TIP: A simple and safe way to desolder SMD SOICS etc

[N_Cook]

I discovered this via needing to isolate Vcc of a SOIC where the trace
is under the IC and a multilayer board.
I have a reel of 0.08mm diameter tungsten wire , tungsten probably not
critical and diameter only needs to be small enough to go under SOIC pins.
I've since tried, pin by pin, of an 8p SOIC on a scrap board with a
success rate of 8 out of 8 isolating each pin , first time. Leaded
solder, PbF situation not tried yet

Feed length of the wire under a pin , long enough to grab hold of both
ends. Pull outwards and slightly upward while quick soldering iron melt
of the pin solder. The wire pulls through, but surprisingly , 8 out out
8 times so far, the solder has not rejoined across the gap. Then a
sliver of mica under the pin , in my Vcc situation, to fully
isolate for testing purposes.
But of course where hot air is out of the question , the IC is required
for salvaged re-use etc, this method pin by pin could be used even in
restricted space,
to salvage a whole SMD IC. I wonder if a single loop per side of a DIL
or quad footprint IC , and progressive single pin melting , would work
with 8 or more pins in one operation - worth trying some time

 

[Mr. Man-wai Chang]

Surface-mounted stuffs are not supposed to be repaired as traditional
chips, right?

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora 19 i686) Linux 3.9.9-302.fc19.i686
^ ^ 22:54:01 up 21:06 0 users load average: 0.00 0.03 0.05
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[N_Cook]

I'd try to break the trace leading from the pad first.

Not accessible in tnis case, its buried inside the board somewhere unknown.
While at it for traces that are accessible. Sometimes in those sets of
Cintride/"diamond" abrasive addons for Dremmels, a diabolo shaped one.
Like 2 cones back-to-back, the 3mm or so diameter outermost cone base is
excellent for cutting neat + tight slots in fine pcb copper traces.

I will try kevlar fibre from an optical data link instead of tungsten
wire, to see if that works also

 

[N_Cook]

A twist of kevlar fibres works if you can get the cord under and around
the pin

 

[N_Cook]

I detailed this procedure a long long time ago on here. I use
etch-revision wire, fed under the pins, and tacked to a convenient
solder point at the far end. With the wire pulled against the first pin,
and slightly upwards, applying the tip of the iron to the pins in quick
succession results in them popping off the board remarkably easily and
cleanly. I have used this technique with total success to remove SM ICs
without physically damaging them , and where hot air cannot be used, for
many years now.

Arfa

Etch revision wire must be compatible with solder, I would have thought
the technique would require an incompatible material

 

[N_Cook]

I detailed this procedure a long long time ago on here. I use
etch-revision wire, fed under the pins, and tacked to a convenient
solder point at the far end. With the wire pulled against the first pin,
and slightly upwards, applying the tip of the iron to the pins in quick
succession results in them popping off the board remarkably easily and
cleanly. I have used this technique with total success to remove SM ICs
without physically damaging them , and where hot air cannot be used, for
many years now.

Arfa

Must have been on a thread I was not following. I think I would have
tried it myself if I'd seen it.
Just tried removing a 50 pin quad format SMD IC. Convenient tiny plated
hollow vias all over the board to pass the tungsten wire through. First
run I decided to run flux over the pins to assist desoldering but that
made the solder recoalesce on half the pins. As tungsten wire is awkward
to handle, used an artery forceps to clamp the wire under the pcb via
and another to hold the free end of the wire, and tension angles just
slightly off line to avoid bending the pins.
Tried another 50 pinner and 4 runs with satisfying twang on release each
time, with only 2 pins having rejoined solder under , which of course is
no difficulty to deal with individually when the other 48 are released.
Only a couple of minutes to desolder a 50 pin SMD

 

[N_Cook]

I've not come across the term or actuality of revision wire, only pcb
trace repair copper foil of various thicknesses/widths. Tungsten is
surprisingly strong for only 1/10mm diameter, which is probably an
advantage for this use and quite rigid for the insertion under the pins
process. Ill probably ditch the masking off and hot air desoldering of
ICs and move over to this method. I used tungsten wire there but only to
apply lift-off force to the IC while applying hot air.

 

[N_Cook]

I find solder perfidious stuff. Cut a trace for diagnosis purposes and
the solder will not bridge the gap when it comes to repairing the cut.
Resolder one pin of a SMD and there is a good chance of soldering to
the neighbouring pin and for good measure some of the solder migrating
under the IC. I would never of believed that running a wire between pcb
pad and pin of a SMD swathed in molten solder , would split the solder
into 2, the vast majority of times, so would never have considered
trying it.

 

[N_Cook]

I detailed this procedure a long long time ago on here. I use
etch-revision wire, fed under the pins, and tacked to a convenient
solder point at the far end. With the wire pulled against the first pin,
and slightly upwards, applying the tip of the iron to the pins in quick
succession results in them popping off the board remarkably easily and
cleanly. I have used this technique with total success to remove SM ICs
without physically damaging them , and where hot air cannot be used, for
many years now.

Arfa

"here" being usenet but on uk.d-i-y, end apr 2013, a place I rarely
visit except to post a buildings type query

 

[N_Cook]

I just noticed the production date on my roll of tungsten wire - 1954, I
think it came from auction sell off of stuff at CEGB Marchwood,Hampshire
labs

 

[notme]

Surface-mounted stuffs are not supposed to be repaired as traditional

I believe he was describing a diagnostic procedure, not a chip repair.

I believe he was being humorous. (note smiley...)

 

[N_Cook]

That sounds like the 30 AWG wire that used to be used to wrap wire
backplanes. I have several rolls of solid wire in that size that is
tinned, and has a thermoplastic insulator, commonly used for ECOs
(Engineering Change Order)

or even just plain old wire-wrap wire, it has to be strong enough to be
pulled tight over those square corners of the pins without fracturing.
If I can remember where it is I'll dig out some of it and see how
solderable it is and whether it works for this SOIC removal purpose

 

[N_Cook]

That sounds like the 30 AWG wire that used to be used to wrap wire
backplanes. I have several rolls of solid wire in that size that is
tinned, and has a thermoplastic insulator, commonly used for ECOs
(Engineering Change Order)

Tried some ancient Bicc-Vero wire-wrap wire (Kynar sleeving stated on
the drum, order code faded with age), surprisingly as made just down the
road from me, stated as 30AWG. I don't know how to strip off long
lengths of the sleeving but for 6 inch length, grab stripped wire end in
pliers and compress by pushing with fingernail+finger, the Kynar/ptfe?
along and off the wire. Incidently , by heating and stretching this , in
effect sleeving, will form even finer spaghetti tube. The wire does take
solder and I can confirm can be used for SMD SOIC removal