|
Index |
atmosPHeric NOISE at HF
amateur bands (Accepted by 'Electron' magazine) Introduction In a related article on
'Reporting', also to be found here,
it has been suggested to always add the noise background in the signal
report. This background noise signal may take many shapes depending on
frequency, the hour of the day, type of antenna etc. Also the type of
radio-'traffic' will be an important issue as may be understood when
comparing intelligibility of a signal in a pile-up to the same signal against
'atmospheric noise' in a quiet part on the ten meter radio-amateur band.
Furthermore one may experience a notable difference when listening to a weak
signal against 'natural background' as compared to 'side-band spatter' having
the same 'rhythm' of human speech. In the latter situation, the 'in-ear
filtering' is much less effective. In all situations, background
noise is an important factor that is calling for a more detailed analysis to
explore type and energies one may expect at HF amateur frequencies.
Noise with a natural
background Natural noise is consisting of a
limited number of components; here we shall deal with the most important
only. This component consists of
lightening discharges between the clouds and between the clouds and the earth
surface. Especially the latter is generating more energetic discharges and
hence more 'noise'. Lightening discharges may be found by the millions each
day all over the globe. These numbers translate to hundreds per second of
which only the nearest will be perceived as 'impulse' noise. The energy of lightening
discharges at a larger distance is traveling towards us along a number of
'roads' and is therefore less energetic but will last longer in time. This
phenomenon is called 'dispersion' and is fundamental to all wave-spreading,
including light and water. Lightening discharge energy at larger distance
therefore will be perceived as a constant hiss, rather than an impulse type
of noise. Impulse noise as well as
background hiss will vary with the weather, the season and the position at
the globe. Usually lightening storms are more frequent and more energetic in
the South-Eastern part of the world, and so communication in the lower
HF-bands is more of a change matter is this area. This is showing natural
noise to also exhibit a strong statistical spread in time. In general one
will measure some 90 dB difference between noise values that are exceeded
during 0,5 % of the time compared to those that exceed 99,5 % of time. There
also is a strong correlation with the time of day. Total background noise
will be quiet different at
This type of natural background
noise will exhibit a wide power spectrum over frequency, dropping off at
higher frequencies. Starting around 4 MHz., general galactic noise will take
over as the dominant factor up to around 1 GHz.. Around this frequency the
noise by the sun is becoming more important and will be predominant at
focused antenna systems looking (accidentally) directly into this direction. Although natural atmospheric noise
is important below 4 MHz., it is not the most important factor between 0,1
MHz. and 4 MHz. Except for heavy local thunderstorms, most important
contribution to background noise will be provided by human interference. In a
rural location far from the big city and industrial zones, this man-made
noise at e.g. 1 MHz. may readily be 50 dB stronger than the natural
background. This level is easily surpassed by another 20 - 30 dB in city
centers and industrial area's.
This man-made 'electro-smog' is
consisting of many components like ageing house hold appliances, sparking
electric switches, train and light-rail based interference and in the older
days also by automotive ignition noise; the latter nowadays has been replaced
by all sorts of PC-based noise, industrial and medical appliances etc. Together with the 'sparking
source' one should also consider the aerial that is radiating the energy.
Most electrical appliances are connected to a wall socket or to each-other,
with these (inter-)connections acting like the electromagnetic radiator. This
is another reason why this man-made electro-smog is prominent in the 0,1 - 4
MHz. frequency range. The combined effect of radiator length and the
'artificial noise source' make for an efficient radiating system at this range.
The noise source antenna's are running in all sorts of directions reason
why electro-smog usually does not exhibit
polarization properties. This is less so for very local noise sources.
Especially for this type of noise it can be very effective to pick op the
signal with a relatively small antenna and apply this unwanted signal in a
noise-bridge to improve DX reception in the main antenna. Many circuits to
this extent may be found on the internet and also MFJ is delivering devices
like MFJ 202B and MFJ1025 noise canceling bridges. This type of noise
cancellation is requiring some skills since amplitude and phase have to be
tuned simultaneously to show effect. In general this type of equipment is
capable of repressing local noise by 30 dB or more (5 S-points).
It is often suggested this type of
local noise to exhibit more important electric than magnetic components and
more often be vertically than horizontally polarized. If so, these properties
may also be put to good use when trying to compensate. These types of local
effects may also be at the base of 'quiet antenna' behavior that is
attributed to one or other antenna. Although usually the particular antenna
is said to exhibit this particular behavior, it more likely is the local type
of the electro-smog that is at the base of the observation; for local noise
sources the antenna is always operating in the very near field. Identical to natural background
noise, the man-made noise is also subject to dispersion, reason why more
distance noise sources also will be perceived as a general and
uncharacteristic background. Again ionospheric conditions are providing a
different noise background in the evening and/or periods of turbulent solar
activity. when this back-ground noise may be received from a larger area. Radio noise levels have been
observed for some time. Quite early in radio-history noise levels have been
measured at various times and places around the globe. These measurements
have been collected in a report by CCIR, Comité consultatif
international pour la radio, that later has evolved into ITU-R,
International Telecommunication Union - Radio Communication sector. The
report is periodically updated and is currently available as ITU
Recommendation P372-8. The report consists amongst many others issues of a
summary on these continuous measurements at different places and times,
normalized into a set of generally accepted graphs. Not long ago the noise data in the
report have been revisited in order to have a set of agreed noise figures as
a reference for comparing expected future levels of noise as generated by
PLC. The ITU noise measurements proved general noise figures not to have
changed very much over the last 20 years. PLC in this respect meaning Power
Line Communication, the commercial plans by Electricity companies to exploit
their electrical power lines for wide-band data communication. As these lines
are optimized only for frequencies of tens to hundreds of cycles per second,
it is clear these are unfit for efficient data transport of much higher
frequencies.
Noise graphs over frequency for
our HF-amateur bands as derived from this ITU report may be found at
internet, 'Googling'
with key-words: 'Atmospheric and man-made noise'. As an example one may find
the graph of figure 1.
These noise graphs are expressed
in the noise figure Fa, representing the excess
noise relative to a resistor at a certain temperature and bandwidth. When
calculating for a resistor of 50 Ohm, a temperature of 300 K and a bandwidth
of 3 kHz. (customary in amateur EZB communication), we are able to translate
these graphs directly into a noise voltage at the antenna input of our
receiver, provided the impedance is 50 Ohm at this frequency. The antenna
gain is to be taken as 0 dBi. In a quiet rural environment one
may calculate from this graph at 3,6 MHz. a total noise voltage of 1,82 μV. The voltage will drop to 0,48 μV at 14,2 MHz. and 0,22 μV at 28,5 MHz. Which the
receiver calibrated according to the information in the article on 'Reporting', i.e. S-meter reading S-9 at
a measured antenna input voltage of 50 μV.,
these noise voltages will translate into S 4,2 at the Local factors Considering these generalized
noise figures it is clear that reports on back-ground noise figures below S-2
on The origin of antenna loss figures
have been discussed in the chapter 'Where
does the HF power go'. An important contribution to this
loss story comes from the antenna feeders, especially at high SWR. It
therefore may be a good idea to always consider termination conditions,
particularly with RG58 coaxial lines over A cable choke will prove to be of
good value to keep unwanted signals out, either from noise sources or your
own transmission signal; therefore such cable choke in general is always a
good idea. More on cable chokes and baluns may be found here.
The above noise figures will be
about the lowest signals you may find on HF amateur frequencies. Higher noise
figures are no exception either when considering noise figures may easily be
up to 20 - 30 dB more in city-centers and industrial environments. Noise
floor may then be measured at S 8-9 at From the above it is clear total
atmospheric background noise already is high enough as it is. This is another
reason why this 'PLC - 'hold-up' should be stopped before it will kill a
world-wide communication channel.
Bob J. van Donselaar, on9cvd@veron.nl
|
|
