Author Topic: Dielectric strength of PCB perpendicular vs parallel  (Read 1254 times)

0 Members and 1 Guest are viewing this topic.

Offline PoeTopic starter

  • Frequent Contributor
  • **
  • Posts: 250
Dielectric strength of PCB perpendicular vs parallel
« on: March 05, 2020, 05:33:02 pm »
Laminate suppliers only list one "Dielectric Strength" so I assumed that orientation didn't matter.

Then I discovered that there are two voltage breakdown tests.  Test method IPC TM-650 2.5.6.2 is used for testing perpendicular to the plane and 2.5.6 parallel to the plane. 

The 2.5.6 test appears to have a much lower passing threshold.  So much so that IPC2221B guidelines for internal clearances come into question. 

Does the claimed "Dielectric Strength" value of a PCB laminate really only apply to voltages perpendicular to the plane (2.5.6.2)? 

Is there another name for the test results of 2.5.6?

Thank you.

 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15799
  • Country: fr
Re: Dielectric strength of PCB perpendicular vs parallel
« Reply #1 on: March 05, 2020, 06:31:37 pm »
The difference between the two is related to a phenomenon usually called "surface flashover". You can look that up - not an extremely simple thing - but basically, voltage breakdown between conductors on the surface of an insulator occurs at much lower voltages/lower distances than through the material (perpendicular).

 
The following users thanked this post: Poe

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Dielectric strength of PCB perpendicular vs parallel
« Reply #2 on: March 05, 2020, 09:01:26 pm »
Related to that, if there is delamination, then the same can happen inside the board.  Especially between inner conductors (since, thru holes have outer layer annular rings closer together than the holes).

Even without delamination, I would also be curious what the lateral breakdown is.

There's no particular reason to expect it to be isotropic -- the fiberglass weave is not evenly distributed, and is oriented in plane.  There may also be gaps, voids, defects in the layup at edges of inner layer copper -- the edges are mechanical displacement, stress raisers and concentrate the electric field.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Poe

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9003
  • Country: us
  • Retired, now restoring antique test equipment
Re: Dielectric strength of PCB perpendicular vs parallel
« Reply #3 on: March 05, 2020, 09:28:04 pm »
In general, the breakdown voltage gradient (V/mm) along a discontinuity in materials (such as the boundary surface between a solid insulation and air) will be substantially less than through a homogeneous material.
 
The following users thanked this post: Poe

Offline PoeTopic starter

  • Frequent Contributor
  • **
  • Posts: 250
Re: Dielectric strength of PCB perpendicular vs parallel
« Reply #4 on: March 09, 2020, 05:59:22 pm »
Thank you for the replies, but I'm still struggling with this.

The 2.5.6 test doesn't appear to be testing surface flashover.  Rather breakdown through the laminate itself.

https://www.ipc.org/TM/2.5.6b.pdf

The test sample is immersed in an oil bath to avoid flashover and the instructions say to retest if the breakdown didn't happen though the material.

TimFox - I guess that would explain why the voltage parallel to the plane has a much lower passing value than parallel.  There are numerous continuous barriers along the glass weave from end to end.

What I find surprising is that the IPC2221B generic standard clearances for internal layers (Table 6-1) are insufficient for the IPC TM-650 - 2.5.6 passing voltage.  They appear more appropriate for the perpendicular test (aka "Dielectric Breakdown" in the laminate datasheet).
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf