Just wondering and haven't tired it... are the later electric start (mine is 1920) Ts designed to start on mag as well as on battery? Or just battery?

They can start on either spark generator.

They will do either. Just like starting by hand, set your throttle and spark appropriately for either battery or magneto and then just use the starter instead of the hand crank.

Theoretically, all T's should start on MAG. But in the real world Ford provided a battery connection for a reason. The reason is that magnets can weaken over time, and a magneto may provide enough juice for running but not enough for starting. My runabout's magnets are well charged and I can start on MAG most of the time, especially when the car is warm. But sometimes it gets stubborn and I switch to BAT to get going.

My 27 RPU starts on Mag all the time. Battery or magneto should make no difference as long as the magnets are charged.

For those who haven't started on MAG and want to try it, I should mention that you don't fully retard the timing as you do when starting on BAT. For a MAG start the lever should be five or six notches down.

Steve is right, but the 5 or 6 notches down but is always puzzling. The linkage used to actually move the commutator does not seem to be as precise.

It begs the question on what notch does the kick back action start?

I think Steve's setting is not arbitrarily accurate. Every T has the timer linkage set differently (especially after some use).
So I suggest trying it fully retarded and then slowly advancing to find the best position for YOUR car

Dan
The answers may be found in this article Steve Coniff and I prepared many years ago. The Model T Ignition System and Spark Timing
It is important to remember this assumes the initial timing is set at 15.5 degrees ATDC and the spark lever linkage has no slop.
As you can see from Figure 8 when starting on magneto advancing the spark lever 5-6 notches will result in the first spark occurring at 4 degrees ATDC. At that point the cylinder will still have adequate compression to fire the fuel/air charge. If you did not advance the spark lever you would be trying to start the engine at 26.5 degrees ATDC which would be difficult because much of the compression is gone. This is why Ford recommended advancing the spark lever for hand cranked magneto starting.
Now contrast that with Figure 7. Starting on battery and advancing the spark lever 6 notches would result in the coil firing BTDC and a kick back is highly likely.
As we pointed out in the beginning of the article the Model T ignition system is commonly misunderstood and why we prepared this explanation to try to de-mystify it all.
It is important to understand that the Model T ignition system (and coils) operates differently on battery and magneto.

Where the spark lever is set when starting on a mag all depends on how the engine timing is set. Many folks including my set the timing just barely after TDC when the spark lever is fully retarded. My car starts well on mag or bat with the spark lever fully retarded.
https://www.youtube.com/watch?v=jYNulylgVgI

Jim

Jim
That may be fine for some Model T's where owners are careful about getting their initial timing properly set and all the slop out of the spark advance control mechanism.
In my experience most owners are unwilling to pay much attention to such details, just wanting to drive the car.
Setting the initial timing so closely to TDC is very dangerous if all other factors have not been satisfactorily addressed. I cannot recommend doing so UNLESS you know for sure.
I believe Ford set the Model T initial timing at 15.5 degrees ATDC for safety reasons and published the hand cranking starting instructions accordingly...

Ron,

With the timer segment needing to be advanced by 34 degrees from fully retarded before the next fire point (18.5 degrees BTDC), it would appear that you would have about 12 notches of advancement on the quadrant (about half way down) before you may have to worry about a kick back occurring. This is assuming properly adjusted coils and no slop in the linkage. Is this correct? I don't plan to test this theory, but basically want to confirm my understanding that because the coils are not constantly firing, as is the case with battery operation, that there is some margin inherently built into the system. For this reason, 5 to 6 notches of travel would get you within the range of 4 degrees ATDC firing point but 7-8 notches would as well, with no threat of a kick back.

I now understand the advantage of the magnets being advanced by 7 degrees. But the paper stops short of saying that the Ford engineers did this purposefully to achieve these results. It seems implied, but is that the case?

Thanks.

Dan
Your first sentence is correct IF you delete the word "next".
Yes, about 12-13 notches, but remember the timer case movement is not exactly linear as we pointed out.
Trying to start the engine at 18.5 degrees BTDC is the same as trying to start it at 12.5 ATDC. So it may or may not kick back, but I wouldn't suggest trying it.
When we wrote the article we could find no specific information as to why the magnets were mounted with 7 degrees advance.
But, other sources refereed to C. Harold Wiles testing pre-production Model T's and finding out what worked best for starting and running. That is only a guess, but obviously someone knew/understood what was going on in this area.
From your comments you seen to have a pretty good understanding of this complicated subject.

Ron and Steve's article on the Ford Ignition system is a great reference. The reason why Ford Engineers knew to advance the magneto magnets relative to the flywheel (and crank shaft /piston position) is because of the elementary phase relationship between AC magneto voltage pulse and the resulting AC current pulse in the ignition coil. The ignition coil is an inductive load which causes the magneto current pulse through it to be delayed in time with respect to the magneto voltage pulse. The magneto voltage pulse is synchronized with the flywheel/crank shaft/piston position. The ignition coil current lags behind the magneto voltage so that means the ignition coil current peaks (and spark fires) AFTER the magneto voltage pulse and since the flywheel/crank shaft is rotating, that delay means spark fires Later/Retarded; by about 7 crank shaft degrees at mid engine RPM for a 3.3mH ignition coil.

The phase lag between magneto voltage pulse and ignition coil current pulse (hence spark firing delay) is very small at low engine speeds but increases with engine speed. The piston would be too far advanced at higher engine speeds if the magneto magnets were not advanced with respect to the flywheel/crank/piston position. I agree, the decision to advance by 7 degrees was likely a compromise Ford Engineers determined from experimentation on what worked best at mid to high engine RPM.

Note that magneto magnet orientation relative to the flywheel/crank/piston position is completely independent of the spark lever adjustment governing when the timer activates the ignition coil relative to the CAM/Crank shaft/Piston position. The magneto magnet advance with respect to the flywheel/crank/piston position is what determines the timing between the availability of magneto power and when spark fires relative to the piston position when operating on magneto power. The spark lever/timer position only selects which magneto pulse is used to power the coil as described in Ron and Steve’s excellent article.

Thanks Ron and Mike for taking your own time to post follow up explanations. I had read that paper a couple times before but I think it finally clicked.

I think the satisfaction of gaining more knowledge is only eclipsed by gaining appreciation for the minds that created this ignition system and who made it work in this application. I can see how researching this information and connecting the dots could be incredibly satisfying.