11    Modulation

Please look in your dictionary, ANY dictionary, and look up ‘Amplitude Modulation’. I am going to bet you any money you like that the description of AM reads something like this……….‘In AM the carrier wave is varied in amplitude in sympathy with the  modulating signal’ It says this in every dictionary that I can find and also in the 2005 edition of the Radio Communication
Handbook and shows the usual curvaceous oscilloscope waveform to prove it. This description is completely and utterly WRONG! Because, in fact, the carrier amplitude does not vary one jot whether there is any modulation or not! This can be proved mathematically. The usual picture of the oscilloscope screen, showing a 100% amplitude modulated carrier wave, happens because the carrier frequency and both sidebands are added together in the first amplifier of the oscilloscope. When using AM, a PA actually produces 3 signals, all at once, the antenna likewise, ………the carrier frequency, together with an upper and lower sideband. To better see what is happening, we really need to look at things on a spectrum analyser. This shows signal amplitudes vertically as usual, but with lower frequency signals displayed on the left of the screen and higher frequencies on the right. With an audio tone, of say 700Hz, fed into the mic socket, all you should see on the screen are 3 vertical lines. The lower sideband 700 Hz to the left of the carrier frequency, which is in the centre, with the upper sideband 700 Hz equally spaced on the right of it. The carrier should be at least 100% bigger than either sideband. If you now reduce or remove the modulating audio, the size of the carrier does not alter at all. However, if we reintroduce the tone signal again and increase the microphone gain too far, other
signal spikes appear each side of the carrier. These are usually caused by amplifier/mixer stages being over-driven in the Tx, producing harmonics and mixtures of these. So a very wide bandwidth of rubbish is being generated. And still, the carrier amplitude does not change! The important thing is that those unwanted signals are taking away power that was being used to generate the voice sidebands you really want to be heard at the receiving end. Objectionable sideband splatter is being produced. Yuk! On SSB, if we used just one modulating tone, all we would see would be a single RF sine-wave. We must use a 2 tone audio generator to modulate the Tx. Why? To get that signal mixing in the oscilloscope first amplifier again, so that we can see the ‘usual’ modulated waveform. Using two tones (which must not be harmonically related, e.g. 700 & 1900Hz) we then get the ‘usual’ modulated signal displayed. However, on a spectrum analyser, the two tones produce two signals on screen, on LSB to the left, or, on USB, to the right of where the carrier would have been, if it had not cancelled out in the balanced modulator. Now, if we increase the mic gain too far, or speak into the microphone too loudly or have too much speech compression on, unwanted audio harmonics appear to the left and right of the wanted signals. I have my K2 and FT857 rigs available to do tests. When either rig was starting to be overdriven, I could actually hear the audio harmonics being produced, just by listening carefully to the other
Rx. Because ham rigs have a ‘communications’ 300Hz to 3000Hz audio bandwidth, with the K2 crystal filter shorted out I was amazed at the bandwidth of rubbish being produced when the FT857 was over-driven. So gentlemen, I give you a challenge. Call your friendly neighbourhood ham who produces an S9+ signal in your Rx and really listen carefully to each other and adjust
the mic gain and processing controls until the bandwidth is contained within 3 kHz or so of your dial reading. Remember that receivers can produce the same effects if you overdrive the Rx RF stages, so keep the RF gain down. Use the widest bandwidth you have or short out the I.F. filter like I did, so that you can hear what’s really going on. Better still, make a 2 tone generator and,
as we all haven’t got a spectrum analyser, get your nearby ham to tell you if strong tones other than the two being injected are heard. If they’re there, then there is non-linearity somewhere &/or mic gain needs adjusting. We all owe it to other band users to keep our bandwidth within limits. You’ll learn a lot by building and using a 2 tone generator and carrying out tests which our licences require us to do! Get careful bandwidth reports from others or better, build a 2 tone generator and a single band direct conversion Rx, (in a screened box) so that you can you test your own transmissions sometimes? They’re all you need.
But an oscilloscope is useful too! You can always borrow one from the club, if you are a paid-up member.

That’s about it for this time. Vy 73. Mike G3JKX