Len Over 21 wrote:
In article , "Jim Hampton"
Hans Brakob came out of the blue with a personally insulting comment
about "bad research," something that I allegedly "always do," solely
about "the first voice transmission by radio." Hans claimed, but did
not reference or otherwise verify that it was done with a SPARK
TRANSMITTER.

I quoted an Electronics magazine special edition text paragraph that
stated this 1906 Christmas transmission was done with a 1 KW
ALTERNATOR, probably using a water cooled microphone in the
antenna line.

Just to clarify, this is what the article at

http://ewh.ieee.org/reg/7/millennium...fferences.html

says about it:

At the turn of the century Fessenden was using a spark transmitter,
employing a Wehnelt interrupter operating a Ruhmkorff induction
coil. In 1899 he noted, when the key was held down for a long dash,
that the peculiar wailing sound of the Wehnelt interrupter
could be clearly heard in the receiving telephone. He must have had
a detector of some sort that was working for him, even at this
early stage in the development of wireless. This suggested to him
that by using a spark rate well above voice band (10,000 sparks/sec),
wireless telephony could be achieved; and this he did transmitting
speech over a distance of 1.5 km on 23 December 1900, between 15 metre
masts on Cob Island, MD [Belrose, 1994a; 1994b].

In autumn of 1906 Fessenden had his HF alternator working adequately
on frequencies up to about 100 kHz. About midnight in
November, 1906 Mr. Stein at Fessenden's Brant Rock station was telling
the operator at a nearby test station at Plymouth, MA how to
run the HF alternator. It was usual for these two operators to use speech
over this short distance. However his voice was heard by Mr. Armor at
Machrihanish, Scotland with such clarity that there was no doubt about
the speaker, and the station log books confirmed the report

[...]

Fessenden's greatest success took place on Christmas Eve 1906, when he
and his colleagues presented the world's first wireless broadcast. The
transmission included a speech by Fessenden and selected music for
Christmas.

So, while the first radio voice broadcast was made at Christmas, 1906, using
an alternator, the first radio voice transmission of any sort was made 6 years
earlier using a spark transmitter.

The description of how the spark transmitter was used to transmit voice
makes perfect sense to me. If you take a band-limited continuous waveform
and transmit discrete samples of the waveform taken at a rate which is twice
the maximum continuous frequency (i.e. 10,000 amplitude-modulated sparks
per second for 5 kHz bandwidth audio), then low-pass filtering the received
signal back to the original bandwidth will reproduce the original continuous
waveform. This is from the Nyquist(-Shannon) sampling theorem, though if
I recall this was done a quarter century before Nyquist's paper on digital
(telegraph) signals over bandwidth-limited analog channels and a half-century
before Shannon explained what this meant for discrete sampling. That's
pretty neat.

Dennis Ferguson