When did the true HiRel world start sliding>COTS ? And why? And the slow
sliding ended in avalanche. How come? I think one of many reasons is the
fact that semiconductors have been incredibly much better. How often do you
have a processor problem with your laptop? And don't forget that they
contain millions of gates today. There is the trick: we think that the rest
of the components do the same improvement. Long nose. Cheated.  Signal
lamps die, press buttons crash, PWBs corrode, switches stop working. And so
do many components. Would you install USB 2.0 parts in your nuclear sub?
 If I had the responsibility, not one single USB had passed the entrance,
no matter what application.  Now, in reality that will hopefully not
happen...yet!

The HiRel requirement design needed lots of people with a solid background
in physics, chemistry, environment testing, failure analysis etc.  That
kind of experienced people are probably not so many today.  Question is if
we won't need them again one day. Or will profit hunger still dominate over
Reliability?

Inge


On 13 June 2013 20:12, Bob Landman <[log in to unmask]> wrote:

> Re thickness of the coating, a friend doing R&D on coatings comments:
>
> "Sometimes the type UR coating sticks poorly-enough and the whisker can
> peel up the coating (if it is thick enough and the whisker does not tunnel
> thru it) and create a circus-tent, and the whisker grows and collapses by
> Euler Buckling.
>
> Sometimes the type UR coating sticks well-enough to the tin surface that
> the whisker lifts and tears the coating and perhaps escapes.
>
> Sometimes tin whiskers.
>
> PCBS usually don't grow epoxy whiskers, and the Mil and IPC specs for
> coatings require they adhere well-enough to the laminate substrates of
> various test coupons, so folks rarely worry about pcb adhesion.
>
> Unless, of course, someone used some silicone thermal compound and left
> silicone fingerprints on the boards they sent over to be coated.
>
> Sometimes..."
>
> And then I read this article....
>
>
> http://www.militaryaerospace.com/articles/2013/06/lockmart-arci-contract.html?noredir=true
>
> The Navy has awarded Lockheed Martin a contract to upgrade the submarine
> fleet sonar with "COTS" computers!
>
> A friend at NASA comments:
>
> "There are many reasons to be afraid of COTS besides tin whiskers, and
> most important is the fact that we simply don't know how these parts and
> assemblies will perform. Will there be a delayed failure mechanism jumping
> out at us in a year or two (tin whiskers, electromigration of aluminum of
> Si die, cracks in capacitors, corrosion due to poor hermetic seals, etc etc
> etc)
>
> One may decide to screen the parts before putting them into the final
> assembly, but the beauty of MIL parts has always been ESTABLISHED
> RELIABILITY, which means that past performance could give us a piece of
> mind about their performance. It also meant that there was accountability
> for the failures. With COTS parts and assemblies - we have no idea how
> parts are made, how they performed, or who will pay for fishing out the
> submarine, once it runs into a coral reef because the sonar didn't detect
> it.
>
> Just yesterday I was listening to the talk about possibly implementing
> automotive-grade commercial parts into HiRel systems. The automotive world
> seems to have lots of checks and balances, that one participant who used to
> run factories producing automotive parts described as 'extremely invasive
> and thorough'. But those checks and balances are performed for individual
> companies buying up parts in millions and millions, and therefore expect to
> get such level of participation from the manufacturer. MIL world used to be
> the main customer on the block 50 years ago, which gave rise to the
> oversight that provided us with the highly reliable parts. But nowadays - a
> few submarines, a few airplanes and a few satellites are peanuts as
> compared to fleets of cars out there. And our guest that used to run
> factories honestly acknowledged, that if we were to step into the factory
> and ask for the level of scrutiny he agreed to for his automotive high
> volume customers - we would be shown the door."
>
> Bob Landman
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Stadem, Richard D.
> Sent: Thursday, June 13, 2013 9:03 AM
> To: [log in to unmask]
> Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> whiskers
>
> Regarding #3 in Joyce's post below, tin whiskers in space are several
> thousand times more likely to lead to catastrophic electrical failure than
> tin whiskers in electronics here on earth.
> In space, in a (nearly) perfect vacuum, there is little or no oxygen. A
> tin whisker short can handle much more current than it can here on earth.
> Most tin whisker shorts on earth actually go undetected, because the power
> going through them will burn them away before the current surge can cause
> any real harm in most cases. In space a tin whisker will pass many
> milliamps of current for a very long time,  enough to cause the device to
> fail.
> So as Inge pointed out, although tin whiskers are rare, once is too many
> times for a satellite costing millions, or for any avionics, traffic
> controls, etc, where people's lives are at stake. Or for braking systems in
> automobiles, or life sustainment systems in the medical field, or......
>
> And I totally agree with Joyce's point number 4, why a total ban if as
> little as 3% lead in the solder will greatly reduce any chance of
> formation? RoHS regulations do not allow more than .1% lead in solder. Why?
> But then again, why any ban in the first place?
>
> Because, in the European Union, they did not perform due diligence before
> passing the lead ban into law. The politicians fed on the public's
> ignorance. A total ban was the only thing acceptable, even though there was
> no proof that the presence of ANY AMOUNT of lead in solder presented any
> health threat to us. In fact, there was plenty of evidence at the time to
> the contrary, but who cared about that? The witch hunt for lead was on.
>
> As each and every day passes, the stupidity of the RoHS ban of lead in
> solder continues to be made more manifest than it was yesterday.
>
> Has there been any re-consideration by RoHS legislators regarding the
> total ban of lead in solder?
>
> If not, it makes the stupidity even more convincing.
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Joyce Koo
> Sent: Wednesday, June 12, 2013 10:01 PM
> To: [log in to unmask]
> Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> whiskers
>
> (1) Saw it in early 90s on connectors. Pure tin plated. It is hard to do
> FA on field return, you lost the evidence when you un plug the board.
> (2) Bad case when you have tin or zinc plated Tiles for cleanroom or on
> the roof.  Fan or air condition can blow them around - random failure in
> equipment.
> (3) NASA case is more persevered may be due to lack of disturbance in
> space.
> (4) Still don't understand why ban Pb, if add few % can eliminate it.
> (5) Good design - material and process can prevent it for sure if it NOT
> follow the politically correct way. It is a man made problem
> (unfortunately, those men are dress better, more powerful than you and me).
> --------------------------
> Sent using BlackBerry
>
>
> ----- Original Message -----
> From: Inge Hernefjord [mailto:[log in to unmask]]
> Sent: Wednesday, June 12, 2013 05:43 PM
> To: [log in to unmask] <[log in to unmask]>
> Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> whiskers
>
> NASA have the most spectacular registry of whisker growths, some of which
> are compatible to sort of metallic psitacosis. E.g. whole cable 'ladder'
> supports covered with Tin 'wool'. If we limit the tour through the NASA
> Whiskerland to PWBs only, then how many of you have seen real schrecklichen
> cases? I have been working for a company that has produced 10ths of mllions
> of boards, and I have just found a handful serious examples. We DID have
> some really pudgy issues (loved to perform analysis), but they were beside
> the PWB world. The vigilance on the whisker theme at Ericsson has been
> active for years, despite that very few boards were reported from the
> myriade of customers. Therefore I have the question: what are the risks of
> tin whiskers today?
>
>
> On 12 June 2013 21:28, Stadem, Richard D. <[log in to unmask]>
> wrote:
>
> > Only a lurdane engineer would disagree with you on that clarification,
> > Wayne.
> >
> > -----Original Message-----
> > From: Wayne Thayer [mailto:[log in to unmask]]
> > Sent: Wednesday, June 12, 2013 1:54 PM
> > To: Stadem, Richard D.; TechNet E-Mail Forum
> > Subject: RE: [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> >
> > Perspicacious?  Damn, Richard, that's a big word!  Had to look that
> > one up.  No disagreement with the basic sentiments you've expressed:
> > Using the data as a guideline for how to prevent problems with tin
> > whiskers would be wrong, and peeling conformal coating is rarely a good
> thing!
> >
> > -----Original Message-----
> > From: Stadem, Richard D. [mailto:[log in to unmask]]
> > Sent: Wednesday, June 12, 2013 2:24 PM
> > To: TechNet E-Mail Forum; Wayne Thayer
> > Subject: RE: [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> >
> > I am not sure I follow your first statement, Wayne.
> > It would be foolish to treat this as just a strange anomaly from just
> > a conformal coat bonding perspective.
> > It would be perspicacious to treat it as a reliability issue from the
> > perspective of a catastrophic failure due to electrical shorting
> > issues, whether through or under the coating.
> > NASA purposely selected a rather benign environment to show that
> > whisker growth can take place both through and under coating in an
> > office environment (to illustrate that even in a best-case scenario
> > they will
> > grow) and they were using tin-plated brass simply because that is more
> > prone to provide the stress interface that tends to produce tin whiskers.
> > However, in real life, and depending on the product, you would
> > typically have a much worse thermal cycling scenario, and plenty of
> > stress interfaces to grow plenty of whiskers. Just because they did
> > not penetrate the 2-mil thick coating does not mean they won't short out
> under the coating.
> > Not to sound like a metal farmer, but tin whiskers do not grow every
> > so far apart like planted corn; in certain situations they will sprout
> > like grass under a rail fence in an Iowa pasture.
> >
> > The separation of the coating is a secondary reliability concern
> > compared to the electrical shorting possibility under the tent. But
> > both are a concern.
> >
> > -----Original Message-----
> > From: TechNet [mailto:[log in to unmask]] On Behalf Of Wayne Thayer
> > Sent: Wednesday, June 12, 2013 12:44 PM
> > To: [log in to unmask]
> > Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> >
> > Awesome presentation, thanks for posting the link.  But it would be
> > foolish to treat the results as anything but a strange anomaly which
> > may or may not have applicability in "the real world".
> >
> > For those who didn't check it out, this is a NASA study where they've
> > been watching tin whiskers grow on a piece of tin plated brass sitting
> > in an office environment for 11+ years.  The polyurethane conformal
> > coat is on half of it.  No whiskers penetrate the 2 mil conformal coat
> > because the combination of this urethane's high internal cohesive
> > strength, it's flexibility, and it's relatively poor adhesion to
> > plated tin cause any whiskers to push the coating off of the
> > surrounding flat (unwhiskered) tin plating, creating a "circus tent"
> > structure around the whisker, with the whisker as the center pole.
> > The material continues to delaminate until the load on the tin whisker
> is high enough to cause it to fail by buckling.
> >
> > Whether the polyurethane adheres to the pcb assembly being coated in a
> > comparable way is open to speculation.
> >
> > -----Original Message-----
> > From: TechNet [mailto:[log in to unmask]] On Behalf Of Phil Bavaro
> > Sent: Wednesday, June 12, 2013 10:28 AM
> > To: [log in to unmask]
> > Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> >
> > I did find the paper that started the discussion about the magic
> > thickness being .002" but I still agree with the general rule stated
> below.
> >
> >
> > http://nepp.nasa.gov/whisker/reference/tech_papers/2010-Panashchenko-I
> > PC-Tin-Whisker.pdf
> >
> > I will patiently wait for Dave to come back.
> >
> > Phil
> >
> > From: [log in to unmask] [mailto:[log in to unmask]]
> > Sent: Wednesday, June 12, 2013 6:10 AM
> > To: TechNet E-Mail Forum; Bavaro, Phillip @ MWG - TW
> > Subject: Re: [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> >
> > Phil,
> > While this is an answer I "should" know, I don't.  Dave Hillman
> > regularly attends and presents at the CALCE yearly conference on
> > whiskers and so he keeps up on all of that.  At present, my esteemed
> > colleague is bumping his head on rocks, kayaking upside down, on some
> white water in North Carolina.
> >  He should be back in the office on Monday and will no doubt answer then.
> >
> > From our discussions, the general rule is still "no conformal coating
> > prevents whiskers".  A thicker coating may cause the whisker to expend
> > more energy punching through and yet more energy to punch through an
> > adjacent coating on a lead (usually resulting in buckling), but I have
> > yet to hear about some magic thickness of any kind of coating that
> > completely mitigates whiskers.  But I could be wrong.
> >
> > Dave?
> >
> > Doug Pauls
> >
> >
> >
> > From:        Phil Bavaro <[log in to unmask]<mailto:
> > [log in to unmask]>>
> > To:        <[log in to unmask]<mailto:[log in to unmask]>>
> > Date:        06/11/2013 02:26 PM
> > Subject:        [TN] minimum thickness of Type UR Conformal coat and tin
> > whiskers
> > Sent by:        TechNet <[log in to unmask]<mailto:[log in to unmask]>>
> > ________________________________
> >
> >
> >
> > Doug et al,
> >
> > Is there a disagreement in the industry as to what minimum thickness
> > of urethane is required in order to mitigate tin whisker concerns?
> >
> > I am hearing that the .003+/-.002" does not provide enough of a minimum
> > thickness and that the number is as high as .004".   I can understand
> > wanting the minimum being raised to .002" but higher than that would
> > seem to make the process much more difficult to control.
> >
> > I have a potential customer asking if we measure the thickness on the
> > individual component leads which is another can of worms it seems.  We
> > always used flat samples to document our thicknesses.
> >
> > I did not get to attend this years APEX so I might have missed the
> > latest data.
> > ________________________________
> > This message and any attachments are solely for the use of the
> > addressee and may contain L-3 proprietary information that may also be
> > defined as USG export controlled technical data. If you are not the
> > intended recipient, any disclosure, use or distribution of its content
> > is prohibited. Please notify the sender by reply e-mail and
> > immediately delete this message and any attachments.
> >
> > ______________________________________________________________________
> > This email has been scanned by the Symantec Email Security.cloud service.
> > For more information please contact helpdesk at x2960 or
> > [log in to unmask] <mailto:[log in to unmask]>
> > ______________________________________________________________________
> >
> >
> > ______________________________________________________________________
> > This email has been scanned by the Symantec Email Security.cloud service.
> > For more information please contact helpdesk at x2960 or
> > [log in to unmask]
> > ________________
> >
> > ______________________________________________________________________
> > This email has been scanned by the Symantec Email Security.cloud service.
> > For more information please contact helpdesk at x2960 or
> > [log in to unmask]
> > ________________
> >
> > ______________________________________________________________________
> > This email has been scanned by the Symantec Email Security.cloud service.
> > For more information please contact helpdesk at x2960 or
> > [log in to unmask]
> > ______________________________________________________________________
> >
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
>
> ---------------------------------------------------------------------
> This transmission (including any attachments) may contain confidential
> information, privileged material (including material protected by the
> solicitor-client or other applicable privileges), or constitute non-public
> information. Any use of this information by anyone other than the intended
> recipient is prohibited. If you have received this transmission in error,
> please immediately reply to the sender and delete this information from
> your system. Use, dissemination, distribution, or reproduction of this
> transmission by unintended recipients is not authorized and may be unlawful.
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
>
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please contact helpdesk at x2960 or [log in to unmask]
> ______________________________________________________________________
>


______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________