I still don’t buy the shaft shearing 
I would accept that any “low rpm” (starting or shutdown isn’t really relevant) is not going to do things much good if you have a massive load on and the system is marginal. Take the well known Cessna 400/TTX anti-ice alternator and all the problems with that. It was always a marginal setup (in terms of pulley shaft diameter, and belt dimensions) and perhaps some people had the system on at low revs.
Otherwise, modern planes should not draw so much. I have an 80A alternator but my plane draws ~27A with everything on, plus more with gear operating but one would normally not be drawing the full gear pump current from the alternator only.
Peter wrote:
About 3x more due to the belt ratio.
That would reduce the torque by a factor of 3 – so even a 100A generator would put less stress on the shaft.
Therefore yes: That shaft shearing is a myth
I’ve never heard of an alternator shaft shearing. I have heard alternator belts squealing, and being burned. And, as I mentioned, the front mount alternator IO-520’s have a very expensive slip coupling, which are known to slip then fail. I spent a year part time trying to develop a replacement part, and gave up, for lack of success. All other big continentals drive the alternator through the starter adapter, which does not handle excess loads well – they are an expensive wear item, and some have an AD for cracked shafts. O-200/O-300’s are known to wear a very expensive drive coupling/gear if too much demand is placed on them. I keep accessory loads as low as possible, other than at cruise/high engine power settings.
I always start my Seneca with Alt Off due to high current taken by field when engines are stopped.
Very often switch one off during flight, then the other one to check the load. Never had problems.
However these are not belt driven.
@Raven but that is not in the POH?
Another Q:
If you have two same size alternators (on the same engine, or on two engines) driving the same bus, how is the load shared?
This is a classic problem in electronics and the solution isn’t trivial. Done properly, it involves current measurement. Otherwise, you need very accurately matched regulators and then the nonzero source impedance will take care of the current sharing.
With a “backup” alternator, like the 15A B&C one I have, the implementation uses a lower voltage reg setting on the backup (26.5V) so if the main alt fails, the bus drops immediately from the normal 28V and when it falls to 26.5 (a fraction of a second) the backup reg field current gets driven up.
TB10 is alternator ON after start and after confirming the warning light is lit.
RobertL18C wrote:
but that is not in the POH?
It is not.
Peter wrote:
Another Q:If you have two same size alternators (on the same engine, or on two engines) driving the same bus, how is the load shared?
In my case the load is not shared. 
Left gen takes all the load. Only when I switch it off – then the right gen takes over. I have two amperometers so I can see what is going on.
Probably need to adjust regulators a bit. Left gives me about 0,3v higher.
I assume if I match them to about the same voltage (+- 0,1v) they should share the load more or less and that would be a good situation.
Left gen takes all the load. Only when I switch it off – then the right gen takes over.
Assume you mean alternator, the turboprop STC market didn’t address the Seneca :)
Something is definitely wrong on your set up, at the very least you should be getting some kind of Alternator failure indicator warning. I have mainly flown the I, II and V versions, so your set up might be different?
