Post mortem on an IEC connector

[Eeyore]

I'd agree with your analysis. I have seen both hairline cracks such as you
mention and 'cold joints' caused by failure to use an appropriate soldering
temperature, or soldering time when soldering components with large thermal mass
such an IEC connector PCB pin ( pressure will keep it working for a while but
it'll finally go high resistance ). Even seen that on TO-220 devices.

p.s. ( The TO-220 device acting as a heatsink to the solder via its leads, hence
preventing the alloying connection needed ). I've literally seen TO-220 devices
'fall out' of a PCB after a year or two leaving 3 nice rectangular holes in what
look to be perfect solder joints.

p.p.s The hairline crack around soldered joints is especially common where the
drilled hole size is not ideal ( oversize ) for the lead diameter. I have seen
manufacturers specify hole sizes very poorly with the inevitable outcome. Hence my
PCB footprints often use my own estimation of the appropriate drill size rather than
the data book one. Plenty of other people do the same.

Graham

 

[N_Cook]

60/40 is still readily available here in pretty much all of the varieties
that it always was - I got a new roll just a couple of weeks ago. What was
interesting though, is that they have jacked up the price to match - almost
exactly - the cost of the dreaded lead-free, which up until recently, was
like 50% more expensive.

Arfa

Do you know the most authorative source/group, academic or industry
expert/s, on tin-pest, lead-free solder, in-service degradation of solder in
connjunction with temperature and vibration. ie those issues, as distinct
from tin whiskers aspect that seems to have more coverage.

 

[N_Cook]

I spoke by email with a couple of industry advice bodies, when I was
preparing an article on lead-free solder for a mag. I'll see if I can find
the references for them.

Arfa

I did not find , on this site, the expected date for the results of their
study

CALCE Long-Term Pb-Free Study
http://www.calce.umd.edu/lead-free/longterm.htm

Generally, it seems to be just anecdotal evidence so far

 

[N_Cook]

My Dad - who taught me to solder some 50 years ago - use to say that a
good solder joint started off as a good mechanical joint.

As a teeenager, I probably spent thousands of hours wiring up
uniselectors and GPO relays with tinned copper wire and sleeving, on
every joint, the wire was either wound around the tag or passed through
a hole, if it was a lead through a circuit board, the wire was bent over
at right angles on the solder side.

We manufacturerd slot machines btw.

Ron

Very true , unfortunately assembly time/cost preclude that.
Hence the notorious guitar input socket problem. The sockets used to be
wired between the chassis mounted socket and pcb , so mechanically decoupled
from routine use of the socket. Now all soldered to the board and the
insertion/removal forces are transmitted to the pcb solder joint.

 

[Eeyore]

My Dad - who taught me to solder some 50 years ago - use to say that a
good solder joint started off as a good mechanical joint.

As a teeenager, I probably spent thousands of hours wiring up
uniselectors and GPO relays with tinned copper wire and sleeving, on
every joint, the wire was either wound around the tag or passed through
a hole, if it was a lead through a circuit board, the wire was bent over
at right angles on the solder side.

We manufacturerd slot machines btw.

Doubtless needed it to cope with the environment.

Graham

 

[Eeyore]

Even then if you bother to bend the tags over when replacing them, the
problem is less likely to recur.

Good point. I gather that Cadac 'hand wire' even pots to the PCB via short lengths
of wire btw.

Graham

 

[Adrian Tuddenham]

As far as I know, at this point in time, the avionics industry still has an
exemption from having to use the stuff. Now I wonder why that could be ... ?

They may be exempt, but what might they do when they can't get the
proper stuff any more because no manufacturer can afford to run two
production lines for the same product simultaneously and daren't risk
getting them muddled?

1) Set up special production lines.
2) Stop making avionics.
3) Change the rules and use lead-free.

 

[N_Cook]

There are no plans at the moment, from what I can gather, for manufacturers
to stop producing leaded solder, nor for distributors to drop it. There is
much 'legacy' equipment out there, which is manufactured in leaded solder,
and should not, as a consequence, be repaired using lead-free solder,
according to published wisdom from those who should know. There is no legal
requirement to use lead-free for the purposes of effecting such repairs. As
well as avionics, there are a number of other exemptions in important fields
such as medical equipment, and some military areas. Again, I wonder why that
could be ... ? As far as I understand it, the American military will not
tolerate the stuff being used in any of their equipment. How eminently
sensible of them. Pity we haven't got the same courage of our convictions to
stand up to the green mist brigade on this side of the pond ...

Arfa

For anyone in production :
I cannot reveal the company , Germany and UK , makes equipment that can be
found, not exclusively, in laboratories.
They got a derogation, allowing the continued use of leaded solder because
about 10 percent of their output goes to medical labs, not operating
theatres.
And deemed inappropriate to have to be forced to run 2 lines , 2 stocking
regimes etc

 

[Sjouke Burry]

Maybe after the Eurofighter fleet is grounded because of avionics
failures linked to LF solder?

Hope they don't use it to solder them Airbuses together

Nah... They use superior glue to hold them together....I hope..

 

[Eeyore]

They may be exempt, but what might they do when they can't get the
proper stuff any more because no manufacturer can afford to run two
production lines for the same product simultaneously and daren't risk
getting them muddled?

Fair comment. Most of the semi manufacturers' output goes to normal commercial
users who artifically 'need' lead free.

1) Set up special production lines.
2) Stop making avionics.
3) Change the rules and use lead-free.

Believe it or not there is now a new industry re-plating component leads with
Pb-Sn solder !

Graham

 

[Eeyore]

Pity we haven't got the same courage of our convictions to
stand up to the green mist brigade on this side of the pond ...

" green mist " ! I like it. We used to call it tomfoolery.

Graham

 

[Eeyore]

Hope they don't use it to solder them Airbuses together

But they're made of plastic ! ;~)

Graham

 

[Eeyore]

Nah... They use superior glue to hold them together....I hope..

Did you know the world's first jet airliner ( the DH Comet ) used 'redux' glue to
hold the majority of the airframe together ? Shame they didn't use it round the
windows and access panels, thinking rivets were more 'belt and braces' not
realising about microfractures from the drill holes. Caught out by a previously
unknown phenomenon. I can hardly believe they had prototype Comets flying around
1948 ! Britain was 100% in the lead at that time.

Graham

 

[N_Cook]

So do I, my friend, as I am about to get on one for the first time in
October. All of my previous cross-pond jaunts have been in properly built
747s, which have a proper yoke for the driver to hang on to, and
'automatics' that can be switched off. There's something fundamentally wrong
about a plane that has to be flown with a left-handed joystick, and which
employs a robot driver hidden away somewhere, which believes it knows more
about how to fly a plane, than the human guy and his chum in the co-seat,
who have 40 years flying experience between them ... :-\

Arfa

Fly by wire uses 3 hidden "driver" computers running 3 different bits of
firmware, then a democratic voting system that decides which 2 drivers to go
with, before making a control surface movement.
Now we recently know, in the UK, what such a voting structure produces in
the way of a parliament.

 

[Bob Larter]

That's Paul Goodman at ERA Technology. The final report is here.
http://ec.europa.eu/environment/waste/weee/pdf/era_study_final_report.pdf

The interesting bits are from page 73 onwards. Scary stuff!

 

[Bob Larter]

As far as I know, at this point in time, the avionics industry still has an
exemption from having to use the stuff. Now I wonder why that could be ... ?

Medical also.

 

[Eeyore]

So do I, my friend, as I am about to get on one for the first time in
October. All of my previous cross-pond jaunts have been in properly built
747s, which have a proper yoke for the driver to hang on to, and
'automatics' that can be switched off. There's something fundamentally wrong
about a plane that has to be flown with a left-handed joystick, and which
employs a robot driver hidden away somewhere, which believes it knows more
about how to fly a plane, than the human guy and his chum in the co-seat,
who have 40 years flying experience between them ... :-\

Shame that the most common cause of airliner accidents is still 'pilot error' !

Graham

 

[Eeyore]

there have been more than a few instances where true
fly-by-wire sytems of this type, have made a totally wrong decision in a
given unusual set of circumstances, and the driver and his chum have been
blocked from taking control to correct the situation.

Can you give a specific example ?

Graham

 

[N_Cook]

Is that one a 'fact' or a guess ? Again, I'll ask my mate tomorrow if he
knows any stats on that ...

Arfa

There was that French ? air show one, where the pilot quite gracefully, (no
frantic movements appeared on control surfaces or engine speed) flew into
woodland.

 

[Eeyore]

Is that one a 'fact' or a guess ? Again, I'll ask my mate tomorrow if he
knows any stats on that ...

It IS a fact actually. Go to any of the professional aviation sites.

You know than Spanair MD-82 ? that crashed in Barcelona on takeoff for example
not long back ? Pilots forgot to deploy the flaps on take-off because they were
in a hurry after a maintenance delay that COULD have alerted them to the problem
( WOG fault ).
http://en.wikipedia.org/wiki/Spanair_Flight_5022

That's a nice recent one. I could give you loads more.

The Turkish one at Amsterdam, stalled into the ground ( called CFIT in the trade
) Controlled Flight Into Terrain. Pilots not monitoring instruments and flight
attitude combined with an instrument fault ( pilots could EASILY have caught it
- they had 45 seconds to avoid a stall ).
http://en.wikipedia.org/wiki/Turkish_Airlines_Flight_1951

http://en.wikipedia.org/wiki/Comair_Flight_5191
Pilots took the wrong runway !

http://en.wikipedia.org/wiki/Comair_Flight_3272
Total failure of pilots to actually FLY THE AIRPLANE ! Captain reacted to the
stick pusher in EXACTLY the wrong way.

And it goes on and on.

Graham

WOG = wheels on ground btw.

due to the hugely increased level of spam please make the obvious adjustment to
my email address