Ideas About The Future Of The Medium Wave

some arguments for wide-band-AM, some thoughts about DRM

by Peer-Axel Kroeske (DL2LBP).

Sound Quality must be improved

The poor sound quality is probably the main reason that listeners stay away from the medium wave. In the UK, Spain, America and parts of the third world it is still alive though. In these countries MW programmes exist, which cannot be heard on FM. This is the key to MW attraction. But MW stations still have less success because of the poor sound.

Is DRM the future?

The digital standard DRM has tried to push forward for a couple of years now. The break-through still seems far away. And I remain sceptical. Of course, there is a potential. In theory you could extend DRM with a flexible frequency concept like RDS on FM. An intelligent way of frequency switching is desperately needed on LW/MW/SW even more than on FM when you think of 24h coverage for large areas. The main problem is probably the lack of cheap receivers. As long as there aren't enough potential listeners, stations won't invest into DRM (and DAB). While DAB is becomming affordable, the cheap chipset for DRM is still missing, let alone the antenna problem.
The competition is high: FM usually offers about 20 stations, internet radio via cell phone networks (where available) offers the ultimate variety. DRM could become attractive for commercial stations, at least on shortwave, because you cover large areas with only one transmitter. On medium wave it's probably difficult to set up complex SFN-DRM networks. Crowded frequencies at night don't leave much space for adequate DRM signal to noise ratios (SNR). There would be an option for local stations which couldn't get FM frequencies. But they should rather use VHF frequencies to avoid interference problems (see below). DRM on the former TV band I frequencies has already been discussed. And DAB still offers capacity in most places.

One advantage could be the sound quality. The DRM developers compare it with mono FM sound. But 24kbps AAC is far from FM Hifi sound. The DRM signal is heavily compressed resulting in a very artificial sound impression. Cosmetics by SBR technique help only little and could theoretically also be implemented to analogue AM. 48kbps requires double channels (20 KHz). I'd recommend that but it doesn't look like any station is using this in the near future. As far as sound is concerned DRM is still an improvement to nowaday-AM. However, the main potential for DRM can be found on short wave.
A main problem of DRM is interference: I've experienced that DRM needs absolutely free frequencies. If there's only one weak carrier (could be from an AM station using the same channel) SNR drops down and the signal disappears. Indoor environments are full of interfering signals by computers, power supplies and various electronics. While AM can still be received in such a situation, many DRM signals stay silent indoors.

DRM introduction takes time

If you've ever had a look at receivers in 3rd world countries this becomes clear: People still use receivers from the 60ies and 70ies, they repair them again and again. New DRM receivers are still expensive. It would take much time until all receivers are exchanged.

In the rich part of the world people can afford new receivers. But only a few would buy them. The big majority is satisfied with FM and alternatives like the internet as mentioned. International foreign services do not really attract too many people. Again: DRM must become affordable. And a number of attractive stations (Information or popular/special music formats) must start at the same time.

Is a co-existence of DRM and AM possible?

Yes. But not on the same frequency. As mentioned earlier, DRM causes hard intereference on AM and vice versa. Both modes should get their own frequency segments.

First aid: Wide-band AM

AM does not sound dump by nature. It's the restrictions. Right now, the official audio bandwidth in Europe is poor 4.5 kHz, not much better than phone sound. Actually, many countries have already lifted that up: in the UK and the Netherlands there seems to be a common agreement on 6.3 KHz, Southern Europe uses 9 KHz and more, in America, Australia and Japan the standard is 10 KHz. That's close to hifi if it's a local undistorted signal. And it's 100% compatible to every receiver. The only problem is the question of filters. Some stations enhance treble to improve the sound in receivers with narrow filters. But if you listen with a wide filter it all sounds terribly manipulated. Wideband can be easily be introduced in the daytime when MW is empty, only occupied by local stations. In the night neighbouring channels could interfere with each other.

Wide-band AM doesn't require more frequency resources

If stations suppress one sideband, wide-band AM wouldn't even occupy more frequency bandwidth and would not produce additional interference in the night. There is one disadvantage: Sideband suppression could lead to some distortion. The solution could be in suppressing only parts of the sideband: reduced power or a limited frequency range. This would have a similar effect as if you tune a receiver one or two KHz from the center frequency to catch up more treble. Synchronous AM-detectors (SAM) can help to solve the problem: Signals get distorted as soon as the carrier is notched out for a while. But it's easily possible to amplify the carrier in the receiver. This technique is called synchronous AM and should be implemented in more AM receivers.

Intelligent DSP for AM

With digital sound processing you can improve sound quality of AM signals much further. Chip power can not only demodulate DRM but also boost up AM signals. It could adjust frequency phasing, notch out beeping carriers, simulate stereo and so on. And as long as it's AM it would still be downward compatible.

Day and night on medium wave

But which stations should be on medium wave at all? Coveage completely changes from day to night. During light hours, it's a local band. In the night signals reach out a few 1000 km. Unfortunately, the frequency coordination in Europe seems to ignore this unconvenient fact. In most cases transmitters are allowed to broadcast 24 hours with power only a little reduced at night. The result: in the day MW is relatively empty, at night an interference chaos rules. So what can be improved? It would be a bad idea to create a MW scenario with different stations at day and night I think. No comercial station could afford parttime presence. Think of the northern regions: it's dark during 6-8 am prime time in the winter time, in the summer the sun is already high in the sky. The better concept would be 24h with one nationwide program on one frequency and a flexible usage of transmitters. With some intelligence the unconvienient propagation characteristics could be turned into an advantage.

Build up more single frequency networks (SFN)

There are networks which use up to 20 frequencies for the same programme. This has historic reasons but today it leads into chaos. In the old days it was difficult to modulate transmitters simultaneously. As soon as you received two transmitters of one program, there would be a terrible echo. Today with satellite feeds, SFNs are no problem anymore. You might argue that SFNs sound bad as soon as you receive more than one transmitter at once. This results in frequency shaping and stronger distortion (without S-AM). But it's still better than the existing mix of interfering signals. 80% of the channels can't be used at the moment when it's dark. Listeners appreciate to know their station in one and the same place. Transmitters should be launched close to every big cities. Single or double-frequency networks with small-sized transmitters close to the cities would reach many people. A small number of big transmitters would back up the signal in the countryside.

International coordination only for strong transmitters

Unfortunately it is almost impossible in many countries to find new MW frequencies for new stations. International coordination is a big hurdle to take. But is international coordination really neccessary for local daytime stations on medium wave? Transmitters of let's say 50 kw(day) and 5kW(night) don't reach much further than 100 km in the day (with simple radios). With a simplified coordination procedure, many wide-band-AM stations could be launced around the metropolitan areas.

Foreign services should transmit from the target country

The principle that foreign services use transmitters in their own country seems to be old-fashioned nowadays. This is a more or less a relict of the cold war when stations weren't allowed to transmit their propaganda from the other block. Today with freedom of speech this should be posible. Something like World Radio Network on satellite could work on medium wave, too. Here you hear different international services every hour like Bulgaria at 20UTC, China at 21UTC, Radio France at 22UTC (just an example, real times are different). You could even build up national networks for such a service.


So what we need is a "Zerobase" (re coordination from scratch) for the medium wave with wide-band AM. The potential of good old amplitude modulation is big and has not been discussed enough yet. Clear channels without interference and a better sound quality are the keys. DRM could also be worth a try but only on exclusive channels, best with 20 KHz bandwidth. But AM is just easier to handle.


Jacques (ON5MJ) wrote:
You suggest on your web site to enlarge the bandwith of AM stations. This is not that simple to do as increasing the audio bandwidth because at the same time the power must be reduced in order to respect the maximum power dissipation in the final amplifier of the station. It is well known that SSB permits 12 times more power efficiency than AM due to one side band and carrier suppression. Nevertheless one can imagine a tradeoff in using a single side band modulation WITH the carrier remaining (because usual AM receiver don't have a BFO in AM). This would be compatible with the present AM usual receivers and be more powerfull than simply increasing the audio bandwidth.

Questions? Corrections? Ideas? Comments?
Mail me:


This page is based at:

Back to