This was an offlist response by a friend to a copy of the "best
desoldering tool" thread. As it deals as much with soldering as
desoldering and as it gives a fairly detailed commentary on a variety
of methods it seemed worth positing in its own right.

____________________________

Ken says:

We have a Hakko 936 temperature-controlled soldering station, a Hakko 850
hot-air SMT rework station, and a Hakko 474 vacuum desoldering station.

All three tools work very well and I'm confident to use them to solder and
desolder pretty much any type of component having any lead pitch on boards =
of
any number of layers (with the exception of BGA and similar high-density la=
rge
pin-count leadless packages).

Soldering fine-pitch SMT devices using the 936 requires a wedge type tip, l=
iquid
flux, and sound technique  - but once you get used to it and have practiced=
 a
bit you can for instance lay down and solder a 240-pin 0.5mm pitch QFP in l=
ess
than 30 seconds.  I believe that you can get special SMT soldering tips tha=
t are
in the form of a sort of "spoon" that can be charged with solder  - but I've
never bothered as I find that the standard Hakko SMD wedge tip (Hakko 900M-=
T-K)
works well enough for me.

The basic technique is to first place the IC on the board, accurately align=
 all
leads with the pads, apply a small amount of liquid flux to the corner pins,
press down on the package gently (I used the tips of a pair of fine tweezer=
s)
and tack a couple of the corner pins.  Next apply liquid flux liberally to =
the
leads along one side, charge the wedge-type tip with enough solder to do al=
l the
leads on one edge of the IC, and then just drag the sharp edge of the along
(pulling the iron towards you) in contact with the pads using the lead-ends=
 as a
guide.

Sometimes you will get solder bridges between leads but these are easily de=
alt
with by either wiping the tip (with as little solder on it as possible) alo=
ng
the offending leads in an inwards-to-outwards direction, or if necessary by
using desoldering braid.  Extra liquid flux is also often helpful when deal=
ing
with solder bridges.  For IC's having leads on four sides solder the opposi=
te
side next and then the other two.  If some leads fail to solder on the first
pass just apply a bit more flux (and perhaps put a bit more solder on the t=
ip)
and try again.

Soldering passives with two leads using the 936 is also very easy.  You pla=
ce
the component (say an 0805 resistor) so it overlaps the pads by about 25-30=
% (no
need to be too fussy with alignment) and apply liquid flux.  Charge the wed=
ge
type tip with an appropriate amount of solder (it takes a while to get used=
 to
judging that), and bring the tip into contact with the two pads (as near to
simultaneously as possible).  The solder will flow across the pads and when=
 it
reaches the device the surface tension of the solder will pull the device i=
nto
proper alignment relative to the pads, and the tip can then be withdrawn.  =
Done
right the soldered joint should be every bit as good as one done in a reflow
oven.

I generally set up a number of passives and then solder them all one after =
the
other.  Once you have practised the technique you can lay down and solder
passives and other 2-leaded SMT devices surprisingly quickly.

I almost never use the 850 for soldering as solder paste is a pain.  Not on=
ly
does it take a comparatively long time to apply to the pads, it always drie=
s out
eventually in the syringe and you are for ever cleaning the hypodermic.  You
also have to be much more careful in placing the IC after applying the past=
e.
It's much better to use the 936.

Desoldering IC's etc. from SMD boards using the 850 requires an appropriate
"nozzle" for the target package  - and they are relatively expensive.  We h=
ave
built up a reasonable selection of nozzles and have on occasions temporarily
adapted a nozzle to suit a different package using adhesive metal foil, or =
thin
tinplate or aluminium.  Just heat all leads of the device evenly, while
maintaining a gentle sideways force on it using the tips of fine stainless =
steel
tweezers.  Alternatively use thin piece of stainless steel wire slid under =
one
corner of the device.  As soon as the device moves sideways withdraw the no=
zzle
and flick the device off the pads.  With a bit of practice the device will =
come
away cleanly on the first attempt.  Don't apply too much force and don't ap=
ply
it too early or you run the risk of lifting pads.  Clean up the pads with
desoldering braid and the 936  - taking care to always wipe along the longer
axis of the pad from outside to inside, and clean any flux using an appropr=
iate
flux cleaner on a rag.  If done right the board should look as if an IC has
never been in place.

Smaller SMD devices with relatively few pins can best be desoldered using a
round nozzle on the 850.  We have a 3mm nozzle (good for discrete semis in
SOT23's and small IC packages) and a 5mm nozzle (good for larger discretes =
like
rectifiers and somewhat larger low pin count IC's).

Note that if a high pin-count fine-pitch QFP is KNOWN to be faulty, the saf=
est
approach to its removal is to use a scalpel to shear off all leads right at=
 the
point where they enter the package moulding, let the package body fall away=
, and
then use the 850 with a 3mm nozzle and tweezers to remove the leads a few a=
t a
time.  Clean up the pads as described above.

The 474 excells are removing leaded components and has enough heat capacity=
 to
desolder large component leads (as on large electrolytic capacitors) from h=
oles
embedded in large copper planes.  The tricks are to use the right temperatu=
re
setting, keep all the filters clear (replace regularly), empty the solder
reservoir regularly, and to use a nozzle with the right hole size.  I have
adapted a number of smaller nozzles by drilling them out so I can better de=
al
with larger leads.  I also split the ceramic fibre filters in half by peeli=
ng
them apart.  Not only does this improve air flow but you only have to buy t=
hem
half as often.

One trick with the 474 (essential on boards having minimal solder) is to fi=
rst
apply solder to all leads of the target component so there is a generous am=
ount
of solder on each joint.  It is a bit counter-intuitive but it's necessary =
to
get a good thermal path from the tip of the iron to the joint (and the flux=
 from
the solder may help also).  The other trick is to apply the tip to the join=
t and
wait for the joint to heat up (and the for the solder to melt completely al=
l the
way through the hole and on the component side of the board) before pulling=
 the
trigger to apply vacuum.

The technique is to pull the trigger and then sort of rotate the tip so it
orbits around the lead a couple of times.  This causes the lead to orbit in=
 the
hole and helps encourage the solder to be sucked out.  Don=92t apply too mu=
ch
sideways force to the lead so there is a danger of bending or breaking it, =
and
apply only the minimum pressure to the pad (otherwise there is a danger of
damage while the substrate-copper adhesive is hot and soft.

Regards,

Ken Mardle
_______________________

_soldering

-- =

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