[TR] Electrical diagnosis: Update

[Randall]

> In the meantime, I'll be driving myself nuts trying 
> to figure out 
> if a  battery can fail suddenly, while the engine is being 
> cranked and 
> suddenly the  ignition light goes out and the engine stops 
> cranking, and also 
> even if that can  happen 

Although certainly not common; it has happened to me.  The battery had a
broken conductor inside the case, probably from over-enthusiastic tightening
of a clamp.  All normal tests would show that the battery was working fine
(which it was), but then when the engine was cranked, it would heat up the
break and cause it to quit conducting entirely.  Then any sort of fiddling
with the battery (like removing it for testing or tightening a clamp) would
rub the broken ends together and cause them to start conducting again.  The
effect was a "perfectly good" battery that would not start the car!

An even weirder problem happened to me on a non-Triumph.  On that car, the
battery ground cable ran to both the body and the engine (so it doubled as
the body/engine ground strap).  The stud on the engine was horizontal, and
the nut had worked loose, so the cable end was just hanging on the stud.
When cranking, the cable end would micro-weld to the stud and pass enough
current for the starter.  Immediately after starting, the voltmeter would
indicate that the battery was being charged, but soon engine vibration would
break the micro-weld and the indicated voltage would drop back to battery
voltage (even though the alternator was putting out 14+ volts).

My point in all this is to get the problem to happen, and make the voltage
checks with everything connected and undisturbed.  Just blindly replacing
components can drive you crazy!

Randall