On Fri, 18 Jul 2003 12:43:17 GMT, "Carl R. Stevenson"
wrote:

I guess it may depend to some extent on the quality and type of
fixtures/bulbs
that you install.

Doesn't really matter much. Most noise optimized tubes are really
audio noise issues.

My shack is in a finished "walk-out" basement (the house is dug into
the slope at the top of the hill, such that from the front it appears to
be a 1-story structure, but from the back side, it's 2-story).

The entire basement area, ham shack, electronics lab, office, family room,
woodshop, etc. is VERY well illuminated by an array of fixtures set in the
grid of a suspended ceiling. Area is about 2000 sq. ft. and there are 14
4 ft., 4-tube fixtures.

I have checked and I can't see a noticable change in RX noise in the HF
bands between all lights off and all lights on.

If you get a tube that's starting to go, or a bad ballast, that could cause
a dramatic change ... something that bodes in favor of replacing bad
tubes.

(Of course, my antennas are outside, away from the lighting, and the
house is covered with aluminum siding for low maintenance, which
may help to contain any noise from the lighting ... YMMV, depending
on such shielding, antenna placement relative to the lighting, etc.)

73,
Carl - wk3c

Hi All,

My situation is very similar to Carl's, and my experience is very much
like his as well.

I went a little further to provoke problems to confirm all sources of
noise within my home.

The first was to establish a noise baseline (done at various times of
day and night and time of year, across the many bands) by using my rig
on battery power (it is actually always battery powered, but employs a
floating charger) with the charger and all accessories disconnected
and ground removed (no ground loops). I also throw the main breaker.
There are any number of hidden power loads that you cannot disable or
otherwise isolate. One of the most egregious examples is the doorbell
transformer that in older homes can turn into an S9 noise generator
that will plague many "experts."

That was Step One. There is not much point in going to Step Two if
you cannot isolate sources. Many succeed however, through weeks of
hair pulling and a magical connection (or break) that could also kill
them.

Step Two is to throw the breaker back on. If this seems a huge chore
of having you go through the house resetting clocks, let me console
you with the evidence that that end of the gene pool has been thinned
out considerably by the twice annual event of daylight savings and the
invention of the digital clock. Survivors of that Darwinian task can
now take another noise reading (again observing the list of times and
frequencies already mentioned).

Step Three (this is a progression back toward your conventional shack
design) is to start adding loads to the house circuit. Again, take
readings (you need not note times as you are now introducing
correlated noise). Remove power from all such loads.

Step Four is to add back your conventional ground (that is, the ground
through which power, not RF arrives). This introduces the possibility
of a ground loop. Repeat Step Three.

Step Five is to add back your conventional accessories. This too
introduces the possibility of a ground loop. Repeat Step Four (which
recursively repeats Step Three, of course).

All told, this should result in a grid of readings with several dozen
(maybe 100) data points. This will take far less time than trying to
solve a noise problem that persists for more than a day (and possibly
on the WRONG day).

Any one of these tests could result in the appearance of noise.
Sometimes this is correlated to the last addition of a load, but as
tantalizing as that is, it too refuses cure (and you cannot live
without that load).

Such problems then need to be isolated as to being over the air, or
through conduction.

Over the air problems are often fixed by Carl's observation of having
the antenna away from the house. More telling is the addition of a
choke BalUn at all feed points (even those remote from the house, this
is especially true for long wavelengths).

Problems of conduction are in fact ground loop based. You are sharing
the same circuit as the noisy load. I have fluorescents above my
station, and dimmers on every incandescent bulb in my house and yet I
suffer very little noise arriving by conduction (and find it perturbs
over the air by 3 or 4 dB above baseline before I go turn off that
light I rarely do that because atmospherics and remote noise often
overwhelm this local contribution).

A simple way to confirm this noise path is to break it by powering
your equipment off another circuit (circuit meaning the wire path to a
separate breaker at the distribution panel). This can make 10-60 dB
difference for dimmers alone (or the other killer, the fish tank
heater). Simply employ a sufficiently long extension cord so that you
do not change your shack configuration (which would also force another
baseline evaluation).

The single next greatest contributor of noise is through peripheral,
accessory connections. The lighted Daiwa SWR meter has a lump you
plug into the wall. Both the lump and the connection are potential
problems if you disregard them and one or the other turns out to be
the problem. It is easy to disregard such an inconsequential item, so
to are computers with modems the third most common source of ground
loop.

If you wish to preserve those accessories, buffer your power through
an isolation transformer with a faraday shield connection. Connecting
the shield is also not obvious (there are reasons to connect it to the
sink, there are reasons to connect it to the source - but never both).
My battery charger is so buffered.

Well, these are the heavy hitters of noise. There are more, but their
application becomes a matter of simple logic and practice when you
have a basis.

73's
Richard Clark, KB7QHC