McMURDO SILVER was the author of that CQ magazine article? I thought McMURDO SILVER was a brand of old time radios.
TIL (That I Learned): McMURDO SILVER was not just a brand of old time radios but was actually the name of the company founder. See https://www.worthpoint.com/dictionary/p/tools/manufacturers-vintage/mcmurdo-silver-radios-manufacturer

That article claims that 50% modulation is only 3 dB down from 100%. It is not stated where Mr Carr, the pre-war CE of WTHT got the Fig. 1 graph. The one in Fig. 2 is correct.

Modulator power is a function of the SQUARE of the percentage of modulation. The modulator power required to achieve 50% modulation is only 25% of that required to achieve 100%. This means only 25% of the sideband power of a signal modulated 100%. 25% power is 6 dB down, not 3 dB. Those two graphs do not line up; it is incredulous that (sideband) output power would not be directly proportional to the modulator power generated to produce it. Someone please let me know if I have missed something here.

The article is correct in that the average percentage of modulation with the typical human voice is only about 25% (actually, more like 30%) when the positive peaks reach 100%. IOW, the average sideband power of unprocessed voice modulation is about 10 dB down from peaks of 100%. The same also applies to a SSB signal. With no "processing", a SSB signal peaking at the magic 1500 watts is typically radiating only about 150 watts average power, hence the near-universal use of processing in SSB rigs. This is why we need compression and peak limiting with AM as well, and why nearly all AM broadcast stations use it. This also explains why it is advantageous to have some positive peak head-room capability in the transmitter beyond 100%.

A further note, it is not "illegal" to modulate an amateur transmitter beyond 100% positive. The rule limiting modulation to 100% was eliminated decades ago with the final R & O of Docket 20777. It always was a grey area regarding positive peaks beyond 100%; the FCC enforced it only for negative peaks, as long as the positive peaks remained clean and un-clipped. Technically speaking, any output from a SSB supressed-carrier signal would have been illegal, since there is no carrier to begin with.

Along with the rule on 100% modulation, the FCC also deleted the prohibition against "simultaneous" amplitude and frequency modulation, which explains why Timtron (WA1HLR)'s SBE rig (the SlopBucket Eliminator), a high level modulated oscillator, is now legal (another grey area).

Don k4kyv

I thought it was an interesting article Kevin WA6JKN had tossed my way. I was trying to drum up some conversation. Some of it seems a little off from what I know because I always have gone by the 100% modulation rule is what to strive for in general for a proper AM signal.