[BC] The AM bandwidth issue

[Robert Meuser]

Barry

I agree with your comments but the problem runs deeper. Let me address a
number of comments in different threads. First the FCC has always been a
political entity and before that so was the FRC. In the early days of radio
it was the FRC that first established a social policy that determined both
the content (in general) and allocation for AM stations. This was the first
crack in the dike in allowing more small stations into markets. Subsequently
both the FCC and international treaties began to ignore the true physics of
AM by allocating channel spacing and protection that allowed a significant
overlap of first adjacent channels. Fast forward to post world war two when
many new marginal stations were licensed counter to any engineering
standards. At this point a convergence of technology and economics came into
play. By the late 60s when more and more stations were overlapping each
other and receiver manufacturers were able to switch from tuned IF circuits
to ceramic filters. This accomplished several things, it reduced assembly
cost as well as parts count and it enhanced selectivity, necessary due to
the increased number of stations. It is at this point that broadcasters
began to boost high frequency audio response to counter the narrower
filters. Ultimately this led to increased splatter and more interference and
ultimately the unending cycle of narrower bandwidth vs more high end boost.
It took less than 2 decades for ceramic filters to seriously contribute to
the decline of AM.

The FCC has historically written rules that favor administrative convenience
over technical necessity. Ignoring the result of sideband response in
directional arrays is but one example and FM allocations are another.

Those who think wideband AM is commercially feasible are merely deluding
themselves. Physics dictates that more bandwidth equals more noise. Being
that AM is a noise sensitive medium there will be a direct relationship
between bandwidth and noise. I do not think that in the current environment
10 KHZ audio is a reasonable possibility and wider bandwidth is off the
table if the signal strength contours now accepted remain. I was an advocate
of wideband AM and had some serious argument with Chris Payne when he was at
Motorola but then I had the good fortune to build AM stations where most of
the city grade was near 100 MV/M and the fringe was agricultural with much
lower noise and few interfering stations. The reality is different today.

What is disappointing is that with DSP there can be some mitigation of
physical limitations but for whatever reason there are no really decent AM
digiceivers. This could reverse the problem addressed in the 70s with
ceramic filters regarding costs vs. selectivity as digiceivers are cheap to
build and can deliver much more fidelity vs. bandwidth than can ceramic
filters. 

In my own empirical experience a receiver response that is -3 DB at 5 KHZ
with a Bessel shape that has extra zeros so that everything is gone before
7.5 kilohertz and a complimentary transmitter response (+3 DB at 5 KHZ and
zeroes at 7.5 kilohertz) would be a good compromise. With DSP this is easily
accomplished. There are also algorithms to eliminate impulse noise as well
as other noise reduction algorithms for noise closer to white noise.

Bottom line is that broadcasters and government set the path for the demise
of AM since almost the beginning of the service and there are laws of
physics that limit what is realistically possible.

On 11/25/10 2:49 PM, "Barry Mishkind" <barry at oldradio.com> wrote:

> 
>         The manufacturers, by then, had *already* rolled back
>         the bandwidth, as a protection. The original NRSC
>         was not mandated. It was a "serving suggestion."
> 
>         And since the NRSC was designed to make the best
>         of bad receiver bandwidth, it was only a "holding"
>         place. Again, I will defer to Bob or Greg's recollection,
>         as they were a major part of it. However, the
>         manufacturers had no incentive to "open up" no
>         guarantee that things would ever get better.