Phil,
You raise an interesting point. The real issue has to do with voids between
high voltage points within the structure. Interesting things happen at high
voltage and one is partial discharge arcing. Spacing, void size and voltage
between points where a void resides are all critical but if there is
partial discharge present internally then you can run into reliability
problems. It gets interesting when one starts this testing because there
are two distinct PD mechanisms present, Townsend discharges characterized
by fast duration (<10nS) and low charge transfer (<100pC) at corona
inception voltage (CIV). Townsend type discharges are normal and usually do
not lead to long term problems. The other and what you need to avoid is
streamer type discharges characterized by longer duration (>100nS) and high
charge transfer (>120pC) at CIV. These types of discharges are detrimental
to long term reliability and must be avoided. This requires PD test
equipment typically made by Biddle and a few others and using these tools
will allow you to have an objective acceptance criteria. Clearly you will
need to characterize what is good and bad and then perform a lot of cross
sectioning to verify your results.

I enjoyed working in this arena and found it fascinating finding and
correlating these types of defects in the past. Have fun.

John Maxwell


>Good morning everyone.
>
>Today we have a situation where we have our "Lore and Legend" of
>business (or "that's the way we've always done it") versus what a vendor
>will supply.  Now let my unscramble this for you.  We make systems that
>use high voltage coils encapsulated in epoxy.  Our practice for many
>years is to not accept any voids (that we could see on the surface) in
>the finished encapsulated item.  Generally we can achieve this by
>pouring the epoxy into the mold while it is at roughing pressure or
>under low vacuum.  If there was a void we never accepted the unit even
>if it was patched because of a concern about punch through because of a
>poor bond, dirt or other "conductive" gunk possible trapped under the
>patch.  We won't go into the waste generated or process control here,
>that's a separate discussion.
>
>We now have a unit that is being potted by an outside vendor.  One unit
>came in with a visibly patched void.  We rejected it because of our
>undocumented culture of "no voids".  The outside vendor says they will
>not guarantee void free potting.  So now we can either eat the cost of
>this unit or use it.
>
>What is a generally accepted practice for encapsulation or potting
>regarding visible voids and patching?
>Do you flat out reject it or do you consider the application and the
>location of the void/patch?
>
>I'd appreciate your views as we will be meeting on this subject and I'd
>like to have some external opinions as reference when discussing the
>merits of accepting repaired voids.
>
>Thanks in advance,
>
>Phil Nutting

---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
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 delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------