Phil
I don't believe FR-4 is suitable for what you are doing.
1) If you take the prehistoric MIL-P-275 (its successors may still say
the same), the max voltage gradient was c. 160 V/mm bare and, I think,
the double conformally coated, let's say 333 V/mm for convenience. If
your oil can be likened to a liquid conformal coating, this means the
spacing for conductors at 100 kV DC should be 30 cm. Is it???
2) Epoxy resins contain ionic species. Under high voltage gradients,
these can dissociate and migrate, which would promote tracking, by
setting up a series of conductive sub-molecular islets throughout the
material. At low voltage gradients, this is not important as any islets
are encapsulated within the resin matrix.
3) TBBPA, used to render the FR-4 flame resistant, has 4 bromine-carbon
covalent bonds per molecule, which is within the resin structure itself.
The Br-C bond is notoriously weak and, even though it is non-ionic, it
cannot be guaranteed that the odd bromine may not be liberated under the
stress of a high voltage gradient (in the same way as under thermal
stress). If such Br atoms encountered any free hydrogen ions or
radicals, also possible under imperfect stoichiometry during
polymerisation, the resultant HBr would form a lovely ionic conductor.
This is hypothetical, but entirely possible.
4) Glass is an ionic liquid of super-high viscosity. Ions therefore are
little mobile, which renders it normally a good insulator. Under high
voltage gradients, it, too, can break down. However, the glass is
treated with a member of the silane group, such as
chlorotrimethylsilane, to render it wettable by the epoxy resin.
Silicon-hydrogen covalent bonds are notoriously unstable and are labile.
Who knows what would happen under the stress of an enormous voltage
gradient?
My recommendation would be to replace the FR-4 with a linear polymer,
such as polytetrafluoroethylene, without reinforcement, if this is
possible. This would be chemically resistant to the oil and electrically
resistant to tracking, even at relatively high voltage gradients. It has
disadvantages: hellish to work, mechanically unstable and not the
cheapest PCB laminate. Alternatively, some of the styrene copolymer
laminates may be OK, but they may be less chemically resistant, possibly
resulting in stress cracking over time.
Hope this helps.
Brian
Phil Nutting wrote:
>
> Hi folks,
>
> Today I have a question about high voltage tracking. The actual set of materials and applications are as follows.
>
> FR4 material, 0.090" thick, two layer, plated through hole, LPI solder mask, submerged in Shell Diala AX dielectric oil, operating at 100kV. We are seeing hv tracking marks that run with the grain (fiber directions) of the FR4 material. This burn marking is in the LPI, but also in the FR4 material as can be seen in the photo where the LPI has been scrapped off. The measling is from removing the copper from the board.
>
> Here is the photo http://www.stevezeva.homestead.com/files/HVTracking.jpg. It is the top left hand photo entitled High Voltage Tracking. Thanks, Steve Gregory for posting this on your page www.stevezeva.homstead.com.
>
> We have seen this tracking on other boards which do not have the LPI solder mask, they are bare boards, so that eliminates the LPI as the cause.
>
> One of our engineers had seen this in a former life and it was attributed to the small air bubbles in the substrate where the fibers were.
>
> Is there a different type of material that will not exhibit this tracking?
>
> Thanks, 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 e-mail to [log in to unmask]: SET Technet NOMAIL
> 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/html/forum.htm for additional
> information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
> ---------------------------------------------------------------------------------
---------------------------------------------------------------------------------
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 e-mail to [log in to unmask]: SET Technet NOMAIL
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/html/forum.htm for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
