I believe that If you are having trouble soldering to both the PCB and the 
components your problem lies in the solder paste or the reflow profile.

We have been using Ni/Au plated boards on one line for several months now with 
no real solderability issues. We are using no-cleans also, and your WS-609 
aqueous paste should be even more active than our paste. I don't believe playing
around with nickel and gold thicknesses will help you for the two reasons. 

First your gold layer is an probably an immersion (chemical displacement) 
deposit with limited plating thickness deposition. I don't believe you have the 
option of increasing gold thickness as the plating stops at about the thickness 
you mentioned in you text. This is also too little gold to cause gold 
embrittlement, although your solder joints may net be as bright as those made to
HASL PCBs.     

The nickel thickness should have no effect on your solderability.

Since you are using nitrogen the areas I would focus on would therefore be bad 
solder paste or the incorrect solder reflow profile.

Check the recommended solder reflow profile with Alpha for the WS-609 and try to
get a fresh batch of paste.

Regards,

George Ead

     George,
     
     You should be able to help this guy....


______________________________ Forward Header __________________________________
Subject: soldering gold pcb,s
Author:  [log in to unmask] at Internet
Date:    10/6/95 10:46 AM


     
     We are having some problems soldering to an electroless gold over 
     nickel card.  specifics are:
     
     5-10 microinches electroless gold over 20 microinches electroless 
     nickel.  .020 thick double sided smt card being soldered in a nitrogen 
     covection reflow oven. components range from 0603-1206 passives, sot 
     23 devices, and a 25 mil ssop. 
     
     Soldering is being done using Alpha 609 water soluble 63/37 paste with 
     a fairly standard profile.
     
     Results thus far are that the joints are granular, cold appearing, 
     with marginal wetting to both card and components.  There are also 
     numerous occurrances of voids/pinholes eminating from the lead pad 
     interface.
     
     Any insights would be helpful in the areas of:
     
                Profiles
                Pastes
                Methods for checking incoming solderability of the cards. 
                Any thoughts on optimal gold/nickel characteristics
     
     Thanks for the help.
     
     Steve Ethier
     Mgr. New Product Engineering
     Lockheed Martin Commercial Electronics 
     Tel: 603-885-8432