Surface finish on bare printed circuit boards
For those of us who can, let us cast our mind back to the days when 12 thou gap and track was still a technological challenge and there was a rumour someone could drill and plate 0.6mm diameter holes. Surface finish was simple in those days. You had roller tin for single sided complete with trailing blips and reflow, complete with the cosmetic looks of Joan Rivers after she has her make up on.
Each finish was simple and few people offered gold in those days, although there were companies who used electro plated gold as an etch resist and suffered the undercut problems associated with the electro chemical reaction at the track edge.
Then came the options, Hot Air Solder Level, what a fantastic idea. Selectively plating the desired Pads with the very material you use to connect the components, solder. The buzz around about solder mask on bare copper, (SMOBIC, can’t remember what the I was for) was really exciting. Selectively putting solder only on exposed copper…..genius.
The process is simple, you heat up a bath of solder, dip the boards in and draw them out at a controlled speed across 2 hot air knives which both clean out the holes of solder and level off the pads. In the early days, this was a hugely expensive machine so a lot of companies sub contracted the process but the finish made it worthwhile. It was a major leap as it meant no solder under the resist so no chance of reflowing the solder during assembly with the chance of a breakdown in the resist integrity. The boards could now endure multiple heat processes with no problems
The fact that it reduced the meniscus effect or ‘’loaf of bread’’ as it was often described was a great plus for surface mount devices.
The machines have become more sophisticated and now, the finish is really desirable for 95% of products with excellent consistent flatness and solder ability, even using the Lead Free Hot Air Level.
Again, you are using the very medium that you connect the components with to coat the board making the solder ability reliable.
There have been a couple of finishes that have come and gone but the other ones that have survived are silver, Organic Surface Protection, and of course nickel/gold.
The major benefits of these processes is that they chemically deposit the finish so the surface accurately reproduces the copper plated finish. It is not prone to shorts at the tighter gap and track which on quad packs and the like is a great advantage.
These processes are self limiting on the protective deposits,once they have covered the copper, they slow down to almost zero.(Nickel in the case of ENIG). So OSP, silver and gold, although the nickel used underneath the gold will continue to build up on it self and is generally specified at 5 microns. The nickel does 2 things, it gives a hardness to the gold and also prevents the copper migrating through the gold which can cause unreliable joints if stored too long. It is very uncommon for gold to be put directly onto copper.
General comments:
Silver. It has emerged as a successful finish but cosmetically, especially on ground planes, gives the visual impression of it oxidising. Customers don’t trust this. As it burns off during the assembly process, it is deemed a non-functional fault. It is becoming less popular giving way to LFHAL and ENIG. It is very flat and follows exactly the copper plating.
OSP, it needs careful handling as acidic fingerprints will break it down. During test, the pressure often needs to be increased to break through the layer causing a crater in the pad. It needs to be stored carefully and the soldering processes will need to be changed to suit, possibly a more aggressive flux. It is a very good process, especially if the boards are straight to line.
One of the problems is that there is no visual change between the finish when it is good and not good. The only way to check is in process. Careful monitoring at stores is required to prevent problems.
ENIG (electroless nickel gold.) This finish is the preferred finish for any tight gap and track jobs. It gives a cosmetically pleasing product and the solder ability is excellent. It has a good shelf life promoted by the nickel interface and some customers specify it on all product, even though it has a cost increase over LFHAL. The main disadvantage is cost.
The surface finishes are subjective and there is a bit of ‘well that finish has always been used here’.
My personal preference is LFHAL every time from a quality/cost view point, until the gap and tracks reduce then straight to nickel gold.
I will be happy to discuss or put on the web any views or comments you may have.
Any suggestions for topics next month. If not, I will talk about metal backed products and other substrates.