SPECTRUM OF FM ATV
By G8MNY
Most people are familiar with FM modulation and the sidebands it produces,
but Colour ATV with intercarriers is more complex and it is very difficult
to see and understand what is going on.
ATV COMPONENT PARTS
SOUND SUBCARRIER
The signal feed to the FM modulator has many components, the simplest
of which is the 6MHz FM subcarrier sound. This has a peak deviation of
+/-50KHz. The audio pre-emphasis network should give 50uS pre-emphasized
(+15dB @ 15KHz Treble lifted) audio, to obtain best signal to noise from
the FM process. Audio Bandwidth 30-15KHz @-3dB is easily possible.
LUMINANCE and SYNCS
The Luminance and Syncs signal has a bandwidth from near DC (10Hz) to
5.25MHz.
A flat gain, phase, and group delay, over this frequency range is essential
if the picture is to remain intact. Frame sync are at 50Hz and Line is
at 15.625KHz, and all the picture information is repeated at these rates.
COLOUR SUBCARRIER
To the luminance is added the complex 4.43MHz PAL colour signal. It is
a double sideband signal made with a Phase & Amplitude modulator driven
from the colour difference signals. It produces a colour signal bandwidth
of 1.3MHz.
FM TELEVISION
PRE-EMPHASIS
As with the FM audio, the TV signal to noise ratio, is much improved when
Pre-emphasis and de-emphasis is used. This is because the FM modulation
and demodulation process suffer from high levels of HF noise.
 With
the Video signal frequency range, simple CR time constant pre-emphasis,
will not work due to the number of octaves it contains. The emphasis standard
used for all FM Video is the CCIR 405, that describes a 14dB HF lifting
curve, made from 5 components. This gives lift only to the picture detail
like, edges and colour information, above 2MHz that suffer the most from
noise, leaving the bulk of the signal unchanged. It is applied to the
composite video signal before modulation, but not the sound subcarrier!
FM MODULATION PROCESS
Rather than attempting to explain the complexity of all these signals
applied to an FM modulator, lets go back a step and consider what happens
when we FM modulate with a simple square wave.
With a square wave you get two CW carriers switched on and off at the
square wave frequency both with sets of AM sidebands. The depth of deviation
only separates the two CW frequencies. So high deviation separates these
signals, low brings them together so there sidebands overlap. Increasing
the Square wave frequency, just increases each of the CW sideband widths.
Now you know what is going on, lets take a look at simple video signal
consisting of 8 levels of Luminance and syncs (standard grey scale) this
gives 9 CW carriers all pulsing on and off in sequence at line rate.
However if we use 6MHz sound subcarrier with no video, we generate 2
main subcarriers 12Mhz apart, at a level dependent on the injection level
similar to AM.
Put back the grey scale video with the subcarriers and we have not 9,
but 27 CW signals. Consisting of the 9 instantaneous carrier frequencys
each with a pair of lower power sound subcarrier sidebands at +/-6MHz.
From this you can see there are repetitions of all the Video information
on the sound subcarriers.
BANDWIDTH
Without taking the higher order of FM sidebands into account, it can be
shown the full channel needed is 19MHz wide. If we look at the above example
the two sound subcarrier are 12MHz apart. When the recommended +/-3.5MHz
peak deviation of video is applied, each subcarrier frequencys are also
moved over the same +/-3.5 MHz, making
12MHz + 7MHz = 19MHz total subcarrier displacement.
FM SIDEBANDS
Go on forever! Yes its is true, unfortunately with FM not only the first
order of sidebands are generated. The levels of the other orders (Fm x
2,3,4,5,6,7,8,9,...) depending on the Modulation index (Dev/Fm), but they
are always present at some level!
However the video signal varying in intensity, pre-emphasis, and sub-carriers,
the use of Modulation Index to try to describe what levels of higher order
products you actually get is just not possible. In practice these higher
orders products decrease depending on their fundamental injection level.
FM ATV DEVIATION
To limit bandwidth used (unlike satellite TV where MI=1.5 with 6 or more
subcarriers) ATV uses only a peak deviation of +/-3.5MHz (MI=0.7). However
with CCIR pre-emphasis the low frequency (<2MHz) deviation is practice
-14dB below the +/-3.5MHz peak deviation which gives;-
0.2 x +/-3.5 MHz = +/-0.7MHz.
So most of the signal energy is near the carrier, only the colour &
subcarriers are spread.
When using a Satellite Rx for ATV, this narrow deviation causes problems
of low video gain and high noise (& adjacent channel RADAR).
 SUBCARRIER
LEVEL
The ATV recommended subcarrier levels are -18dBc.
In the receiver this level of subcarrier just about provides FM limiting
for the sound demodulator, when the picture is about P4 grade. But as
the Picture grades reduce the sound will be totally lost at a P2 while
a noisy colour picture remains.
The result of this subcarrier level however is to make the ATV signals
19MHz edges -51dBc with successive 6Mhz products falling off at a rate
of 18dB/6MHz.
FILTERING
Although narrow Rx is effective on FM for weak signal, the effect on good
signals when the weaker parts of the spectrum are lost due to over filtering
causes all sorts of video distortions like;-
Sparkly edges,
Colour problems,
Video on sound buzz,
Poor syncs.
For Transmitters as well as the above problems, tight filters cause SWR
increases as modulation is applied, introducing AM modulation mixing problems
in the PA.
Filtering below 19MHz flat bandwidth will cause some of the above on
good P5 signals, but will not be noticed on weaker ones. Filtering below
12MHz on weak signals is very useful where sound & quality are not
wanted.
|
