Why are there no frequency modulated broadcasts on the long/medium/short wave, but only on VHF? I'm sure there are many historical and political reasons (long/medium/short wave were already in use for AM when FM radio was developed, VHF still had lots of free bandwidth), but are there also technical reasons?

FM radio gives a better sound quality than AM. Common FM on VHF also uses a much broader bandwidth and interchannel spacing than MW AM, so using the same bandwidth per channel much fewer FM channels would fit in the MW band than AM channels. But narrowband FM can be used with similar interchannel spacing as is typical for AM, so this is no reason not to use frequency modulation.

So are there any technical reasons why narrowband FM broadcasting is not used on long/medium/short wave, and if so what are they? I'm assuming that even narrowband FM would give better audio quality than AM, am I wrong about that?

(disclaimer: maybe there have been some FM transmissions on MW that I'm not aware of, in that case please read this question as "why are there nearly no FM broadcasts on the L/M/SW ...")

(I'm asking out of historical interest. I'm not planning on setting up any FM on MW or something)

  • $\begingroup$ Nitpick: The reason you do not hear AM on the HF bands is because using SSB allows for double the number of simultaneous conversations within the same band. Same thing happened to CB. $\endgroup$
    – SDsolar
    Commented Aug 20, 2017 at 20:38
  • $\begingroup$ @SDsolar Also, SSB has a much better S/N ratio. $\endgroup$ Commented Aug 20, 2017 at 20:49
  • $\begingroup$ I don't have the time for a good answer but here's a direction to look into; Bandwidth of ww2 radios used in tanks. It's pretty big. Also the marine vhf used 50 kHz spacing for a while. The 12.5 kHz FM is still pretty new and has not completely replaced older versions. It might be that it wasn't possible o have nbfm back then. $\endgroup$
    – AndrejaKo
    Commented Aug 20, 2017 at 21:08
  • 3
    $\begingroup$ As an aside, there is FM to be found at the top of the 10m amateur radio band, around 29.000-29.700 $\endgroup$
    – Scott Earle
    Commented Aug 24, 2017 at 12:54

3 Answers 3


So, the bandwidth argument is the dominant one. You need to incorporate the full Carson Bandwidth in your channel spacing, not only the frequency deviation! As explained in my answer to your previous question, that must be significantly larger than the bandwidth of the same audio signal in AM.

Aside from that, as always in Ham usage: historical reasons.

If you'd ask me, no sense in doing FM, either. There's digital voice modes that have a far superior spectral efficiency (that's a measure for how much info you can get across per second per bandwidth).

I'd recommend looking at DRM (Digital Radio Mondiale), a broadly adapted narrowband broadcast mode than can give you very clear voice over noisy 3 kHz channels. That's what India and Brazil are doing to replace their AM MW broadcast stations. Not "small scale bleeding edge". This is literally "state of the art for billions of people" tech; using FM to make more of the MW spectrum is half a step back, not a step forward.

If you need really narrow band communications (and that's what you'd want if you're working MW with a lot of people), then the digital voice modes of FreeDV are what I'd look into. 1.25 kHz channels with good voice communication, in any situation superior to AM, and not comparable to FM at all, because FM wouldn't even remotely allow voice communication at 1.25 kHz CBW. You don't need fancy hardware to use it.

So, FM was never an option on low frequencies simply due to bandwidth constraints, and digital modes are the logical step from AM.

I can only recommend to head over to the Codec2 site, and listen in – Codec2 is the audio codec used within FreeDV. Listen to the 1300 bps sample and ask yourself whether a 1.25 kHz AM channel would have offered a comparable quality (and before the "digital modes don't degrade gracefully" people show up: at SNRs where this works, you won't be hearing anything but static over AM. Also, FreeDV by default uses FEC, so yeah, it degrades graceful – by not degrading at all at first, and after that, the codec is still pretty robust against bit errors and does indeed degrade rather graceful, compared to AM, especially in the presence of burst interferers).

  • $\begingroup$ I'm asking this out of historic interest mainly, and also to better understand radio. Of course with current technology digital encoding is better for nearly anything. Nobody should be moving AM to FM nowadays, but I do wonder why nobody did 50 years ago. $\endgroup$
    – JanKanis
    Commented Aug 20, 2017 at 20:09
  • $\begingroup$ The first reason I thought about was also the bandwidth usage. But the Wikipedia article I link to says that FM with a peak deviation of 2 kHz can use an interchannel spacing of 8.33 kHz, which is less than the 9 or 10 kHz spacing used on the MW. That doesn't leave a lot of audio bandwidth, but AM sound is not so great either. So unless 2 kHz deviation FM gives worse quality than AM, that would have been a step up (before digital radio became viable). $\endgroup$
    – JanKanis
    Commented Aug 20, 2017 at 20:18
  • 1
    $\begingroup$ I've never heard of someone doing a $\Delta f$=2kHz, CBW=8.33 kHz FM, but sure, you could – at the expense of having but ~ 2kHz of audio bandwidth, which is really crappy. $\endgroup$ Commented Aug 20, 2017 at 20:33
  • $\begingroup$ @Marcus European CB radios use deviation of at most 2 kHz. I can' t remember the max audio frequency at the moment, but it fits nicely in the 10 kHz CB raster $\endgroup$
    – AndrejaKo
    Commented Aug 20, 2017 at 21:04
  • 1
    $\begingroup$ JanKanis To your question of 50 years ago - Before 1989 AM broadcasters could transmit audio frequencies up to 15 kHz. The older 15 kHz rule is a holdover from AM's mono music broadcasting days. Music on AM wasn't too bad! Stations weren't shoulder to shoulder like now. This was changed to 10.2 kHz to allow more stations in the 10 kHz slots. Technically you could do this 15 kHz with FM using a deviation small enough to match AM's bandwidth at a cost of S/N. To do the same with FM with broadcast quality S/N would require at least +/- 30 kHz. Simply put, you get more channels with AM than FM. $\endgroup$
    – JSH
    Commented Aug 25, 2017 at 15:00

The sidebands generated by frequency modulation are not as simple as those with AM, as their distribution is affected by both the amplitude & frequency of the modulating signal. A normal VHF station with a "max deviation" of 50 kHz will actually deviate 50kHz above & above the unmodulated carrier frequency, for a total bandwidth of 100kHz. In practice, a stereo FM station will have sideband components occupying out to near 200kHz in total.

In contrast, an AM Tx with a similar audio bandwidth will occupy a tad over 30kHz RF bandwidth. In Australia, MF AM stations normally transmit an audio bandwidth of 9kHz, amounting to an occupied RF bandwidth of 18kHz-----other countries use much more constrained bandwidth.

OK, why not reduce deviation so an FM signal fits into the same bandwidth as an AM station?

As you reduce the deviation of FM, two limiting factors start to occur. Narrow band FM cannot reproduce the same bandwidth audio as AM or VHF WBFM. Unfortunately, there is also a "double whammy", in that its noise rejection deteriorates.

If you increase the deviation to avoid these problems,you are back to "square one".

NBFM is used for communications work, where 'speech quality"audio is satisfactory.

Unintended long distance propagation of WBFM VHF broadcasts, or even analog VSB TV with its accompanying FM sound signal is not uncommon during Tropospheric ducting conditions. This can be 100s to 1000s of km.

Amateur Radio operators( "hams") using VHF & UHF NBFM radios more consistently make use of this phenomenon, usually during summer.

Hams do use FM below VHF, as does the old EU 27 MHz CB band. Long distance reception of both full sideband AM & NBFM via ionospheric propagation can suffer from "selective fading". This is because the various sidebands are refracted slightly differently by the ionosphere, so they are shifted in phase, & begin to cancel.

That said, under good ionospheric propagation conditions, a distant 29MHz NBFM signal sounds very much like a distant 146MHz NBFM signal received under good Tropospheric propagation conditions.

It is almost impossible to answer this question in a few paragraphs, as it is difficult to avoid introducing terms which require study in their own right.


Technically from a technological point of view any mode of transmission could be used on any frequency assuming that signal propagation is stable enough to insure reliable communication and bandwidth used was not an issue. Signal propagation on HF,MW and Lw is dependent on time of day with regard to atmospheric absorption and signal path whether skywave or groundwave .

Skywave means propagating upwards and being reflected down towards the Earth one or more times whereas groundwave propagation is line of sight. AM and SSB are less affected by skywave propagation but FM would be unreadable. If strictly using groundwave propagation then all modes would work. On shortwave FM is used on the 10 meter band.

FM is used on the VHF and UHF bands (the other modes are used as well) because propagation is pretty much line of sight which makes for a very stable signal which is required for FM to be readable.

One of the reasons why AM is noisier than FM is that impulse noise is generally amplitude modulated ( although some noise is also frequency modulated as well) and FM receivers are designed to detect frequency variations and reject or limit amplitude variations.

  • 1
    $\begingroup$ And by "FM would be unreadable [by skywave]", is that because frequencies are reflected differently by the ionosphere and FM uses a wider bandwidth/more frequencies? Or what happens? $\endgroup$
    – JanKanis
    Commented Aug 24, 2017 at 5:44
  • $\begingroup$ Has more to with way FM receivers decode the signal. I can't at the moment remember all the details (as it is very late-tired). Google FM receiver design and the principles of frequency modulation. $\endgroup$
    – Old_Fossil
    Commented Aug 24, 2017 at 6:47
  • 1
    $\begingroup$ Needing line of sight for FM is wrong. Ask your car radio. Do you constantly have line of sight to a broadcast tower when driving? Does radio break down every time you pass a house or a truck? $\endgroup$ Commented Aug 25, 2017 at 17:49
  • $\begingroup$ @MarcusMüller In North America, broadcast FM is 88-108 MHz and the AM BCB is 540-1700 kHz. Is that different than in Europe? $\endgroup$ Commented Aug 25, 2017 at 18:21
  • $\begingroup$ Broadcast FM transmitters are usually atop tall buildings and or mountains where the line of sight would easily overcome any small objects like trucks and houses. The is also multipath reception, that is reflected signal as well. Houses that made of wood would be no impediment to RF. $\endgroup$
    – Old_Fossil
    Commented Aug 26, 2017 at 6:11

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .