AM Broadcast Signal

[ssaunders]

Folks,

I'm not entirely sure where to post this, so apologies if I've hit the wrong target.

One for the broadcast engineers…
This week I'm interested in the characteristics of AM broadcast signals - specifically what the audio bandwidth and modulation depth of your work-a-day AM broadcast is?

Do broadcasts characteristics vary between stations and/or short / medium / longwave?

Is there some kind of international standard covering this kind of thing?

Thanks

Simon

[Jolly 7]

Having listened to many AM stations in and from countries across the globe, definitely the audio quality of some live broadcast stations can be worse than others. My guess is that higher broadcast fidelity costs more money to achieve and a compromise is made in most cases.

[dominicbeesley]

Not an easy read but this goes into details of modulation depths https://www.itu.int/dms_pub/itu-r/op...2018-PDF-E.pdf

D

[paulsherwin]

Almost all AM broadcast stations use very heavy audio processing to make them sound louder and maximise their reception areas, so it's difficult to comment on the static parameters. All European stations are supposed to use 4.5kHz brickwall filters to stay within their 9kHz channel allocations.

[Oldmadham]

AM Broadcast audio bandwidth depends on what country the Station is in.
Some countries restrict the audio bandwidth to around 5-6 kHz, whereas in others it is around 9KHz
(In Oz, it used to be 10kHz, but it was reduced to 9kHz).

[russell_w_b]

Analogue HF (short-wave) broadcasts emanating from UK or their relay stations overseas are 10kHz bandwidth; 9kHz for domestic MF / LF - or they were up until 2013 when I was last involved, anyway. I suspect some domestic broadcasting exceeds the 4.5kHz audio bandwidth nowadays, getting away with more and more as stations become fewer.

AMC (Amplitude-Modulated Companding) was/is used apart from on the few high-level modulated AM sets that remain: 6db compression for HF so that the peaks of mod at 100% equate to unmodulated carrier level. AM broadcast used 3db AMC. But all this may have changed nowadays.

'Optimod' or similar audio processing at the transmitters provided the brickwall cut-off characteristics - if set up accordingly.

The ITU document should cover it all.

[paulsherwin]

9kHz is the channel width and translates to a 4.5kHz audio bandwidth limit when both sidebands are taken into account. Not all territories insist on brickwall filters to stay within the channel though - Britain didn't until the 70s. I don't know the current position in Oz.

[kalee20]

Originally Posted by russell_w_b

I suspect some domestic broadcasting exceeds the 4.5kHz audio bandwidth nowadays, getting away with more and more as stations become fewer.

Wishful thinking perhaps - but as MW / SW stations become fewer, we might see an improvement in quality as the remaining stations open-up their audio bandwidth.

Sure, they will spill outside their 9kHz allocated channel, but with not so many other stations to interfere with, it may become tacitly accepted.

[paulsherwin]

I'm pretty sure Ofcom frown on out of band splatter, and as it's all done in the audio processors now it needs a conscious decision by somebody to change it. I suppose small community stations without professional Optimod setups may go out of band either deliberately or by mistake. Radios built since the 80s normally have ceramic filter IF stages which result in a brickwall effect anyway though.

[G6Tanuki]

Originally Posted by kalee20

Wishful thinking perhaps - but as MW / SW stations become fewer, we might see an improvement in quality as the remaining stations open-up their audio bandwidth.

Sure, they will spill outside their 9kHz allocated channel, but with not so many other stations to interfere with, it may become tacitly accepted.

I doubt it will happen - spreading the radiated energy across a wider bandwidth tends to reduce communications-efficiency if you're using AM (this is one reason why SSB became the default for both ham and 'professional' communications on SW) - it also increases the likelihood of selective fading (where one sideband fades/changes-phase with respect to the other, causing a particularly objectionable form of audio distortion).

Also, for the last 50-odd years SW/MW/LW receivers have generally aimed at narrowing the audio-bandwidth to avoid interference - I doubt any manufacturer's going to rejig their receiver's filtering. MW/LW/SW coverage is these days very much an afterthought for most radio-manufacturers.

Finally, most LW/MW/SW broadcast-transmitter installations are now 'legacy' infrastructure, which a lot of the broadcasters would like to shut down altogether! Spending any time/effort above the bare minimum needed to keep them going in accordance with their existing licence-conditions isn't likely to be high priority.

[Ian - G4JQT]

Probably a couple of decades ago now while at work, I got some info from the Engineer In Charge at Droitwitch regarding the audio BWs of the stations being broadcast from there.

To my irritation I can't lay my hands on the info right now, but I do remember it depended on the station being broadcast, Virgin having wider bandwidth than the predominantly voice content BBC stations.

[Davewantsone]

Originally Posted by Ian - G4JQT

Probably a couple of decades ago now while at work, I got some info from the Engineer In Charge at Droitwitch regarding the audio BWs of the stations being broadcast from there.

To my irritation I can't lay my hands on the info right now, but I do remember it depended on the station being broadcast, Virgin having wider bandwidth than the predominantly voice content BBC stations.

I remember reading an article which was in an old edition of Wireless World from the mid 1930s. It showed the frequency response of a signal sent by land line from London to Droitwich radiated and picked up by the BBC in London. The graph showed that the frequency response went up to 8kHz, So It would appear that in the 1930s the transmitted bandwidth went up to 16kHz. The article also went on to report on the poor stability of overseas broadcasters to maintain correct broadcasting frequencies.

[Oldmadham]

In the Oz case, I was referring to the audio bandwidth.
At 9KHz, this translates to 18KHz channel width.

I have done more frequency response tests on ABC MW AM Broadcast transmitters "than most people have had hot breakfasts", so the figures are burnt into my memory.
I don't think the bandwidth has been reduced since, but with the obsequious attitude of the regulators to "world's best practice" anything can happen!

Back when they were 10kHz audio bandwidth, the limiters were set for 3dB compression.
The corresponding figures at that time, for the "regional shortwave transmitters" VLX & VLW (now defunct) were 7kHz & 6dB.

These latter were internal broadcasters for the (then) vast areas of Western Australia which had no access to MW broadcasting.

[Synchrodyne]

Back in 2005, when I was living in Sydney, I asked the ABC about AM transmission bandwidth. Here is the response I received:

“The ABC MF AM transmissions have audio bandwidth that is restricted by the licensed spectrum (20kHz total per service, i.e. +/- 10kHz), in order to avoid interference to adjacent channels.

“The primary determinant in the transmitted bandwidth is a sharp-cutoff lowpass filter included in the audio processing prior to the transmitter. The actual cutoff frequency sometimes has to be varied in order to meet spectrum and operational requirements. In general however, a minimum realisable bandwidth would be about 7.5kHz, and some services (ABC Radio National in Sydney for example) have realisable bandwidth out to about 9.7kHz.

“Pre-emphasis is very widely used in AM transmission to compensate for the high-frequency rolloff inherent in most types of AM receiver. The characteristic used on ABC MF services is the US NRSC characteristic curve. This is basically equivalent to an R-C time constant of 75 microseconds, with an added shelving at about 8700Hz.

“Other factors in the program chain are generally controlled to allow the transmitted signal to achieve the potential performance. In particular as many program distribution paths can also be used for FM transmissions, their quality is not a limiting factor for MF services.

“Given the above, the use of a wideband receiver can yield an audible improvement in many cases.”


More generally, the subject of AM transmission bandwidths is recurrent here, and much has been said about it. Amongst others are these three threads:

https://www.vintage-radio.net/forum/...=1#post1072168
https://vintage-radio.net/forum/showthread.php?t=120815
https://www.vintage-radio.net/forum/...ad.php?t=70496


Cheers,

['LIVEWIRE?']

The first and third of those links both produce a '404 File not found' message, but the second is working. That thread, as some members may not know, contains information about European MW stations which were due to close in 2015/16. Not that I'm a regular listener to AM radio these days, but, AFAIK, most, if not all those mentioned in that thread, did indeed close down 4 years ago.

[lesmw0sec]

Originally Posted by Oldmadham

In the Oz case, I was referring to the audio bandwidth.
At 9KHz, this translates to 18KHz channel width.

I have done more frequency response tests on ABC MW AM Broadcast transmitters "than most people have had hot breakfasts", so the figures are burnt into my memory.
I don't think the bandwidth has been reduced since, but with the obsequious attitude of the regulators to "world's best practice" anything can happen!

Back when they were 10kHz audio bandwidth, the limiters were set for 3dB compression.
The corresponding figures at that time, for the "regional shortwave transmitters" VLX & VLW (now defunct) were 7kHz & 6dB.

These latter were internal broadcasters for the (then) vast areas of Western Australia which had no access to MW broadcasting.

When I lived in Queensland, I recall thinking that the AM transmissions were nice and crisp, compared with the woolly sound in the UK!

[Hybrid tellies]

This graph shows the response of an AF band limit filter used by most AM transmitters to limit interference to adjacent channels. As you can see it does not cut off at 4.5Khz but tails off at 6Khz.
AM receivers that have a similiar response curve the audio can sound reasonably pleasant even with music.
The problems occur when AM receivers, especially those built in the late 1970's and for much of the 1980's were fitted with brickwall or narrowband IF filters which cut off all audio above 4Khz. The AM audio quality on such receivers was horrendous especially with the heavy compression that was used at the transmitters back in those days.
Another point to bear in mind is that even if the broadcasters were able to provide a wider bandwidth most AM receivers would not be able to benefit from this.

[boxdoctor]

A rough comparison can be obtained by looking at some of the transmissions as received at the Utwente S.D.R. in the Netherlands. http://websdr.ewi.utwente.nl:8901/
For illustration a part of the long-wave frequency spectrum is shown, and it is possible not only to approximate the bandwidth, but also observe to an extent the amount of compression some stations use compared to others. (Fairly severe on R.T.L.)
On short wave, Radio China do not appear to believe in bandwidth limitation at all on some of their outlets. Tony.

[Synchrodyne]

Originally Posted by 'LIVEWIRE?'

The first and third of those links both produce a '404 File not found' message, but the second is working. That thread, as some members may not know, contains information about European MW stations which were due to close in 2015/16. Not that I'm a regular listener to AM radio these days, but, AFAIK, most, if not all those mentioned in that thread, did indeed close down 4 years ago.

I thimk that these will work:

https://www.vintage-radio.net/forum/...=1#post1072168

https://www.vintage-radio.net/forum/...ad.php?t=70496

I had copied the three over from this post:

https://www.vintage-radio.net/forum/...0&postcount=22

Looking at that, it would appear that the 1st and 3rd links were abbreviated, but not the 2nd. Perhaps that's why the 1st and 3rd didn't work when copied over as they appeared in the thread....


Cheers,

[Synchrodyne]

Originally Posted by Oldmadham

I don't think the bandwidth has been reduced since, but with the obsequious attitude of the regulators to "world's best practice" anything can happen!

The irony is that Australia and New Zealand probably represented best practice as far as AM broadcasting was concerned. The change from 10 to 9 kHz channelling c.1978-79 was probably neither needed nor beneficial downunder, but was done to conform to international practice.

The Australian and NZ equipment makers, such as Audiosound, Allen Wright, AWA and Fountain offered decent, wideband AM tuners through the 1970s and into the 1980s. But very little imported equipment offered this facility, and in that period most of the Japanese manufacturers (Sansui being a notable exception) fitted quite atrocious AM sections to their hi-fi equipment. “Electronics Australia” (EA) magazine called out this practice. As an example, in its 1979 January review of the Technics SA-400 receiver, it said: “Once again, the AM tuner in the SA-400 is just about as poor as any in typical Japanese stereo receivers.” And in a 1985 December review of the Sony ST444ES tuner, it said: “It’s a pity that the AM performance is but the merest shadow of the standard for FM. While the signal to noise ratio was quite respectable at 61 dB with respect to 100% modulation, the audio bandwidth was particularly narrow, with a -3 dB response between 200 and 950 Hz.”

EA also included wideband AM tuners amongst its constructional projects, but it also had some designs with conventional narrowband AM sections. One of the latter was the Playmaster AM-FM Stereo Tuner-Clock, described in the 1978 November and December and 1979 January issues. Of its AM performance EA said: “While the AM performance is nothing to become excited about, the AM specifications of this tuner are actually better than those obtained from many commercial tuners. This may be hard to believe, but it’s true.” The AM frequency response specification was 30 to 3000 Hz at the –3 dB points!

In respect of transmitted quality, a comment that was made by Allen Wright about the two wideband AM tuner models he had produced: “Both sounded like good FM , just in mono, from the superb quality non-commercial government stations. Compressed commercial stations sounded just that - compressed!”


Cheers,