PSN-L Email List Message

Subject: Re: it happened again - RFI defense
From: Ed Thelen ethelen@........
Date: Tue, 25 Jul 2000 20:46:35 -0700

John Tacinelli wrote:
> Well it jumped again tonight, right on schedule.  Fortunately I was right there listening to KOLM radio, 1570 AM.  It jumped 
exactly as they went off the air.  It must be their carrier signal during the day that keeps it in one place and then when it g
oes off the signal changes.  Unfortunately, my homemade aluminum foil box shield did not work.  At least not enough.  The folks
 at KOLM say they are going to get their 24 hour license soon so perhaps I will wait till then and depend on their steady signa
l to keep it centered.  I would like to shield it however.  I could add more tinfoil but it seems unlikely to help much.  Perha
ps a wire cage of some sort?
> John Tacinelli

Good - sounds like you have isolated the effect to 
a one to one with the radio station.  

Battle plan - basically your amplifier (yes amplifier) should
be shielded from the effects of the cruel world.

(In theory, you can make a research project to find what exact
 component/configuration is giving the trouble.  But life is
 short, and you may wish other adventures  :-)

(A hint though, if you are using diodes as non-linear element
 to gain some say logarithmic or other function - I would suspect
 RF coupling into them first.)

The following is brute force, kill the fly with a hammer, approach
to shielding.  (Shielding should be a science, but it is also
a black art, complete with magic incantations (I think :-)

0) check all solder joints and connections between your sensor coil
   and amplifier.  make sure they are GOOD - no oxides, no bad solder
   joints, ...  (the darndest things can cause rectification -
   and Murphy usually wins.)

   Also, if you can, use twisted twisted wire between the sensor
   and amplifier - shielding does not hurt either.

   You might wish to put a SMALL capacitor directly across the outputs
   of the sensing coil - like 0.001 micro farad - and hope that you
   don't pick a value that resonates with the AM band - experiment
   later - actually, with the very thin (high resistance) wire
   most of us use, resonance should be minimal - like Q < 1

1) get a shielded (say aluminum) box (say from Radio Shack)
   large enough to contain your amplifier.  
   Yes, I know they cost too much, but they cost too much
   at every retail outlet - just pay up.

   General practice is to ground the amplifier to the chassis,
   and ground the chassis to earth (the Brits use the correct word ;-)
   (One can be picky and try to use single point grounding,
    to avoid "ground loops" which might not bother much here,
    but something to think about.)

2) You probably have a low pass RC filter with a cutoff
   at something like 10 to 20 hertz at the input to your amplifier
   - if not, ya jus gota make one.
   Of course the common of the filter must connect to the 
   amplifier common.

3) While messing around, filter the power supply inputs
   to the amplifier common, and might just as well
   do a little filtering on the output of the amplifier
   also - cheep & easy 

4) Close the aluminum box, and put in a sheet metal or other
   conductive fastenings on each side of opening to 
   give reasonable conductivity between parts of the box.

With reasonable luck - your RFI effects should be below noise level :-))

  Ed Thelen - been there, done that  ;-))

Public Seismic Network Mailing List (PSN-L)

[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>