[RFI] DSC Alarm Panel 75 Meters/False Alarms

[Jim Brown]

On Wed, 17 Nov 2004 11:27:07 -0500, Pete Smith wrote:

>Wouldn't the bypass capacitor to ground take care of pin 1 problems by 
>making a good chassis connection at RF?  

A bypass cap on the signal conductor to the chassis does not help a pin 1 problem. A 
bypass cap from the shield to the chassis may help, hurt, or not be enough. It would help 
by lowering the Z of the path to the chassis, but the circuit board is still seeing the IZ 
drop across the cap (which may have some inductance of its own). It could hurt if the 
cap forms a parallel resonance with the L of the path via the circuit board. The best fix is 
to connect the shield (or signal conductor) properly to the chassis/enclosure. 

>I was just using ZIP cord for the speaker leads. 

Replacing ZIP cord (even expensive ZIP cord) with twisted pair will kill a lot of RF 
problems on loudspeaker lines. 

>Again, if this is low-speed data or DC, mightn't bypassing 
>both sides to a single chassis ground help?

Yes, with the limitations noted above.  Certainly a suitable bypass helps if the cause of 
the problem is insufficient filtering (bypassing). It all depends on what mechanisms are 
at play. And there can be more than one mechanism -- you stomp down on one of them 
and the other is still there. :)  The old "two-problem" problem. 

For a tutorial discussion of this (which I had a lot to do with writing) see the informative 
annex to the draft AES48, currently out for comment. http://www.aes.org, go to 
Standards for Comment. It's a free download. While the illustrations are for balanced 
wiring, the physics (and the draft Standard) apply equally to unbalanced wiring. 


Jim Brown  K9YC