A possibly dumb question: I'd like to know if commercial AM radio stations broadcast a carrier tone along with signal sidebands. To ask my question is a different way; if I could measure (view) the spectrum of a transmitted commercial AM broadcast signal presumably I'd see the two symmetrical sidebands' spectral energy. But would I also see a single narrowband spectral component, a carrier tone, in the center of the two sidebands? (I've searched and searched the Internet trying to find out if commercial AM broadcast signals are suppressed-carrier AM signals. But I've found no clear answer.)

• – user65586 Nov 30 '15 at 16:37
• Crystal sets (and other historic receivers) wouldn't work if they didn't. – Brian Drummond Nov 30 '15 at 17:07
• Every AM trasnstitter broadcasts carrier and two side bands, else it wouldn't be AM, but DSB or SSB. – Marko Buršič Nov 30 '15 at 17:08

AM broadcast stations do exactly that: broadcast AM. Not single sideband, double sideband with suppressed carrier, or anything else. That means a carrier in the middle and a sideband on each side.

The fact that the carrier is a significant component of what is transmitted should also be obvious from a few seconds thought. By definition, AM changes the amplitude of a carrier to encode the baseband signal. At full or 100% modulation, the carrier amplitude varies between 0 and some maximum, with half that value being the average. Obviously there is significant carrier component in the result.

• Your use of the word "obvious" in your much appreciated answer deserves contemplation. I've been in the signal processing business long enough that I've become gun-shy of things that appear obvious. For example, if someone asked, "Is the sum of two periodic sinewaves itself periodic?" The obvious answer is, "Of course." But the correct answer is, "It depends." – Richard Lyons Dec 1 '15 at 0:47

This afternoon I obtained a 45-year old radio theory textbook and found a diagram of a class-C vacuum tube amplifier circuit. The author claimed that this circuit is used to generate broadcast AM radio signals.

The amplifier acts a nonlinear 'switching device' such that when the sum of the carrier, $f_c(t)$, plus the so-called "modulating" audio signal, $f_a(t)$, is applied to the amplifier the amplifier's output contains a theoretically infinite set of spectral components. But one of the spectral components is of the form:

$$R(t) = \left[1 + f_a(t)\right]\cos(2\pi f_c(t))$$

All the amplifier's output spectral components are then filtered out except $R(t)$ which is the final transmitted signal. So the '1' inside the brackets of the $R(t)$ expression answers my question and agrees with the comments and answer you folks graciously provided me.

• @ Tom Carpenter: It looks like the MathJax syntax that I use on the StackExchange Signal Processing web page doesn't work properly here. Tom, thanks for polishing my answer. – Richard Lyons Dec 1 '15 at 12:19