Why do major radio stations broadcast on odd frequencies?

[Eli]

Originally posted by: JustAnAverageGuy

Originally posted by: beer
Every realizable signal is time limited, therefore it has infinite bandwidth. All communications signals exist on all frequencies, but they are greatly, greatly attenuated depending on the design of the transmitter. Therefore, it comes down to filter design to attenuate the signal at the appropriate frequencies. I believe that with FM in the US, the FCC defines 200 KHz of bandwidth for each station, but the bandwidth is not 'true' bandwidth (which is infinite) but '99% signal power bandwidth', that is to say, the integral of the frequency response curve centered at X MHz, symmetric about each side with integral bounds of 100 KHz +/- the centered frequency would be 99% of the true infinite bandwidth of the signal. Therefore, if you tuned into 90.0 MHz, you would be halfway in the signal spectrum of 89.9 and 90.1, which would be garbage to a human ear.

I barely understood anything you just said, but it was long and used alot of big words so you're probably right.

LOL, my thoughts exactly.

 

[MSCoder610]

Originally posted by: Syringer
How come there's no channel 1?

Assuming you are talking about TV...
http://members.aol.com/jeff560/tvch1.html

 

[beer]

Originally posted by: RaynorWolfcastle

Originally posted by: beer
Come on, speaking of theoretical, there can be no ideal filter that gives 100% attentuation anyways. So yes, the 'theoretical' aspect of infinite bandwidth is there, but so is your ideal LP filter that smooths off exactly at 45 KHz. It doesn't exist. And for what it's worth, you should know that the modulation scheme has nothing to do with the receiver architecture. Signals are modulated onto a certain carrier wave because of the channel's properties. The FCC used the frequency range of 87-108 MHz because it has a fairly direct line of sight, and nothing else. It allowed them to densly pack FM transmitters, instead of AM transmitters that at night can interfere hundreds of miles away due to ionosphere reflection.

A) You're right there are no filters that can give you 100% attenuation, but if your filter attenuates your signal so that its stopband power is below channel noise then that really doesn't matter.

B) My side note was a general musing and not related to any of your comments specifically. I don't know why you feel I need to be informed about receiver architecture since I didn't say anything about it in my post.

C) As you surely know, the modulation scheme doesn't have much to do with the spectrum allocated to a given technology but has a lot to do with the spacing between carrier frequencies. This has a lot to do with how good your receiver is (though not its architecture per say) and how complex you're willing to make it.

We're in agreement. Your channel noise is surely higher than that of either '99% power' or '35 dB attentuation' definitions of bandwidth, too. It was merely presented as a formality. For what it's worth, you wouldn't have to lowpass it anyways, it's an analog signal and voice is naturall lowpass. You're not lowpassing to avoid alaising, so in this case it is unnecessary too.

 

[EGGO]

Totally forgot how that's like now that I have satellite radio.

 

[sharkeeper]

Channel 1 picks up the Taco Bell take out order system.

 

[RaynorWolfcastle]

Originally posted by: beer
We're in agreement. Your channel noise is surely higher than that of either '99% power' or '35 dB attentuation' definitions of bandwidth, too. It was merely presented as a formality. For what it's worth, you wouldn't have to lowpass it anyways, it's an analog signal and voice is naturall lowpass. You're not lowpassing to avoid alaising, so in this case it is unnecessary too.

While voice is bandlimited (to much less than 20 KHz) instruments are not. A cymbal crash for example will have a wide spectral content so it needs to be lowpassed so that you don't contaminate neighboring channels. This is all pretty irrelevant if your source is a CD since that is all "pre-lowpassed" but if you have a DVD-A or a live band as a source it is a problem. You also lowpass to minimize the baseband noise that you're sending to your carrier frequency.
🙂

 

[aplefka]

Because you touch yourself at night.

 

[VirtualLarry]

Originally posted by: Colt45
odd in north america, even in europe, not sure about the rest of earth..

Well, Europe has phased out analog radio, for the most part, in favor of terrestrial digital broadcasts, haven't they?

 

[mercanucaribe]

Originally posted by: beer
Every realizable signal is time limited, therefore it has infinite bandwidth. All communications signals exist on all frequencies, but they are greatly, greatly attenuated depending on the design of the transmitter. Therefore, it comes down to filter design to attenuate the signal at the appropriate frequencies. I believe that with FM in the US, the FCC defines 200 KHz of bandwidth for each station, but the bandwidth is not 'true' bandwidth (which is infinite) but '99% signal power bandwidth', that is to say, the integral of the frequency response curve centered at X MHz, symmetric about each side with integral bounds of 100 KHz +/- the centered frequency would be 99% of the true infinite bandwidth of the signal. Therefore, if you tuned into 90.0 MHz, you would be halfway in the signal spectrum of 89.9 and 90.1, which would be garbage to a human ear.

That's not what he asked. He asked why they are centered on the odd digits and not the even ones.

 

[beer]

Originally posted by: mercanucaribe

That's not what he asked. He asked why they are centered on the odd digits and not the even ones.

If the FCC spectrum spans 87 MHz to 108 MHz, and each interval requires 200 KHz of bandwidth, then then it must be centered at 1, 3, 5, 7, 9, and rolloff gradually into 0-2, 2-4, 4-6, 6-8, etc.

It is important to explain why the spacing exists, and why you can't rolloff more immediately and not need the 200 Khz bandwidth.