[RFI] How much RF does your power company deliver along with the AC?
Christopher E. Brown
cbrown at woods.net
Fri Aug 21 19:44:29 PDT 2009
On Fri, 21 Aug 2009, Jim Brown wrote:
> On Fri, 21 Aug 2009 13:15:20 -0500 (CDT), Christopher E. Brown wrote:
>> With power back, even with everything but a couple lights (filament, not
>> fl), strong RF on the AC lines, and s5 - s7 on the dipole.
>> Starting to think the best option would be a good filter between the
>> disconnect and sub-panel
> Your proposed solution is based on the assumption (almost certainly
> false) that the noise you hear is coupled into your home on power lines.
> The FAR more likely conclusion to be be drawn from your observations is
> that when the power in your neighborhood is off, the many noise sources
> in your neighborhood (and in your own home) cannot produce noise. Much of
> that noise is RADIATED by power lines (both inside and outside of homes
> where that noise is produced), or by other wiring connected to those
> noise sources, and picked up by your antenna. And, of course, faulty
> insulators on power lines may produce noise, which the power lines also
> Unfortunately, there is no magic bullet you can apply at your QTH. That
> noise still must be eliminated at the source. The most you can do is use
> antenna directivity (and common mode chokes) to minimize what you hear of
> Jim Brown K9YC
Perhaps I did not state things as clearly as they could have been, but you
may want to re-read my original message.
2 Antennas about 40 feet apart, the dipole about 10 feet from the closest
AC wiring, and sitting above all of it, the L about 30 feet from the
closest AC wiring and to the side.
Distance from the nearest pair of neighbors is similar for both antennas,
dipole is a litle closer but nearest neighbor pair is off the ends of the
Differance in distance from antennas to all other neighbors rapidly
decreases from there.
Both antennas have fairly hi values of common mode choking action on 20M,
to reduce noise coupled over the outside of the coax.
Before the power outage I was not sure of the relative performance of the
antennas, the dipole was too noisy to use.
During the outage, all potential noise sources within 1000 feet and most
with 2000 feet were unpowered, and the entire local grid and all small
(2-4 home final stepdown) transformers were totally disconnected from rest
of city power system.
During the event, I did a general survey on the L, my normal use antenna
and swept the entire 1 - 30 Mhz range looking for any bothersome signals,
so that I could log them as not coming from my immidiate area. There were
a few, mainly what sounded like PWN chargers or motor controllers. A few
of these I was thinking were nearby RV chargers, but it is clear now I
need to look along the business strip a bit further out for these.
During the outage, I set the 756pro to both pre-amps on, no atten with L
on antenna 1, dipole on antenna 2.
I them swept through the whole band listening and watching the scope while
switching back and forth between antennas. Since this is a simple button
push it made them easy to compare.
Both antennas showed nearly the same background noise levels across the
band, the diff between them was less than the variation over time on the
same antenna. Except for a couple narrow-band sources everything was < s1
across the whole band. Most of the findable spikes were of similar
strength on both antennas, with a few being stronger on the L.
I then switched to SSB and went hunting for stations, I can often hear
europe over the pole around then.
With the very low background I was able to hear a number of stations I
would normally not, I listened in on at least 30 QSOs. Overall both
antennas did about the same with the dipole favoring the central to
eastern US and the L doing much better with europe over the pole, but if I
had a clean signal one one the other could at least copy it on the other.
This was about as expected given the direction of the dipole.
So far my assumtions are
All nearby noise sources our offline
No noise on the 0V AC wiring
Both antennas have the same background noise and same spikes as almost the
same level across the band.
Both antennas seem to have about the same receive performance (on 20M on
40 and 80 the L is much better).
Power returns, everything at my house is still disconnected from AC
Now the dipole is running 25 - 30 db higher across the entire band. I
find a few spikes strong enough to be seen above the noise floor on the
dipole, one is a TV clocking signal on 14.318 from about 4 houses away
(known source). Signal is nearly same strength on both antennas (I
*know* this is a radiated signal, common mode on the 300 ohm twinlead
running the a rooftop antenna. I will be helping the older couple
replace the old twinlead antenna with quad6 from a new antenna to a
DTV converter befor winter.).
I don't trust the exact numbers here, it is just a bandscope reading, but
the scope says 10db/div and has 10/20/30 db atten settings. The L shows a
nice rough pattern in the 0-5db range, the dipole across the whole band
just under 3rd div w/ the little random peaks just over. 20db scope atten
pots the pattern on the first div, and 30 leaves only the small peaks.
The difference in distance to nearby homes between the antennas is not
enough to matter except for the closest, and the closest are directly off
the ends of the dipole and the diff in distance is still < 10%.
Where exactly do you get that 2 antennas that display similar performance
would show a 25 - 30 db difference *only* on local noise??
I can repeat the general process by bringing a small pickup loop within 12
inches of any AC run in the house, or a small directional loop withing
about 2 feet.
I suppose this could be the wiring acting as an antenna and coupling to
the dipole, but conducted common mode seems alot simpler then induced...
I am open to being convinced I am wrong here, but someone will need to
explain where the flaw in my testing/logic is.
To be clear here, what I suspect is conducted common mode over the AC
lines is the wideband hash that pushes the whole floor up. I assume that
the large majority of specific signals are radiated, not conducted as I
get them on both antennas.
This is not a quick fix, or a substitute for noise tracking. I have
put a great deal of effort over the last year finding and eliminating
close radiated sources. This time last year I could not copy anything
much below S9 an any band 20m and down, as my noise floor was s7-s9
minimum, on the external antenna (then a 30ft vertical, later changed to
an L). I just received another box of 2.4in #31 toroids to help in this
The reason I want to "fix" this... I do not have space/location for more
antennas away from the house. I know my radiated interfance levels on
80/40/20 are reasonable/workable these days, it is only when an antenna is
in proximity the the AC wiring that the floor jumps 20db or more.
More information about the RFI