If you want the simplest and probably the most reliable ignition system get
a two cycle two cylinder small outboard ignition system.

Fires the same number of times and 180 degrees apart just like a 2 rotor.

As close as you will come to a A/C magneto.

http://www.iboats.com/

http://www.iboats.com/Johnson-Stators/dm/view_id.271367

http://www.cdielectronics.com/
Down load the complete catalog.

Some of these stators are as low as $20 used.

The high voltage coils supply 300 volts for a simple CD system and the other
coils supply enough
12 volt current for an EFI system. No belts. If the main electrical system
or alternator fails the engine still runs.

This stuff is mass produced in the millions and extremely reliable.

Paul Lamar

Some interesting stuff in that catalog. Some of us wouldn't know exactly
what would comprise a complete system, as it looks like they are all parts
of an upgrade for certain motors.
Did you notice the EFI and ECU units in the catalog? They also note that
these can be reprogrammed for "racing purposes".
Would this be suitable for our purposes? Of course I don't think it would be
a good idea to mention aircraft. One cold say I'm making a marine rotary
conversion.
C. Smith


Good idea Craig. A racing rotary boat motor.

All you need is a source of 300 volts, a 1 ufd 300 volt cap and a switch of
some sort. An FET works fine and they are easy to use. Way back about 30
years ago we discovered that a 50 cent CMOS CD4050 or a CD4049 chip makes a
great FET driver if most of the six gates in the chip are connected in
parallel.
Since that times 1000's of FET gate drivers have been  designed and built
selling for as much as $6 each and so far I have not found one that works
better than a CD4050.
Many of them require a regulated supply which lowers efficiency and cost
more money and parts.

I set up this simple CD circuit on the bench in half an hour and it works
great except for the $150 300 volt supply.
In theory it is 30 watts which should be enough.
Unfortunately the high buck PS does not like to work with CD systems. It
shuts down for several milli seconds right after I fire the cap. It sees the
short. That means the RPM is limited.
The heat sink is essential.

The FET is a IRF740.

I am rapidly discovering it is all about the 300 volt power supply. You need
a robust power supply that does not crap out when it is shorted to ground
and that is where the charging coil in the outboard motor comes into play.

Nothing could be simpler than a large magnet passing a coil to rapidly
charge the 1 mfd cap.

Paul Lamar
--
So now I have more questions. Does one of these 2 cyl systems have enough
juice to fire both plugs at the same time?
Can it be used for multi-spark, or would one need to stack 2 or more with a
bit of angular offset, then utilize your FET controller to switch the spark?
Last, maybe a failsafe contactor to switch out multispark in the event of a
malfunction of computer to revert to stone axe mode? Or am I making it
overly complicated?
C. Smith


I think stacking two would be needed. The charging coils are available stand
alone and very cheap. All we need do is find the curved magnets stand alone.
Jeff can make the rotating magnet holder. The electronics are trivial
as I just demonstrated.  Here is a picture of what the spark looks
like. The picture is a bit out of focus but you can see the spark.

BTW I am thinking I of using a smaller cheaper version (40 pin)
of the Atmel computer to control the stand alone multi strike ignition.
The half bridge FET drivers on the market are not designed for CD system and
I am having the same problems with reliability the Aussie's had with
theirs. Namely blown FET's due to precise timing requirements.

Paul Lamar

K&J Magnetics makes cutom magnets, $150 minimum order. I wouldn't have a clue as to the dimensions or polarization, but I'll bet you do. Looks like a quote is free.
http://www.kjmagnetics.com/custom.asp

C. Smith


One of my hangar neighbors fixes Mercury's. I can get the dims
of one of those flywheels. Probably not for a week as I have
to go to PA on biz next week.

Paul Lamar

Check out the last five magnets at the bottom of this page.
http://www.kjmagnetics.com/products.asp?cat=168
A the very least, it demonstrates it is doable.
C. Smith

Looks too small. We need something 3 inch radius
or larger. Some of these big ones are high bucks.

Paul Lamar