Why is frequency modulation (FM) not used on the current amplitude modulation (AM) radio stations? The Police have FM 2 way radios that use only 5 kilocycles of modulation. Instead of the current AM radio stations that use the same bandwidth of 5 kilocycles. The limiter & discriminator or ratio detector filter out the static & loud pops that are on AM; Thus the FM would have much less static & noise than AM. FM channels would occupy the same bandwidth as AM channels. For example if WABC in New York used frequency modulation over the current frequency of 770 they would need to have the oscillator changed but the power amp would stay the same. But the receiving radios would need to be replaced with a radio to demodulate frequency modulation at 770.
- Anybody can ask a question
- Anybody can answer
- The best answers are voted up and rise to the top
I'll list a couple of main reasons:
First reason is compatibility! You'd have to leave behind a few generations worth of AM receivers, all of which would suddenly be unuseable. Then there's also equipment at existing stations that needs to be changed, which requires more investments. Furthermore, will listen to your FM radio station in the AM band? There's also regulatory angle that needs to be taken into account.
Second, even the analog FM itself is outdated. Today, we're moving towards the digital systems and in the short-wave region, the main systems are Digital Radio Mondiale and HD Radio. Both can provide in certain configurations performance as good as broadcast FM and can be used in configurations which are compatible with already existing AM equipment. There are similar solutions for VHF radio as well, which are intended to replace the broadcast FM stations.
Next, there's the claim of Police two-way FM radios which use 5 kHz bandwidth. I'd really like a source for that claim! Most FM professional radios do use today 2,5 kHz or 5 kHz of FM deviation, but deviation is not FM bandwidth. A popular method of determining FM bandwidth is the Carson's rule, which states that the majority of sideband energy is within bandwidth of $B=2( \Delta f +f_m)$, where $\Delta f$ is the peak deviation and the $f_m$ is the maximum frequency transferred. Common channel spacings for two-way radios are 12,5 kHz and 25 kHz. They both provide good enough quality to understand words, but are insufficient to transfer music. When using 5 kHz deviation, traditionally, highest frequency transferred is 3 kHz, which gives us 16 kHz occupied bandwidth plus a bit of extra space for co-channel interference. In the 12,5 kHz case, 2,5 kHz deviation is used and the highest transferable frequency is also around 2,5 kHz, which gives us 10 kHz of bandwidth and a bit extra for co-channel interference.
So "narrow" FM does give us comparable bandwidth to AM, but maximum useable frequency is much lower. I don't have a good reference for AM broadcasting at the moment, but if I remember correctly, the maximum frequency that can be transferred in medium-wave AM is a bit less than 5 kHz. That's twice as good as narrow FM and still has a bit lower bandwidth.
I forgot the power-amp part that was mentioned in the question.... There is a difference between power amplifiers for AM and FM broadcasting. AM output power is (lot's of hand-waving here!) proportional to input power of the signal, so louder input will provide louder output. On the other hand FM has (again lot's of hand-waving here) power output that's closer to constant and doesn't depend as much in the input signal. This is important, because for same peak output power, FM transmitter's average output will me much closer to peak than for AM transmitter. This requires that FM transmitters be much more conservatively rated.