflow and the spacing between 0805 
  pads. We also see it on 1206 components, but not bad as on the 0805 
  chips.
  
      My question is, should this a really be a concern?...The balls ARE 
  glued, they ain't going anywhere, and they're not big enough to short any 
  conductors anyway. 
  
      But the quality engineer fears that it could lead to cross-talk if 
  there is enough of them in one spot that's close to a trace...is he 
  right? I'm not knowledgeable enough about that stuff to know...BTW, the 
  board operates at 100mhz...
  
                                           Thanks in advance!!!
  
  
       
                       ////  
                                           
                      |  @@ /      Steve Gregory                         
                      C   )        Process Training Engineer   
         ////  
       \_o         Pragmatech Incorporated               
        |  @@ /                    101 Nicholson Lane          
        C   )                      San Jose, California 95134            
         \_o        ////  
        [log in to unmask]         
                   |  @@ /         Phone:(408) 943-1151                  
                   C   )           Fax:(408)943-1461           
                    \_o                                        
   -------------------------------------------------------------
  
  ....the views expressed above are my own, not my employers...