One more thought and I am going to shut up. I am for super clean
boards, and perfect coating please do not misunderstand me but, many boards
are digital circuit type require only 0.6 volts signals which is the diode
turn on voltage. Could that be a reason why not so many problems exist? Or
is it because dendrites take so long to form that by that time the boards
are absolete? Or is it because labs do not give MEs feed back about such
problems?
Regards,
Ramon
> -----Original Message-----
> From: David Douthit [SMTP:[log in to unmask]]
> Sent: Tuesday, September 09, 2003 11:59 PM
> To: Dehoyos, Ramon
> Subject: Re: [TN] Silicone conformal coating
>
> Ramon,
>
> Dendrite growth is a type of corrosion/metal reduction. As such there
> is very little need for "power". Corrosion is a form of ionized metal
> migration
> where moisture combines with contamination/dissimilar metals and creates
> its
> own voltage cells.
>
> Dendrite formation evolves pico/micro amps of current flow. The resulting
> dendrites are microscopic and
> will vaporize with just a few microamps of current.
> When they are discovered it's usually after several dozen dendrites have
> formed
> and vaporized
> in the same location. The most notable situation is when high impedence
> very
> slow logic lines are involved
> (such as reset or interrupt lines). An unknown reset or interrupt is
> generated
> and cannot be duplicated.
> After numerous attempts and repeated resets/interrupts it's time to get
> the SEM
> or high power (200+) microscope out
> and start looking. Sometimes all you find is residue vaporized metal mixed
> with
> carbon from damaged laminates.
> If you are lucky you may find partially formed dendritic paths that have
> not
> reached the anode path.
>
> In the case of BGA's this is very, very expensive and extraordinary
> processes
> are implimented (rigrous cleaning programs)
> to prevent these problems. These issues are of major concern to the US
> military
> because COTS components and circuitry are designed for
> speed and size reduction and not durability or longevity. Depending on the
> end
> use environment it can take 10 or more years for dendrites to form.
>
> You are correct that it "takes a certain amount of power" but the amount
> is
> very low!
>
> David A. Douthit
> Manager
> LoCan LLC
>
>
> "Dehoyos, Ramon" wrote:
>
> >
> > Hello David:
> > My degree is in electrical engineering so I understand something
> > about currents and voltages. Besides voltage you need current. In my
> > oppinion you need cetain amount of power to start the dendrite
> formation.
> > You can hold with your hands the output of a power supply that can
> produce
> > hundreds of amps and low voltage or the output of a high voltage and
> very
> > low current ( in the pico amps ) power source and not be hurt by it but
> you
> > can not hold a power source that can produce 480 volts and a few milli
> amps
> > because it can kill you ( one of my neighbors could hold a 240 volt
> outlet
> > bare wire that is why I chose 480 V ) I have seen so many boards that
> are
> > not cleaned properly. In some cases with hydrolyzed flux on some
> components
> > due to the cleaning agent being to weak or the flux not being cleaned
> > withing the hour or other reasons. . Specially PGA and BGA type of
> componets
> > are extra difficult to clean. Also conformal coating the joint of the
> > previous parts is not easy to accomplish. Yes we spend a lot of effort
> in
> > cleaning our boards at the present job, and try our best to conformal
> coat
> > but other than dipping the boards in CC as we used to do in a previous
> life
> > is nearly imposible to coat every solder joint. Traces could be solder
> > masked and be taken cared of them that way, but not solder joints. I am
> > speaking from experiences and not from lab tests that I have made.
> > My one and a half cent
> > Regards,
> > Ramon
> >
> >
> > > -----Original Message-----
> > > From: David Douthit [SMTP:[log in to unmask]]
> > > Sent: Tuesday, September 09, 2003 1:37 PM
> > > To: TechNet E-Mail Forum.; Dehoyos, Ramon
> > > Subject: Re: [TN] Silicone conformal coating
> > >
> > > Ramon,
> > >
> > > That is a very poor conclusion! Experiments done by IBM in the 70's
> > > showed that dendrites would form with as little as .5 VDC bias.
> > > Further, electrolysis of water can occur with as little as 2 VDC
> potential
> > > difference.
> > >
> > > The limiting issue is the amount of liquid water present. Corrosion
> > > engineers consider a
> > > layer of water molecules 5 deep to be a "liquid" film.
> > > That is the film will behave chemically the same as bulk liquid water.
> > > The limitation here is the volume of water in the film. The deeper the
> > > layer the more
> > > ions the water can transport. The chemistry that converts water vapor
> to
> > > liquid is very complex
> > > but it does occur on surfaces with as little as 60% RH in the local
> area.
> > > Just how long the water is there, the amount of water, and the amount
> of
> > > contamination all combine to
> > > cause several failure modes besides dendrites. There are numerous
> > > variables involved and
> > > the testing can be very expensive. But if you are building high
> > > reliability products this testing is needed.
> > >
> > > David A. Douthit
> > > Manager
> > > LoCan LLC
> > >
> > > "Dehoyos, Ramon" wrote:
> > >
> > > > After reading carefully the inputs from Per-Erik, Brian and
> > > others
> > > > it seems that the biasing between traces is a very important factor
> in
> > > > dendrite forming which are unwanted shorts. I know from first hand
> > > > obsevation that sometimes boards are not cleaned well enough under
> BGAs,
> > > and
> > > > CC under BGAs is done by pure accident in very few cases. So by
> > > conclusion
> > > > the traces between BGA balls do not have high enough bias for the
> > > dendrites
> > > > to form. Because they are not conformal coated , not cleaned
> properly in
> > > > some cases and RMA flux is used.
> > > > Regards,
> > > > Ramon
> > > >
> > > > > -----Original Message-----
> > > > > From: Brian Ellis [SMTP:[log in to unmask]]
> > > > > Sent: Tuesday, September 09, 2003 8:21 AM
> > > > > To: [log in to unmask]
> > > > > Subject: Re: [TN] Silicone conformal coating
> > > > >
> > > > > Per-Erik
> > > > >
> > > > > Unfortunately, I don't have time just now to discuss this further
> in
> > > > > detail. I agree with you on most points, but let's not forget that
> > > > > atmospheric humidity has four components:
> > > > >
> > > > > - H2O in clumps (mist): this is relatively harmless to coatings as
> it
> > > > > wets or sits on the surface, rather than penetrates
> > > > > - H2O as isolated molecules (true humidity). This penetrates all
> > > > > polymers ± easily
> > > > > - H+ as a result of hydrolysis: a very reactive and unstable
> reducing
> > > > > gas which can penetrate and react easily with many groups in and
> under
> > > > > the coating.
> > > > > - OH- as a result of hydrolysis: a very reactive and unstable
> > > oxidising
> > > > > gas which can penetrate and react easily with many groups in and
> under
> > > > > the coating.
> > > > >
> > > > > The hydrogen radical and hydroxyl group may be just as dangerous
> as
> > > H2S
> > > > > and are ever-present in air, constantly forming and recombining at
> > > > > random, under the influence of ozone molecules. The hydroxyl
> group, in
> > > > > particular, can play havoc with any loosely bonded hydrogen atoms
> in
> > > > > organics it comes in contact with, as H-OH is a much stronger bond
> > > than
> > > > > H-C (this is why many hydrocarbon vapours have a short atmospheric
> > > > > residence times: the molecules are literally torn apart by
> hydroxyls,
> > > > > during hydrolysis). As the global average tropospheric hydroxyl
> > > radical
> > > > > concentration is estimated at 8.1 x 10^5 radicals.cm-3 at any
> single
> > > > > moment, this is far from negligible.
> > > > >
> > > > > I feel this must also be a consideration.
> > > > >
> > > > > Best regards,
> > > > >
> > > > > Brian
> > > > >
> > > > >
> > > > >
> > > > > Tegehall Per-Erik wrote:
> > > > > > Brian
> > > > > >
> > > > > > What you are saying is that the function of the conformal
> coating is
> > > to
> > > > > act as an impervious barrier against humidity. Thus, a low
> > > permeability is
> > > > > crucial and the protection afforded is directly related to the
> > > > > permeability of the coating. Let's assume to begin with, that
> > > absorption
> > > > > of humidity will endanger the reliability of an assembly (which I
> do
> > > not
> > > > > agree with if the cleanliness is adequate). Then, why should
> > > permeability
> > > > > be more important than the water absorption level at equilibrium
> (that
> > > is,
> > > > > saturation level)? I guess that many assume that a high
> permeability
> > > also
> > > > > automatically means a high water absorption level but there is no
> such
> > > > > relation. Silicone materials have a very high permeability
> compared to
> > > > > most other coating materials but it is also a fact that the
> saturation
> > > > > level of water is lower for silicone materials than for other
> > > coatings.
> > > > > This means that the amount of water absorbed by a silicone coating
> > > likely
> > > > > is somewhat higher than for other coatings to beg
> > > > > in with, but after some time the other coatings will have absorbed
> > > more
> > > > > water. Which is then most dangerous, a high permeability or a high
> > > > > saturation level? As in many other discussions on TechNet, the
> answer
> > > is:
> > > > > It depends. In this case, among other things, on how long it will
> take
> > > to
> > > > > reach saturation.
> > > > > >
> > > > > > About ten years ago I did an experiment to prove that
> permeability
> > > of a
> > > > > coating is not crucial for the protection afforded. I coated a
> comb
> > > > > pattern on thoroughly cleaned HASL boards with one acrylic, one
> > > > > polyurethane and one silicone coating and Parylene, each with
> three
> > > > > different thicknesses. I then put the boards in a humidity cabinet
> and
> > > > > increased the temperature to 85 degree C but without any added
> > > humidity.
> > > > > When the temperature had stabilised, the humidity was increased to
> 85
> > > %
> > > > > RH. The surface insulation resistance (or rather interface
> insulation
> > > > > resistance in this case) for the comb patterns on the boards
> coated
> > > with
> > > > > the silicone coating decreased instantaneously and almost as fast
> and
> > > to
> > > > > almost the same level as on a non-coated reference board even for
> > > > > thicknesses of up to 150 um. So you are right, the permeability is
> > > very
> > > > > high for silicone coatings and the insulation resistance will be
> > > affected
> > > > > and, if that had been important, silicon coatings would be
> useless.
> > > How
> > > > > ever, the results very practically identical for the boards coated
> > > with
> > > > > the acrylic and the polyurethane coatings. The SIR decreased a
> little
> > > > > slower on the boards coated with Parylene but after about one hour
> the
> > > SIR
> > > > > was on about the same level on a board with 50 um Parylene as on
> the
> > > > > boards coated with the other conformal coatings. In fact, after
> longer
> > > > > exposure to humidity, the SIR on the Parylene coated boards
> actually
> > > > > decreased to the same level as for the noncoated board. That is,
> it
> > > > > decreased more for the Parylene coated boards than for the boards
> > > coated
> > > > > with the other coatings. These results show that SIR beneath a
> coating
> > > is
> > > > > affected very quickly even when Parylene is used as coating but
> they
> > > do
> > > > > not answer the question how long it will take to reach saturation
> > > level. I
> > > > > haven't measured that but the problem with pop-corning of plastic
> > > > > components shows that epoxy materials pick up quite large amounts
> of
> > > water
> > > > > in a few days, so my guess would be that it is a matter of da
> > > > > ys before an acrylic or a polyurethane coating has picked up as
> much
> > > water
> > > > > as a silicone coating.
> > > > > >
> > > > > > Is then a high uptake of water to the level of saturation
> dangerous?
> > > Of
> > > > > course not (under condition that the cleanliness is adequate). If
> had
> > > > > been, a coated board would not pass a humidity test.
> > > > > >
> > > > > > If the cleanliness is not adequate prior to the application of
> the
> > > > > coating, then the situation is different. Hygroscopic
> contamination
> > > will
> > > > > pick up water through the coating and cause blistering and
> > > electrochemical
> > > > > migration. I think that you agree that this will also happen if
> you
> > > apply
> > > > > Parylene to a contaminated surface. Electrochemical migration
> requires
> > > > > three things to occur; contamination, humidity and a bias. Thus,
> if
> > > > > Parylene had been an impermeable barrier against humidity, it
> would
> > > not
> > > > > had been necessary to clean the assembly. I agree that hygroscopic
> > > > > contamination will be picked up much faster if you have a silicone
> > > > > coating. But on the other hand, it is only when you have a bias
> that
> > > > > electrochemical migration will occur. If heat is generated when
> the
> > > > > equipment is used, the assembly will dry up much faster if you
> have a
> > > > > coating with high permeability and low moisture uptake. Therefore,
> a
> > > > > silicone coating may be preferable under such conditions (although
> it
> > > > > should
> > > > > verified). If heat is not generated, silicone is perhaps not the
> best
> > > > > choice (but again it should be verified). In any case, reliability
> > > problem
> > > > > due to contamination needs to be solved by improving the
> cleanliness,
> > > and
> > > > > not by finding a coating that will delay failures somewhat.
> > > > > >
> > > > > > What is then the function of a conformal coating? It is to
> eliminate
> > > the
> > > > > third factor causing electrochemical migration (and blistering)
> from
> > > the
> > > > > surface beneath the coating by preserving a cleanliness that have
> been
> > > > > achieved by cleaning prior to application of the conformal
> coating.
> > > Thus
> > > > > it needs to be an impermeable barrier, but towards ionic and
> > > hygroscopic
> > > > > contamination from the field environment and not towards humidity.
> It
> > > may
> > > > > also need to protect the circuitry from being in direct contact
> with a
> > > > > condensed water film on the assembly. (Thus, air is disqualified
> as
> > > > > replacement for a conformal coating.) I have not seen any data of
> the
> > > > > permeability of various contaminants in conformal coatings but I
> guess
> > > > > that most coating have very low permeability of ionic
> contaminants.
> > > > > Organic contaminants may be a problem since they may be absorbed.
> I
> > > also
> > > > > have to admit that if you have corrosive gaseous pollutions such
> as
> > > > > hydrogen sulfide, silicone coating is a bad choice since it h
> > > > > as a high permeability for most gaseous pollutions.
> > > > > >
> > > > > > Then, why does Parylene generally afford a much better
> protection
> > > than
> > > > > other coatings? It is because it affords better protection against
> > > > > contamination from the field environment, not because better
> barrier
> > > > > properties but because of better sealing properties due to better
> > > > > coverage. With Parylene, you will get a coating with even
> thickness on
> > > all
> > > > > surfaces, even on the tip of a needle. It is very difficult to get
> > > good
> > > > > coverage with other types of coatings since the coverage of
> protruding
> > > > > parts such as solder joints and component leads due to
> gravitation,
> > > > > surface tension and capillary forces.
> > > > > >
> > > > > > Per-Erik Tegehall
> > > > > > IVF
> > > > > >
> > > > > > -----Ursprungligt meddelande-----
> > > > > > Från: Brian Ellis [mailto:[log in to unmask]]
> > > > > > Skickat: den 6 september 2003 09:48
> > > > > > Till: TechNet E-Mail Forum.; Tegehall Per-Erik
> > > > > > Ämne: Re: [TN] Silicone conformal coating
> > > > > >
> > > > > >
> > > > > > Per-Erik
> > > > > >
> > > > > > An interesting idea. Briefly, you are proposing a more permeable
> > > coating
> > > > > > so that humidity can leave, as well as enter. I suggest that if
> > > these
> > > > > > conditions apply, then the best coating would be the most
> permeable.
> > > > > > Probably 90% or more of electronic assemblies use such a
> coating. It
> > > is
> > > > > > called air. It is extremely successful for the conditions under
> > > which
> > > > > > such circuits are normally used. Conformal coating is usually
> used
> > > when
> > > > > > air-coating is unsuitable, for whatever reason. Therefore, we
> need
> > > > > > something as far removed from the conditions of air as possible,
> > > such as
> > > > > > a theoretically impervious skin. The nearest we have to this in
> our
> > > > > > armoury of weapons of mass construction is para-xylylene
> > > > > > (Paralene/Parylene). This is near-ideal, but expensive and
> difficult
> > > to
> > > > > > apply, so we usually look to other materials. I would suggest
> that
> > > > > > acrylics are closest to this ideal, except that they have poor
> > > > > > temperature and chemical resistances. So the choice, IMHO is
> either
> > > air,
> > > > > > with maximum permeability, or a product with minimal
> permeability.
> > > > > > Compromises between the two are probably second-best.
> > > > > >
> > > > > > Brian
> > > > > >
> > > > > > Tegehall Per-Erik wrote:
> > > > > >
> > > > > >>After having thought a little more on this, I will expound my
> > > answer. It
> > > > > >>seems that all assume that high permeability is a bad thing but
> in
> > > my
> > > > > >>opinion it is not that simple and a high permeability may even
> be
> > > > > >>preferable. If you have adequate cleanliness on the assembly
> prior
> > > to
> > > > > >>applying the coating, the moisture penetrating the coating will
> not
> > > be a
> > > > > >>problem and the permeability will then not be an issue. It is
> when
> > > you
> > > > > >>have to much hygroscopic and ionic contamination beneath the
> coating
> > > > > >>that you may get problem with delamination and electrochemical
> > > > > >>migration. Even if you use Parylene, quite a lot of moisture
> will be
> > > > > >>picked up by hygroscopic contamination beneath it in a few days
> in a
> > > > > >>humid environment. If you then turn the euipment on and the
> board is
> > > > > >>heated due to power dissipation, it will take much longer time
> to
> > > dry up
> > > > > >>the assembly if you have a coating with low water permeability.
> And
> > > it
> > > > > >>is during this time period electrochemical migration will occur.
> > > > > >>Therefore, a silicone coating may actually be better than
> Parylene
> > > in
> > > > > >>this case.
> > > > > >>
> > > > > >>Per-Erik Tegehall
> > > > > >>IVF
> > > > > >>
> > > > > >> -----Ursprungligt meddelande-----
> > > > > >> Från: Tegehall Per-Erik
> > > > > >> Skickat: den 5 september 2003 08:38
> > > > > >> Till: [log in to unmask]
> > > > > >> Ämne: Re: [TN] Silicone conformal coating
> > > > > >>
> > > > > >> Steve,
> > > > > >>
> > > > > >> It depends on what you mean with good moisture resistance.
> All
> > > > > >> conformal coatings absorbs moisture but the diffusion rates
> of
> > > water
> > > > > >> in the coatings and the saturation level varies. Water
> molecules
> > > > > >> have a high diffusion rate in silicone coatings but the
> > > saturation
> > > > > >> level is lower than in most other coatings. Which is to be
> > > prefered?
> > > > > >> It probably depend on the application.
> > > > > >>
> > > > > >> However, a general statement of good moisture resistance, I
> > > > > >> interpret as that the coating properties are not degraded by
> > > > > moisture.
> > > > > >>
> > > > > >> Per-Erik Tegehall
> > > > > >> IVF
> > > > > >>
> > > > > >> -----Ursprungligt meddelande-----
> > > > > >> Från: Steve Gregory [mailto:[log in to unmask]]
> > > > > >> Skickat: den 4 september 2003 21:35
> > > > > >> Till: [log in to unmask]
> > > > > >> Ämne: [TN] Silicone conformal coating
> > > > > >>
> > > > > >> Hi All!
> > > > > >>
> > > > > >> I know when you look up properties for Silicone
> conformal
> > > > > >> coating, it usually states that it has good moisture
> > > resistance,
> > > > > >> but is it really that good?
> > > > > >>
> > > > > >> I've heard that it is somewhat permeable to moisture,
> that
> > > if
> > > > > >> you were worried about moisture, and the assembly were
> going
> > > to
> > > > > >> see the outside environment, you would be better off
> with an
> > > > > >> acrylic, or urethane coating.
> > > > > >>
> > > > > >> Any thoughts?
> > > > > >>
> > > > > >> Thanks!
> > > > > >>
> > > > > >> -Steve Gregory-
> > > > > ---------------------------------------------------
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > > ---------------------------------------------------
> > > > > Technet Mail List provided as a service by IPC using LISTSERV 1.8e
> > > > > To unsubscribe, send a message to [log in to unmask] with following
> text
> > > in
> > > > > the BODY (NOT the subject field): SIGNOFF Technet
> > > > > To temporarily halt or (re-start) delivery of Technet send e-mail
> to
> > > > > [log in to unmask]: SET Technet NOMAIL or (MAIL)
> > > > > To receive ONE mailing per day of all the posts: send e-mail to
> > > > > [log in to unmask]: SET Technet Digest
> > > > > Search the archives of previous posts at:
> > > http://listserv.ipc.org/archives
> > > > > Please visit IPC web site
> > > http://www.ipc.org/contentpage.asp?Pageid=4.3.16
> > > > > for additional information, or contact Keach Sasamori at
> > > [log in to unmask] or
> > > > > 847-509-9700 ext.5315
> > > > > -----------------------------------------------------
> > > >
> > > > ---------------------------------------------------
> > > > Technet Mail List provided as a service by IPC using LISTSERV 1.8e
> > > > To unsubscribe, send a message to [log in to unmask] with following
> text
> > > in
> > > > the BODY (NOT the subject field): SIGNOFF Technet
> > > > To temporarily halt or (re-start) delivery of Technet send e-mail to
> > > [log in to unmask]: SET Technet NOMAIL or (MAIL)
> > > > To receive ONE mailing per day of all the posts: send e-mail to
> > > [log in to unmask]: SET Technet Digest
> > > > Search the archives of previous posts at:
> > > http://listserv.ipc.org/archives
> > > > Please visit IPC web site
> > > http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional
> > > information, or contact Keach Sasamori at [log in to unmask] or
> 847-509-9700
> > > ext.5315
> > > > -----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------
|
