Narrowband FM vs Wideband FM on VHF and UHF: Deviation and Bandwidth

Almost every handheld and mobile radio for the VHF and UHF bands has a menu item with two letters: W and N, or wide and narrow. Many operators leave it at the factory setting and never ask what it does. Yet this setting decides how much space your signal uses on the band, how loud your audio is at the other end and whether your voice sounds clean or distorted on a repeater. Narrowband FM is the way in which most commercial radio systems now work, and it is spreading in the amateur bands as the channel plans are tightened. This guide explains deviation, bandwidth and the channel widths in plain terms, shows how to calculate the space that a signal occupies with Carson’s rule and describes what happens when a wide radio talks to a narrow one. It ends with the practical settings that give a good result on a normal contact.

What Deviation Is

In frequency modulation the voice changes the frequency of the carrier, and not its amplitude. The carrier moves up and down around its centre frequency in time with the audio, and the largest excursion from the centre is called the deviation. A louder voice causes a larger deviation, up to a limit. The ratio between the deviation and the highest audio frequency is the modulation index. The Wikipedia article on frequency modulation gives the mathematics, and for practical use it is enough to remember two facts. A larger deviation gives a louder audio signal at the receiver, because the receiver converts the frequency swing into voltage. And a larger deviation uses more bandwidth, because the sidebands spread further from the centre.

The deviation is limited by the channel width. A 25 kHz channel allows the signal to swing by up to 5 kHz on each side of the centre, and a 12.5 kHz channel allows 2.5 kHz. Broadcast FM on 88 to 108 MHz uses 75 kHz, which gives the high fidelity of music but needs a channel of 200 kHz. A voice signal does not need that much, and for speech the narrower setting gives a quality that is good enough with a better use of the spectrum.

The Bandwidth of an FM Signal: Carson’s Rule

The occupied bandwidth of an FM signal can be estimated with Carson’s rule: the bandwidth is about twice the sum of the peak deviation and the highest modulating frequency. For a voice limited to 3 kHz and a deviation of 5 kHz the result is 2 times 8 kHz, or 16 kHz. With a deviation of 2.5 kHz the same voice needs 2 times 5.5 kHz, or 11 kHz. The first signal fits comfortably into a 25 kHz channel and the second fits into 12.5 kHz with a small guard band. The table shows the usual values.

The rule is an estimate. A transmitter that has an audio filter and a limiter keeps the real signal inside the channel, but one that is overdriven spreads energy to the neighbours. This is the reason why the audio level and the deviation are controlled in every FM transmitter by a limiter circuit, and why a shouting operator may sound distorted even when the radio is working correctly.

Why the Industry Moved to Narrowband

The professional services have moved over the last decades from 25 kHz to 12.5 kHz channels, and some to 6.25 kHz equivalents, because the number of users has grown while the spectrum has not. Halving the channel width doubles the number of channels in the same band. The amateur bands have followed in part, and the band plans in many regions now describe 12.5 kHz channels for simplex and repeaters on the VHF and UHF bands. The details depend on the region and the country, so it is best to consult the plan of your national society. Our article on the IARU band plans explains how the regions differ.

The digital voice systems that are popular with amateurs are based on the same idea. DMR carries two time slots in a 12.5 kHz channel, C4FM in System Fusion uses 12.5 kHz and D-STAR occupies about 6.25 kHz. The comparison in our article on DMR, D-STAR and System Fusion gives the details. The result is that a narrow channel is a normal thing today, and an analog FM radio set to wide on a narrow channel is the odd one out.

What Happens When Wide and Narrow Meet

The most common practical problem is a mismatch between the transmitter and the receiver. There are two cases.

• Wide transmitter, narrow receiver. The receiver has an IF filter and a discriminator that are made for a deviation of 2.5 kHz. A signal with 5 kHz of deviation is partly cut by the filter, and the discriminator is driven past its linear range. The audio is distorted and sounds harsh or clipped, especially on loud syllables. The effect is worse when the transmitter is strong and the voice is loud.
• Narrow transmitter, wide receiver. The receiver passes the signal without problems, but the deviation is only half of the expected value, so the audio is about 6 dB lower. The voice sounds quiet, and you may need to increase the volume. The noise level stays the same, so the signal to noise ratio of the recovered audio is worse by the same 6 dB.

Neither case stops a contact, and operators do talk across the mismatch every day. They simply lose some quality. If you want the best audio, you should match the settings of both ends and use the width that is common on your local channel, normally the narrow setting on the busy repeaters and the wide setting where the older equipment is still in use. Ask the other station about your audio and adjust accordingly.

Repeaters, Tones and the Deviation Budget

A repeater system adds another factor to the deviation. The sub-audible tone that opens the repeater, described in our guide to CTCSS tones and DCS codes, uses a part of the available deviation, typically around 15 percent. The rest is left for the voice. If a transmitter has its deviation set so that the voice alone reaches the limit, the addition of the tone can push the signal beyond the channel, and the repeater receiver may hear a distorted signal. Properly adjusted radios allow for this, and the operator only needs to avoid shouting into the microphone. Repeater owners, in the same way, set the output deviation of the repeater so that the channel is used evenly, and a transmitter that is set too low or too high will be noticed by every listener.

The same principle applies to the use of data modes on FM. Packet radio and APRS send audio tones through an FM radio, and the deviation of these tones decides whether the receiving modem can decode them. A tone that is too loud is clipped, and a tone that is too weak is lost in the noise. A level that gives a moderate deviation, a little below the maximum, is the best compromise. Our overview of APRS shows where the narrow or wide setting matters for the network.

Receiver Selectivity and the Neighbouring Channel

A narrow channel plan asks more of the receiver. The IF filter of a narrowband receiver must reject the neighbouring channel only 12.5 kHz away, and a transmitter that is slightly off frequency, or one that splatters because of over-deviation, becomes a problem. The same physics explains why a cheap receiver with a wide filter hears two stations at once on adjacent channels. In a handheld with a selectable width, the narrow setting also reduces the noise that enters the receiver, because the filter admits a smaller band, and the improvement can be about 3 dB in the noise power. The capture effect of FM adds to this: the stronger of two signals on the same channel is demodulated and the weaker one is suppressed, with a difference of only a few decibels needed.

Hearing and Measuring the Difference

You can learn to recognise a mismatch by ear. A narrow signal on a wide receiver has the same sound as a station that is a little too far from the microphone: the voice is clear but weak, and the background hiss rises when you turn up the volume to compensate. A wide signal on a narrow receiver sounds like a loud, slightly rough voice that breaks up on peaks, much as a speaker driven beyond its limit. If you have a second receiver or a software defined radio, you can look at the signal on a spectrum display and see the width directly. A 12.5 kHz signal shows a block that is narrower than the neighbouring wide stations, and a distorted, over-deviated signal shows spikes at the edges that reach into the adjacent channels.

A more careful measurement uses a deviation meter or a service monitor, and many clubs have one. Whistle or play a steady tone into the microphone at a moderate level, and read the deviation. For a narrow channel the peak should stay below 2.5 kHz, and for a wide channel below 5 kHz. If your radio has a menu for the microphone gain, adjust it so that normal speech reaches about 70 to 80 percent of the limit on the peaks. This leaves a little room for the tone of a repeater and for the loud syllables, and it avoids the clipping that makes the audio harsh.

Practical Settings

• Use the width of the local channel plan. On a repeater, set what the repeater owner recommends; on simplex, match your partner.
• Do not overdrive the microphone. Speak at a normal level and a few centimetres from the microphone, and avoid shouting.
• Test your audio. Ask another station for a report in wide and narrow, and keep the setting that sounds better.
• Check the receive width. If the received audio is distorted, switch the receiver to the wide setting; if it is quiet and noisy, try narrow on the transmit side of the other station.
• Leave the programmed channels alone. A good programming tool lets you set the width for every memory individually.
• Keep the frequency accurate. A cheap radio that is a few kilohertz off frequency can cut into the narrow channel next door.

If you load memory channels from a computer, the programming software stores the width as one of the fields of each channel. It is worth a few minutes to set it correctly on the first day. Our article on VHF and UHF simplex and repeater basics gives the other settings that go with it.

The Bottom Line

Narrowband FM uses a peak deviation of about 2.5 kHz in a 12.5 kHz channel, while wideband voice FM uses 5 kHz in 25 kHz. Carson’s rule gives a bandwidth of about 11 kHz for the first and 16 kHz for the second. The narrow system doubles the number of channels, and most professional and digital systems already work in this way. A wide transmitter into a narrow receiver sounds distorted, and a narrow transmitter into a wide receiver sounds quiet by about 6 dB. Match the width to the channel, keep the microphone level moderate and ask for a report if the audio is in doubt.

Frequently Asked Questions

What does narrowband FM mean?

It is FM voice with a smaller peak deviation, usually 2.5 kHz, so the signal fits into a 12.5 kHz channel. Wideband voice FM uses about 5 kHz of deviation in a 25 kHz channel.

Which setting should I use on my handheld?

Use the width of the channel plan you operate on. On many modern repeaters this is narrow, but check with the repeater owner or the local club.

Will a wide radio hear a narrow one?

Yes, but the audio will be quieter by about 6 dB because of the lower deviation, and the signal to noise ratio of the audio is worse.

Why does my voice sound distorted on the repeater?

You may be over-deviating by shouting or by using too high a microphone gain, or the repeater may be set to narrow while your radio transmits wide.

What is Carson’s rule?

It is an estimate of the bandwidth of an FM signal: twice the sum of the peak deviation and the highest modulating frequency.

Is narrowband FM better than wideband?

It is more economical with spectrum, but the audio is a little quieter and less rich. For speech the difference is small, and the narrow channel is the standard for new systems.