It is always fun to read about history, especially regarding technology.
One thing I noticed is that the equation presented seems wrong. The A+B and A-B and pilot injection parts look right but multiplying by 75 kHz would give a double sideband suppressed carrier signal which is what the A-B signal is but not centered at 75 kHz and not including the A+B and pilot signal. Maybe I am missing something here.
Neil
Great Read. Thanks for that.
I worked at an AM station in the mid-60's when they launched a simulcast FM signal in mono. The owner of the station at the time said FM Stereo was nothing more than a fad and it wouldn't last, so the transmitter, all the consoles and everything else (all new) were mono only. However, it didn't take long for him to realize that was a mistake and re-tooled everything, including the transmitter to stereo.
I am guessing 150 kHz at the deviation is 75 kHz from each side of center?
I always remember being told as a rough rule of thumb, fm bandwith is approx 75khz deviation x2, plus 15khz audio x2, both added makes 180 khz.
Paul.
I believe you are referring to Carson's rule (Google it).
The proper way to calculate FM bandwidth involves the modulation index (contains both deviation and audio frequency)and the audio frequency and then use Bessel Functions which give the amplitude of each of the sidebands. Technically, the sidebands extend to infinity but the usual practice is to define the occupied bandwidth as those sidebands summed which contain 99% of the total power. Kinda contorted, isn't it, so Carson's rule is not bad at all.
Neil
Yes i do know and have used the bessel null method with accurate sinewave tone and spectum analyser to look at amplitude null of relevant sidebands for accurate measurement
For quick peak deviation analysis i use the excellent Pira P75 analyser, very accurate !
I do find the basic crt oscilloscope an essential tool, excellent for am modulation, and the trained eye can spot the negative mod level easily under dynamic audio conditions too.
Paul.