I have posted on Usenet seeking collective wisdom on 'universal' ROHS solder
which is insenstitivy to alloying with other solders such that the weld is
reasonably good and close to eutectic behavior. The answers I got were no
encouraging. The 'hard old' (and tedious and bad for boards) method is the one
which works best, i.e. completely cleaning off the residue off the board and
iron tip (!) several times if necessary, and then using the new alloy to solder. 

I heard SN100C is very good for all-around use and low cost, and somewhat well
behaved when mixed with SnPb residues, and some audiophool threads claim that
quad eutectic solder might be 'very good'. I do not know enough metallurgy
physics to judge if bringing in a complex 'outlandish' alloy to the existing one
will somewhat mitigate the effect of the existing alloy on the result alloy.

This is relevant to the thread because of what happens when the alloy formed by
by the new and old ROHS solders is *not* eutectic. The behavior ranges from that
of immiscible alloys swirling together without alloying, to something that stays
liquid for a surprising amount of time after removing heat.

It is this latter behavior which is what I was trying to avoid, and *may* be
interesting for board-friendly chip removal. In short, one could deliberately
add an 'incompatible' (metallurgically, for soldering purposes) ROHS solder to
the joints being processed and then rely on the bad characteristics of the alloy
to remove the part easier. And it does not have to be Indium based ($$$
expensive). Most ROHS solders are not compatible with each other and with SnPb
so just picking two and trying might work.

Also some people use a solder bath to mass remove parts from old boards (in SE
Asia). This is the same solder bath used to make the boards usually. The problem
is that parts stuffed for wave soldering have bent pins which prevent them from
falling out in the bath. The SE Asia method to deal with this is a knife edge
that  rips the parts out. You probably want something more delicate, such as
tying a steel wire under the chip to remove and then use the bath and pull
(gently) on the wire loop. Heat resistant chip inserters/extractors can also be
used. The board is ALWAYS heated from the bottom, whether with a bath, a hot
plate or with hot air.

My favorite method to remove stubborn parts is to preheat the board with a 150C
hot plate. After that, I use normal solder wick and/or suction. The 150C heating
corresponds to the factory preheating used before the solder bath for pre-fluxed
boards (most of them) and is safe in most cases even if prolonged considerably
to achieve heat soaking. Solder flow with preheating is much better than
without, with the same power iron (usually 50W thermostated). By stubborn parts
I mean large non-removable heatsink parts, and certain flyback and SMPSU
transformers and inductors mounted with hollow rivet solder points to reinforce
the board.

Setting the iron temperature is more critical for desoldering than for
soldering, and a high power thermostated iron with a large sharp edged chisel
tip works best. The sharp tip is used to lift the 'bent down' pins of certain
through hole parts. To avoid damage to the pad it is necessary that this lifting
be done such that the other contact point of the 'fulcrum' is off the pad, on a
copper-less part of the board if necessary. The wide chisel chip usually permits
this. It also brings a lot of heat to the part where needed.

Most desoldering braids do not carry enough flux for sustained desoldering work.
I always add pure rosin (solid) to them either as powder or by pulling the braid
 through a rosin block with heat from the iron.

These technologies are now decades old, I worked for 14+ years almost
exclusively with SMD after that. That does not make them less good however ;)

-- Peter


-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist