This has been an interesting thread. But there seems to be a
potentially major problem with ELTs that depend upon a cable and an
external antenna - they may not work after a crash!

First, consider the cable. I make my living using coaxial cables with
BNC (and other) types of connectors, and I can't count the times that
those connectors have failed. True, some were due to poor assembly, but
others (that had been tested after assembly) failed during non-abusive,
normal service. Considering the magnitude of the g-forces that could be
generated during a crash and the likely fragmentation of the fuselage,
how likely is it that the antenna will be electrically connected to the
ELT afterwards?

Second, consider the antenna. There was reference in this thread to a
whip antenna whose instructions specified that it needed to be mounted
within 20 degrees of vertical. There was also reference to mounting
antennas in carbon fiber gliders near the canopy or under a small
fiberglass area in the fuselage. How likely is it in these cases that a
wrecked glider will come to rest with the antenna properly oriented for
transmitting?

It seems strange to have the device itself meet a whole raft of specs
under a TSO (even going to the lengths of specifying the type of
battery) when after a crash the signal might not be able to leave the
ELT. Have the regulatory bodies (FAA, JAA, CAA, etc.) formally
considered this problem, or is it just left to the installer to assess
operational survival probabilities for the ELT system as a whole?

-John