Re: Coherer experiments

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Dr. Duncan Cadd by way of Terry Fritz <twftesla-at-uswest-dot-net>" <dunckx-at-freeuk-dot-com>

Hi Antonio, All!

>Has someone ever experimented with a "coherer", or Branly's tube?
>It's a primitive radio detector, used in the early experiments
>about radio transmission. There are many references about it in the
>web, but I didn't find any modern reproduction.


Nice experiment.  There's a (sadly out of print) book called "In
Marconi's Footsteps Early Radio" by Peter R. Jensen, ISBN 0 86417 607
4  Kangaroo Press Pty Ltd 1994.  In this book the author shows how he
built repros of a number of old radio bits, including the coherer,
pages 88-92.  [If it was an old book I'd post gifs, but as it's recent
I'd better not or I will fall foul of copyright.]  Ideally, it should
be evacuated and the filings specified by Sir John Ambrose Fleming
(Fleming as in diode valve) should be 95% nickel and 5% silver, very
fine, and the metal plugs silver with their ends bevelled to make the
space between them wedge-shaped.  The tube was 5mm internal diameter
and 3-4cm long overall, the gap between the silver plugs being between
0,5-2mm.  (Description due to Fleming via Jensen.)  Obviously you
don't need many filings for so small a gap.  According to Sir William
Preece, the ends of the silver plugs were polished and amalgamated
slightly with mercury, the vacuum used being 4mmHg.

Originally, platinum wires were attached to the silver plugs for
sealing in crown glass, but this does not seal well to modern pyrex
glass (unless you use Apiezon wax in addition) and it would be better
to replace these with wires of composition suited to pyrex glass
sealing.  Also cheaper, no doubt!

The coherer was so named by Sir Oliver Lodge in 1894, after he had
done a lot of work on improving the tube of Prof. Edouard Branly,
1890.  But Branly was not the original discoverer.  The first recorded
work on this was done by Prof. David Hughes in 1878, but subsequently
forgotten and independently rediscovered later firstly by the Italian
Prof. Calzecchi-Onesti and then by Prof. Branly.

Peter Jensen quotes some modern work done on the coherer (but doesn't
say where he got the info - it could be V.J. Phillips "Early Radio
Wave Detectors" 1980 - no other bibliographic details available) as
finding that the breakdown voltage is so repeatable (15V in this
particular case) you could almost use it as a voltage reference, but
that no modern theory exists to explain its action.

FWIW, my guess would be that given there are many point contacts and
given that the voltage breakdown is so precisely defined, we're
looking at something analogous to zener reverse breakdown in silicon
diodes.  It would be nice to know at what voltage and how precisely
your coherer coheres.

Interesting device!

Dunckx