Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


C.K.

#9331

Poor HASL Finish known to cause solder shorts? | 24 September, 1999

Hi Folks:

We've been having bridging problems on our fine-pitch QFP's, despite thorough inspection of paste height, volume, and placement accuracies....profiles look good also.

The bridging is happening after coming out of reflow...one last thing that hasn't been investigated yet is inconsistent HASL finish.

Can a poor HASL finish cause bridging during the reflow process? How can a HASL thickness be measure on a bare PCB?

Thanks.

reply »

Earl Moon

#9332

Re: Poor HASL Finish known to cause solder shorts? | 24 September, 1999

| Hi Folks: | | We've been having bridging problems on our fine-pitch QFP's, despite thorough inspection of paste height, volume, and placement accuracies....profiles look good also. | | The bridging is happening after coming out of reflow...one last thing that hasn't been investigated yet is inconsistent HASL finish. | | Can a poor HASL finish cause bridging during the reflow process? | How can a HASL thickness be measure on a bare PCB? | | Thanks. | I know it sounds hoakey, but my new article should be out in SMTexpress within a week or so. Before that, here's some stuff about thickness.

It is well known that IMC is required for solder joint formation though the IMC layer itself is not a solderable surface. Therefore, say experts, there must be sufficient additional eutectic solder thickness to assure acceptable solder joint formation during assembly. Again, the conflict comes as how thick both the IMC and solder? If the IMC is too thick, and the solder is too thin (as is often the case with HASL coating), acceptable solder joints are not possible. If the IMC continues growth, becoming "extremely" thick - well, you know the outcome no matter the solder thickness. So, how much is too much and how does one control it? One doesn't.

Intermetallic compound thickness can be minimized depending on process parameters used. It is said by some that typical tin/copper IMC thickness is 6 to 12 micro-inches (.15 to .3 microns). In other hot air leveling processes, the IMC layer can exceed 100 micro-inches (2.5 microns). This is pretty contradictory and confusing stuff as "other" may mean those processes not so well managed as in different machine types and processes. That's a whole other concern because so many choices are available as vertical, horizontal, features, capabilities, cost, etc - making it hard for board shops and their customers to make "safe" decisions. How would the average board buyer begin to evaluate a board shop based on all the above?

Though difficult to measure HASL finish thickness, some findings show the range to be 80 to 1000 micro-inches (2 to 25 microns) on the same board. That's a wide range indeed. Again, this points to compromises, variables, and concerns facing all involved and affected. It comes down to how many compromises are acceptable and how each affects solder joint formation, quality, and reliability over time.

I've consulted with many operations wherein are found varying degrees of ineffective process management concerning HASL (there are none truly effective, nor can there be), and all going into it before attempting operations. Often, deleterious causes consist of poor flux management, improper or missing tools and procedures, air knives not adjusted or operating properly, solder pot dwell times and temperatures improperly set or operating, and poor control of boards introduced to the process. The latter often is the main problem as it is impossible to keep and assure copper board surfaces oxidation free. Therefore, the process often is "tweaked" mercilessly in futile attempts to overcome the main problem. This article now focuses on the process itself.

X-Ray fluorescence (XFR) is the primary measurement technique. You can see quite a range concerning thicnkesses. That's why it is not "specable."

Earl Moon

reply »

SP700avi inline stencil printer

Cogiscan Track Trace Control