Hi - I'm looking for some practical advise on using an inner Printed Circuit 
Board "PCB"  plane tied at the corner mounting holes of the chassis as a shield 
from RF exposure coming in on an isolated field circuit relative to another 
digital logic board.

Questions:
  a.. Must my shield layer be the bottom most inner plane to be effective? Note 
that the 2 board solder sides face each other.  Layout wise it is very 
difficult to do without going to a 6 layer board - I am trying to stay with 4 
if possible.
  b.. Would I get some capacitive bypass protection if the grounded shield 
plane where the top inner layer of the field circuit and the isolated field 
circuit common was the bottom inner plane exposed to the adjacent board?  I 
know the field would not be blocked but the capacitive bypass to the chassis 
should reduce the RF intensity.
  c.. Would the placing of the grounded plane between the top PCB components 
and the Isolated common plane disturb the performance of the field circuitry?  
What would this do to the field circuit return path loop area?
  d.. Would a ground pour on the solder side of the adjacent PCB hurt or help?  
It would increase the capacitance between the two boards but it would reduce 
the loop area of the traces on the solder side of the adjacent board too.
  e.. PCB clearance issues within a PCB - What are the clearance/voltage rating 
issues within a PCB?  What voltages can I support from:
    a.. Surface trace to inner grounded plane - Can 240VAC be supported.
    b.. Electrically hot via passing through the grounded plane - What inner 
pad clearance would one use on a 240VAC circuit?
  f.. What do the safety standards say? - I see that UL3111 version of IEC 
61010-1 can treat the PCB as a molded void free material and so the clearance 
issues are not addressed.  I understand that IPC has a spec on voltage ratings 
versus construction - I am looking for it now.

Details:
The circuit worked fine at 10V/M but I am now being asked to take this to 35V/M 
and that is somewhat of a challenge.  
Testing to ANSI C37.90.2 1995 25-1000 MHz.  
My problem area is 400-500 Mhz. 
I am trying to keep RF energy on field circuit from coupling over to an 
adjacent board and corrupting the bus of some microprocessor based logic. 
The two boards are only .200" apart.  
I have experimented with an insulated grounded shield plane placed in-between 
the two boards and it works great.
I can withstand 50V/M WITH 80% modulation!!!
Unfortunately the shield is very difficult to make for production and the cost 
is an issue.

I am trying to put the shield into the 4 layer PCB in the area around the field 
circuit.
The solder sides of the two boards face each other.
The field circuitry has lots of through hole PCB type components and so there 
are leads and trace pads that are exposed on the bottom of the field PCB.
I have trimmed the leads and this helps some.
Even with a 6 layer board and the bottom inner layer as a grounded shield I 
would have the leads sticking through the shield like holes in cheese.
The field circuit construction is:
  a.. top +5V pour plus a couple traces 
  b.. top inner layer presently free - this is where I would like to put the 
grounded shield plane.
  c.. second inner layer (solder side) is the isolated field circuit common,
  d.. Bottom solder side are a number of field circuit traces.
Unfortunately the free plane is not the bottom inner plane but the second from 
the top component side.
The two boards have DC:DC and opto signal isolation on the field board relative 
to the adjacent logic board.
The two boards and are connected at one edge of the board with a pin header.
The field circuitry in question must support at least a 2000 VAC/1minute High 
Pot to ground. 

Hope all this detail is not too overwhelming - I tried to put it all last for 
those who don't want to know.......
Any information you could share would be appreciated!!
Thank you in advance.

Chris Wells
Senior Design Engineer
Cutler-Hammer
Pittsburgh Pennsylvannia - USA
[email protected]






Reply via email to