Dave Saul said in 030910 MMDigest:

> The carbon microphone element is clean, predictable, and capable of
> very fast response.  Placing such a device, or a simple variant of it,
> under each key or at some critical point in a piano action, could, in
> principle, enable the measurement of intensities with which a key was
> struck.

The carbon granule microphone of 1904 was very far from being
predictable.  The granules regularly packed up close together and
required the unit to be given a thump with the fist.  Also -- and
especially germane to piano recording circumstances -- the fixed and
moving carbon block elements had to be mounted vertically.  If they
were horizontal the amount of granules had to be increased to maintain
contact with the upper element and this reduced the sensitivity.

Telephone handsets where the mike could be any way round didn't
appear until about 1906, and then only in office phone systems where
poor performance wasn't critical.  In our radio club in the early
1950s, I somehow obtained most of an office system dating from 1918 and
fired it up and you often had to swing your head wildly about to get
the handsets to crackle into any kind of communication.

Even the mass-produced carbon inserts of the 1950s weren't consistent
in the way you'd need to get a repeatable output for an analogue
transducer like a magnetically-driven pen or electro-stylus.  I was
building an automatic phone exchange (a "central [office]" in US-ese)
and we needed optimum performance to overcome the very high resistance
of our very thin ex-military line wires and we ran resistance trials on
several different mike inserts.  Not only were they all different,
their resistance at rest was different after every movement of the
diaphragm.

> From Welte's perspective, the downside of this was the fact that the
> telephone people controlled and jealously guarded patents on carbon
> microphones and related devices.  It's well known that there was a
> lot of highly charged patent litigation in this area; in the USA, for
> example, between Bell and Western Union.  This might tend to explain
> the extreme secrecy surrounding the Welte recording process.
>
> It might also explain why so many unlikely (and often technically
> unsound) schemes involving carbon rods, pointed wires, and troughs of
> mercury were perhaps "leaked" intentionally into the rumor mill to
> protect legally risky technology from all external scrutiny.

As I said, Hughes's carbon rod transmitter was abandoned after about
1900 and it's unlikely anyone would have tried to protect it.  In our
club we had three old phones from the 1890s, one of which had a Hughes
transmitter.  It gave fainter output than the others, but its
electrical characteristics were unadventurous.  The rest resistance was
always the same.  In lieu of any further suggestion I would certainly
back this device as the prototype for the Welte detector.

Those old phones were already very faint by the 1950s and would be
fainter now.  The reason is that permanent magnets were not (indeed,
still are not) permanent.  The big U-shaped magnets in those old Bell
receivers had to be juiced up every 30 years.  The book said we needed
9000 ampere-turns.  This meant wangling three or four turns of copper
strip around the arms and then short-circuiting a car battery with
them.  We only tried it the once ...!

Dan Wilson, London