Leaderboard

You get more fidelity at a given power level using a wider deviation. You need more power (a net change of 3dB in this case) to reproduce a transmitted signal with the same fidelity. So if you've got a receiver trying to receive your buddy down in the valley and you're in a theoretically perfect hilltop site, you'll have -130dBm of thermal noise coming in. If everything is wideband, your buddy needs to be received with some signal-to-noise ratio to achieve some level of fidelity. Let's arbitrarily set that threshold -122dBm to achieve a good-enough signal. If your buddy switches to a narrowband radio and you switch the repeater to a narrowband configuration, you'll have a narrower IF filter and thus less thermal noise within the IF passband: -133dBm. Passing the input signal at the same signal-to-noise ratio (which would be received at -125dBm) will not reproduce the input signal at the same degree of fidelity -- it'll sound worse. You're trying to do the same thing with less RF, it just doesn't work like that, or we'd have been 6.25 kHz compliant for a long time. The signal-to-noise ratio needs to improve by 6dB each time the bandwidth is halved to produce the same fidelity, and we gained 3dB from the narrower IF filter. To get the same fidelity, we'd need a received signal at -118dBm. The easy way out of the problem is to lowpass the audio so less noise is audible, and that's one of the reasons you'll hear a lot of people say narrowband sounds like junk. The audio loses fidelity in the process, but has less high-pitched hiss from FM noise. The problem is with reproducing the transmitted frequency range. More and more RF power is needed to allow the higher pitches to be demodulated with an audio SNR above a fixed threshold as IF bandwidth goes down. It's a property of FM.

Wideband has a slight advantage (3dB) over narrowband when it comes to sensitivity. Cutting the bandwidth in half requires 6dB more power to get the same signal-to-noise ratio, but 3dB comes from having the noise power on the narrower receive IF filter. On business radio systems, this may not matter; but on GMRS, the users tend to be dispersed further as operating area isn't constrained by the license and minimal frequency coordination takes place. This effect also causes mobile flutter to be more pronounced. On modern commercial radios designed for narrowband use, the strong-signal voice quality doesn't take much of a hit when moving to narrowband channels. We can't get any more pairs because FRS is already authorized for the 467 MHz interstitials. We'd need to go to 6.25 kHz ultra-narrowbanding (NXDN48 or dPMR) and use some odd channels (462.546875, 462.553125, 462.559375, ...), which won't happen anyways because digital voice isn't allowed. Adjacent ultra-narrowband channels would also interfere unless frequency accuracy is well-controlled, which raises equipment cost considerably (particularly for portables) and requires realignment during the equipment's service life. Additionally, wideband users would take interference from any of four ultra-narrowband channels, and narrowband users would take interference from any of two ultra-narrowband channels. The end result is even less voice capacity than before, unless everyone goes ultra-narrowband and maintains their equipment to high standards. The repeater operator always has the option to narrowband should they want to, but there's no benefit unless there's significant adjacent-channel interference, all of the radios are properly configured for narrowband (travelers are probably not), and the radios used on the system have proper 12.5 kHz IF filters (the GMRS-V1 does not). A narrowband mandate would also screw up equipment certifications and cause the FCC to get a big headache over something that is not really a problem at all.

We're looking at this from a different perspective. Hoping to do a blended-function setup with analog voice and SMS via GMRS from a single, non-repeater mobile station to a set of HT receivers in the field within a 1 km radius. Still catching up on the learning curve on this.