Other than putting all ELT's in the bin, as some Aussies seem to suggest, maybe one could return the ELT to the condition of entirely isolated from all of the aircraft elecrical systems.
Folks,
This is NOT what Australian regulations are suggesting. Indeed, Australian regulations require most aircraft, on all but very short flights, (some Ultralights are exempt) to carry an ELT.
What is NOT required is a fixed ELT, for the reasons so comprehensively demonstrated by REAL WORLD results, not theoretical claims about design performance real crash loads, not "designer" crashes.
What is the point of carrying a device that has a failure rate of around 95% (100% in water) AND THIS RESULT INCLUDES LARGE AIRLINE AIRCRAFT not just small GA aircraft.
Quite a number of posters here discuss the merits of a fixed ELT in a way that suggests that the device is actually going to perform some useful function, in the event of a crash. The figures are very clear, the likelihood of any useful broadcast is extremely remote.
In the case of the recent crash at KSFO, I think I am on safe ground in saying that the fixed ELT there in the B-777 (if one was fitted) would not have broadcast a useable signal.
The failure reasons are really quite simple, in the crash sequence, in most cases, either or both the aerial cable or the external aerial are damaged. Although less common, failure of the G switch to activate have been recorded, in circumstance where the calculated G load were such that the system should have worked.
The bottom line here is quite clear, as far as any safety contribution is concerned, fixed ELT are a waste of money and add an extra potential failure mode with on-board fire as a result.
Portable ELTs, on the other hand, have a well established record of producing some quite amazing results, even under the old pre-GPS 121.5/243 COSPAR/SARSAT system, in leading searches straight to the wreck, without even a box search being necessary.
The US Civil Air Patrol, as a result of the Australian research, did a study and found, within the limits of the available data, substantially similar result -- which should surprise no one who has ever spent any time at an aircraft accident site.
None of the airline aircraft I ever flew were fitted with fixed ELT and that never stopped them flying in FAA airspace they were all fitted with with multiple portable ELT, in many cases with a water activated power source, so if it was a survivable land accident, a few blokes were going to have to pee in the plastic bag tied to the ELT for just that purpose. In a ditching, no shortage of "activating fluid".
I have actually had a commercial Li AA size and a 9V NiCad battery have a thermal runaway, in each case the manufacturer said it was impossible. Any exothermic chemical reaction which runs away caused by who knows what ? Impurities acting as an unintended catalyst ? The 9V case was interesting, the multimeter case felt warm, I removed the battery, and it continued to get hotter in my hand until I could no longer hold it so I put it on a bench and watched it finally burnt it all took about 15 minutes from the time the battery was removed from any potential short circuit. A short was unlikely, as the multimeter worked OK when another battery was fitted.