Some FM Basics
In the summer of 1939, radio was
definitely king. It was AM radio, of course, with all that system's inherent limitations.
Static, narrow bandwidth, and interference "birdies" were as much a part of the
entertainment scene as Amos ‘n Andy. In July of that year, an inventor named
Major Edwin H. Armstrong demonstrated that radio need not suffer these particular woes. It
took another decade before FM got any sort of penetration at all, and yet another before a
significant number of people could receive it. And it had to wait until the ‘70s
before it became a major entertainment source.
Nowadays, even with the advent of various digital
broadcasting systems, FM is still one of the prime sources of hi-fi program material.
Here's a basic look at the medium and what it does.
How it works
The development of radio came from a desire to hear sounds
produced in one location at a listening position some distance away. Sound itself is quite
inefficient in this regard because it doesn't travel very far or very fast. What is needed
is something that does travel far and fast (say, light waves, or electromagnetic ones) to
carry the sound signal in some form to the listening position, where it can be turned back
into sound. To this end, electromagnetic waves are "modulated" by the desired
audio information -- that is, they are in some way continuously varied in step with the
audio signal. The radio receiver detects these variations, and creates an audio signal to
correspond.
Sound itself is made up of changes in amplitude. It's not
surprising, therefore, that the original way of modulating a radio wave was to vary its
level (or amplitude) in step with the audio signal. Unfortunately, most of the electrical
noise that abounds out there also translates into amplitude changes, and an AM receiver is
every bit as sensitive to them as to the program material.
Armstrong's idea was to vary the frequency of the
carrier signal, rather than its level. The receiver would detect these frequency shifts,
and translate them back into amplitude variations, to be reproduced in the normal manner.
The receiver could be made impervious to amplitude variations, and so impervious to static
and other sorts of noise.
Is the fi higher?
We tend to think of FM as being a hi-fi medium, and AM as
being less so. Certainly in terms of noise there is some justification for this, and it is
also true that little effort has been made over the years to increase the bandwidth of AM
broadcasting. Technically, however, there is no reason that FM should have a wider range
than AM. But in any event, even FM does not have as good frequency response as you might
suppose. We are accustomed to think of the ideal frequency response for high-quality audio
equipment to be 20Hz to 20kHz within a certain tolerance. FM, however, cuts off pretty
sharply at an upper limit of 15kHz, mainly because of the stereo pilot signal at 19kHz.
Whether or not it is important for a piece of audio equipment to extend beyond 15kHz is a
matter of some controversy, but the fact is that FM does not.
FM stereo
One of the advantages FM had over AM from the early 1960s
is its stereo capability. Several AM-stereo systems were introduced later, but even though
some stations began to broadcast in stereo, both radio manufacturers and listeners ignored
the system. FM lends itself more easily to additional channels of information, in any
case; commercial signals were added to FM programming (for example, to feed background
music to restaurants and stores) long before the advent of FM stereo. It would have been
quite easy to use two channels, one for the left and one for the right.
When FM stereo was proposed, however, one of the main
criteria was that it be compatible with mono. Otherwise, every FM receiver then in
existence would be obsolete. If one channel was used for left, and another for right, a
mono receiver would pick up only one channel.
To overcome this, a system was devised that did indeed use
two channels of information, but not in conventional stereo style. Instead, the main
channel combines the left and right information to produce a mono signal that can be
picked up by existing tuners and future mono radios. The second channel of information
contains a left-minus-right signal, in which the two channels are combined out of phase.
This is then modulated on a 38kHz subcarrier that can either be ignored by a mono
receiver, or used by a stereo one to reconstruct the original two program channels.
In the receiver, the two signals -- left-plus-right and
left-minus-right -- are combined in two ways. In the first, they are added in phase, so
that the right channels cancel out, leaving only left information; in the second, they are
added out of phase, so that the left channels cancel out, leaving right information. All
of this cancellation results in some signal loss, with the result that stereo FM tends to
have somewhat higher noise levels than mono; but this only becomes a problem if you are
trying to listen to distant FM stations.
In addition to the left-minus-right information modulated
on the 38kHz subcarrier, the stereo FM signal also contains a tone of 19kHz that lights
the stereo light on your receiver, and switches it into the stereo mode. The presence of
this signal makes it necessary to chop the audio signal somewhat sharply above 15kHz, or
the pilot might interfere with the upper frequencies of the music.
How far does it go?
Like television, FM signals travel in a line-of-sight from
the transmitting to the receiving antennas (it is similar to television because the FM
band is in fact wedged between the channels of the TV band). AM, on the other hand, has
the curious ability to reflect off the ionosphere and bounce back to earth, particularly
at night. So, while you may be able to listen to AM signals from many thousands of miles
away, you are restricted mostly to local FM signals.
Also like television, FM signals can bounce off nearby
buildings, causing a condition analogous to TV "ghosting." This is called
"multipath distortion," and arises because the same signal is arriving directly
and also a fraction of a second later, after having taken a somewhat longer route to
whatever it is bouncing off and back. Usually, adjustment of the receiving antenna will
cure such a problem.
What to look for
As with other areas in audio, there are certain
specifications that have to be considered when buying an FM tuner. Perhaps the most often
quoted in early days was a unit's "usable sensitivity," which indicates the
smallest signal a tuner can receive while still attaining a specified level of noise
performance. This traditionally has been a quieting level of some 30dB -- that is, the
noise level is 30dB below that of a 100% modulated signal.
Usable sensitivity is not really an adequate measure,
however. It was devised back in the days of mono FM, and it is usually still quoted as a
mono figure, which, as we have seen, is better than a stereo signal in terms of noise.
Also, usable sensitivity is specified in microvolts (about 2 being a common spec), and
this does not take into account the impedance the antenna is working into.
In the 1970s, a move was made to institute a new standard
for sensitivity measurements. The unit was to be the "dBf" -- decibels referred
to one femtowatt of power. As a power measurement, it had the advantage of taking into
account the impedance of the antenna; and as a further refinement, a more realistic
quieting level of 50dB was chosen. Stereo operation was also taken into account. Levels of
about 12dBf in mono or 36dBf in stereo are excellent.
One feature of the FM system is that, if there are two
signals on the same frequency, the tuner has the ability to receive the stronger one, and
reject the weaker one completely, provided there is sufficient difference in signal
strengths. This difference is called the "capture ratio" and is specified in dB,
readings of 1 or 2dB being typical. Related to this is "selectivity," which is
the ability of a tuner to reject nearby signals on the band when tuned into the station
you want to hear. It also is specified in dB, but in this case the numbers represent the
amount by which the unwanted signal is attenuated.
The FM band is divided up into 100kHz chunks, in North
America, anyway; stations that are geographically close should be at least 200kHz apart.
Nevertheless, with today's highly sensitive receivers, there are lots of areas where one
can receive stations on adjacent channels (i.e., only 100kHz apart). Some manufacturers
specify "adjacent channel selectivity," therefore; although most prefer to quote
figures for "alternate channel selectivity" (referring to stations 200kHz from
the desired one) because the figures tend to look better. Selectivity in the order of 60dB
is common in the latter case.
Channel separation is another specification that has some
importance, because it is a measure of the ability of the adding and subtracting circuits
mentioned above to recreate two discrete channels from the in- and out-of-phase
information carried on the FM signal. Separation across the whole frequency band should be
quoted, but quite often manufacturers specify it only at one frequency in the middle of
the spectrum (typically 1kHz), where separation is relatively easy to achieve. Numbers of
35 to 40dB are common, although somewhat less would be acceptable -- psychoacoustically,
as little as 20dB separation is adequate for good stereo imaging.
Surprisingly, some of the "standard" hi-fi specs
are of less importance with FM tuners than with other audio components. For example, we
have seen that the frequency response of the FM system as a whole is somewhat narrower
than we expect to find with other hi-fi components; so a tuner that is flat between the
limits of 30Hz and 15kHz will do just fine. Distortion and noise figures for the tuner are
likely to be far less than that included in the signal at the transmitting end, so such
specifications need not assume any great importance in selecting an FM tuner.
...Ian G. Masters
ian@mastersonaudio.com
|
