Reforming capacitors (technical description, not politics)

[Tothwolf]

On Wed, 29 Jul 2015, drlegendre . wrote:
> On Wed, Jul 29, 2015 at 5:19 PM, Fred Cisin <cisin at xenosoft.com> wrote:
>> On Wed, 29 Jul 2015, drlegendre . wrote:
>>
>>> Incidentally, what exactly differentiates a computer-grade cap from 
>>> any other alum. electrolytic?
>>
>> Maybe computer-grade don't need gold-plated oxygen-free leads?
>
> I get the jab you're taking at latter-day Audiophool idiocy, but you 
> won't find any gold-plated OFC business in any of the vintage gear I 
> typically work with.
>
> But as far as gold plating goes, gold is a good conductor, it solders 
> very well, it doesn't tarnish and its ductility promotes solid 
> connections on screw terminals - it's really these characteristics that 
> make it somewhat desirable in certain applications. So it's not so much 
> that gold "sounds better", it's that it allows one to make connections 
> that work better. But it does tend to wear quickly.

While I don't want to go /too/ OT here, I do want to clear up one common 
misconception regarding gold plating. It isn't that gold itself wets 
easily when soldering...the tin content of the solder actually dissolves 
the gold and the solder bonds to the underlying metal surface.

This can also lead to gold embrittlement of the solder joint, and while it 
generally wasn't as big of an issue with larger through-hole solder joints 
(such as with older semiconductors which had gold plated leads), it did 
become a major issue for SMD boards with gold plated pads, and especially 
with BGA components.

This is why in high reliability applications, industry practice has been 
to pre-tin gold plated surfaces which are to be soldered, using a two step 
solder bath. The first being more sacrificial which will collect much of 
the gold, and a second to catch anything which might still be left.