Okay, Doug,

Defining a void is relatively easy: it is an area which has not been 
wetted by the conformal coating. A bubble is less easy to define because 
there are different kinds of bubbles. In the context, the general 
definition is an area where there is no adhesion to the substrate and 
where the conformal coating forms a skin, proud of the substrate 
surface, and generally containing a gas, which may be air or solvent 
vapour. Depending on the cause, there may be or may not be a thin layer 
of conformal coating on the substrate under the bubble.

Bubbles may be formed by a number of different mechanisms, including:
– uneven spraying of the coating material with unsuitable air pressure
– entrapment of air bubbles in immersion coating
– unsuitable conditions for curing solvent-based coatings, where the 
surface dries too rapidly, leaving solvent trapped underneath which 
subsequently evaporates
– moisture absorbed through the coating that has been applied over 
hygroscopic contaminants (vesication or mealing).
Ideally, it is better to eliminate the cause, rather than to have to 
judge whether the bubbles are acceptable for a given application or not.

With regard to the numbered sentences:
1. If the area under a bubble contains only solder mask or other 
non-reactive surface, it is true that there may be no reduction of 
reliability of the assembly. Nevertheless, I feel that there is the 
possibility of defective application of the conformal coating and the 
cause should be determined before accepting or otherwise the defective 
assembly (I assume that we are talking about high reliability assemblies).

2. I did use the term "two bare conductors"! Whether there is a thin 
film of coating on the conductors or not depends upon the cause of the 
defect. The thickness of the thin film that may be acceptable is 
arbitrary and could depend, for example, on the voltage difference 
between the conductors.

3. I am not aware of any published information. My arbitrary 30% is 
based (from memory) on the acceptability of inclusions in solder masks, 
on the assumption of an analogy. My >10 V is also arbitrary, based on 
gut feeling!

4. Why two bubbles per square decimetre? I would consider that more than 
two bubbles would be indicative of bad workmanship, rather than some 
form of accident, no matter the size of them. As for your second 
question, I think you will find the answer in the other thread which you 
subsequently started.

5. The point I was trying to make here is that the conformal coating 
should be conformal and not under-filling. Depending on the size of the 
component, any under-filling would almost certainly entrap air. In any 
case, if the space under a component is under-filled there may be 
problems due to the coating material having a high volume coefficient of 
temperature, which would risk a drop in reliability.

6. At last!

This leads to a long-felt niggle in industry generally, and in our 
industry in particular. We seem to like every letter i to be dotted and 
every t to have its stroke by having specifications written down in 
great detail rather than having common sense. This may patronise the 
members of our profession which now rely upon standards and 
specifications more than allowing them to come to a conclusion by using 
their eyes and brains. Anyone with more than half a monkey's brains 
could probably judge whether a bubble compromises the quality of an 
assembly or not with a simple glance, instead of which that person may 
prefer to wade through reams of paper in order to find out whether he 
should accept it or not as a defect. This same niggle applies, of 
course, to many other problems that we may or may not have in our industry.

I think that this rant is worthy of a second two eurocents!

I have found the answer to the bailout crisis. I have converted all my 
money into gold bars and buried them deeply in my vegetable patch which 
does happen to require being dug over! :-)

Brian

On 26.03.2013 17:37, Douglas Pauls wrote:
> Brian, thank you for the response.  See my questions in red below.
>
> Doug Pauls
>
>
>
> From:   Brian Ellis <[log in to unmask]>
> To:     <[log in to unmask]>
> Date:   03/23/2013 02:24 AM
> Subject:        Re: [TN] Conformal Coat and Bubbles
> Sent by:        TechNet <[log in to unmask]>
>
>
>
> IMHO, it depends what you mean by bubbles/voids. If they are a result of
> humidity being absorbed by hygroscopic contaminants through the coating
> (vesication or mealing) then the answer is a resounding 'not acceptable'
> under all conditions.
>
> **OK, so Brian, if you had to provide a definition of a bubble or void,
> what would that be?
>
> If they are the result of poor application of a liquid conformal coating
> or solder mask, I would say that they would not be acceptable if any of:
> 1. the substrate is not wetted under any bubble
> **What if the area under the substrate contained only solder mask or other
> unreactive surface?
>
> 2. a bubble bridges two bare conductors
> **I am assuming that the assumption is that the bubble exposes the metal
> of the two conductors.  What if there is still some coating on the
> substrate under the bubble?  Would this still be a defect if I could show
> at least 1 mil of coating still covered the conductors?
>
> 3. a bubble occupies more than 30% of the space between two bare
> conductors with a potential difference of >10 V
> **OK. Is there a study that you can point to upon which such values are
> based?
>
>
> 4. there are more than 2 bubbles per square decimetre on either side,
> not counting the edges of a board where there are no conductors, up to 5
> mm from the edge
> **Oh hell, now I have to go look up decimetre.  Hang on......  Why 2
> bubbles in that space?  Would it not depend on the size of the bubble?
> Second question - do you feel that the sides of an assembly need to be
> coated?  Would it depend on whether the edges of the board were
> routed/sealed vs. punched/unsealed?
>
> 5. a void or bubble can be seen by 10 X visual inspection under a
> component
> Again, would it not depend on whether or not the void or bubble left
> anything open?  I can see operators trying to underfill components with
> conformal coating to get rid of things they perceive as voids under
> components, which would be a big reliability risk.
>
> 6. a paraxylylene coating displays any void
> Here we probably agree.
>
> Some of the numbers in 3., 4. and 5. are arbitrary; common sense should
> be applied, along with the coating.
> Ahhh, but common sense is not all that common.
>
> My 2 eurocents.
> Which are always appreciated.  Glad your banks got their bailout loans.
>
> Brian
>
> On 21.03.2013 17:18, Douglas Pauls wrote:
>> OK, Minions, your next question in the quest to improve J-STD-001 and
>> A-610 relates to bubbles in the conformal coating.
>>
>> I think we can all agree that the ideal conformal coating layer contains
>> no bubbles or voids and is "purdy".    BUT:
>>
>> 1.  Are there bubbles in coatings in areas where they will not impact
>> reliability and their presence should be viewed as a "Process
> Indicator"?
>> If so, where and what limits would you use?
>>
>> 2.  What would you classify as a bubble requiring disposition, i.e.
>> Becomes an actionable defect?  If so, where and why?
>>
>> 3.  Would any of your answers change depending on what KIND of coating
> it
>> is, i.e. Acrylic vs. silicone vs. Parylene?
>>
>> Well, OK, that is three questions.  I'm on a roll.
>>
>>
>> Doug Pauls
>> Chairman, IPC Cleaning and Coating Committees
>> Galactic Emporer
>>
>>
>>
>> ______________________________________________________________________
>> 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] 
______________________________________________________________________