Well done on finding that.
This is a common problem in GA – just as it was on Vauxhall Vivas 50 years previously, whose electrical system build quality all of piston GA has inherited, with superbly waterproofed (not) connections 
Just in case anybody has the same problem I finally discovered the reason why the alternator light was permanently stuck on.
The inline fuse holder is actually located near the battery box and the soldered wire in the end of the fuse holder had oxidised and fallen out. The two ends that parted are marked by the red arrows Luckily the bare end with 12 volts on was lying on some aluminium near the battery box but had never shorted out to ground.
The break is marked on the circuit diagram with a red line below. The indicator lamp (for this lamp only) should actually be 6 volts but like on many Pipers a 12 volt lamp has been installed in the past by mistake. At present it glows only very weakly, if it was the correct 6 volt lamp an alternator failure would illuminate at normal brightness.
The reason why this lamp blows so often is that the “test” button puts + 12 volts on a 6 volt lamp. Best not to hold in the “test” button too long. It is a shame I can’t substitute the “OIL” “VAC” and “ALT” lamps for flashing LED’s – they would be a lot more obvious.
It is certainly true that a lamp will illuminate only when there is a voltage across it I am not taking the p1ss – that is how everything in electrics works. You need a voltage (potential) difference for a current to flow.
However " It is there to absorb the voltage from the ammeter shunt if the lamp fails or to absorb some of the voltage if the 5 amp inline fuse near the ammeter shunt blows." does not make any sense to me. A resistor is not going to “absorb a voltage”. I think Cobalt (#5) has got it right as to the purpose of the resistor. But maybe there is something else…
A resistor is unlikely to corrode to a lower value. Corrosion normally eats away material.
Thank you for the replies. I have not investigated yet as my Aircraft is away from home. On another forum I read the following
The alternator inop light works on the principle of opposing voltages or
potentials. The light is powered by a 5 amp breaker and at the same time by a
line from the ammeter shunt also protected by a 5 amp inline fuse. As long as
there is voltage on either side of the lamp the light is out. If the
alternator stops working the voltage from the breaker goes to ground and lights
the lamp. Use the press to test feature and see if the lamp glows brighter
while testing. There is a 30 ohm resistor going to ground in the Alternator
Inop light circuit after lamp and breaker. It is there to absorb the voltage
from the ammeter shunt if the lamp fails or to absorb some of the voltage if
the 5 amp inline fuse near the ammeter shunt blows.
So there are three likely possibilities. The 5 amp in line fuse is open or
there is a short circuit after the lamp or some other open circuit between the
ammeter shunt and the resistor. Some of the circuitry is also protected by
diodes to prevent voltage from lighting other lamps such as Vacuum Pump, or Oil
except when using the press to test button when all the annunciator lights
should light.
This opposing potential feature is pretty neat but a little unusual or at least
so in the way that Piper did it as there is no dedicated sensor like Beech and
some Twin Cessnas us
I’ll check out the inline fuse as wigglyamp suggests and for some reason I have a gut feeling that the the 30 ohm resistor might have broken down a little and started to short out. Let’s say the resistor has now partially burt out and is now 10 ohms? Would this explain why the indicator lamp gets slightly brighter when the TEST button has been pushed?
As part of the circuit diagram is missing for the OIL and VAC lamps I can be bothered to think about it anymore – I’ll get my pre-historic Avo 8 out – much more satisfying than using my plastic digital test meter and at least it make the plane’s circuits look modern!
I won’t argue with wigglyamp, he has guru status in all things electrical.
If the battery is remaining charged for multiple flights then a failed diode in the alternator isn’t the cause of the lamp illuminating. I’d go for the the blown fuse as suggested by buzz53.
I think I would be looking at the diode pack in the alternator, I have a very clever little test box from snap-on that will tell me if this is the problem in seconds.
How could a low battery cause the light to come on, since the alternator output is the same voltage as the battery, during normal operation?
We had same on PA28-161 on our club fleet this year and it was a u/s battery that kept causing it, new battery = problem went away
Peter wrote:
The bus at the top is labelled “starter and accessories” and “alternator field” at the same time.
These labels are for the circuit breakers below the bus. They forgot labeling buses in a single bus electrical system…
Peter wrote:
the alternator warning lamp is supposed to come on when the voltage difference between (a) the battery + terminal and (b) the alternator output, is high enough to illuminate it.
Yes. One side of the lamp is connected to the bus, and the other one to the alternator output via the ammeter.
But that side is also pulled down to ground through a resistor, so when the fuse (or ammeter) fails, so the lamp illuminates, but a bit weaker than if you use the test switch, which pulls it down to ground
