Why Rotax-Ducati have implemented such a complex kill magneto circuit in CDI modules?

Hi Mike. I've been thinking about your last post for several days.

What you are suggesting is that the MOSFET acts as an active element to "raise the voltage" by making use of the inertia current of the coil (flywheel) and thus be able to properly charge the capacitor. Am I getting it right?

However, speaking in terms of energy, not only voltage, the capacitor should be charged equally. Following your reasoning, if I am not understanding it wrong, the circuit of the figure (that is has removed the circuitry of the MOSFET) would not work. Truth?.

However, if I'm not wrong, 
this is the classic/academic/canonical design of 99% of the CDIs used!!!.

I’ve never looked at any other CDI, so I can’t offer a comparison. 

There’s not a lot of information available regarding the 912 ignition charging coil voltage. I’ve only found one report indicating 30 VAC measured while running and under load (module under power.) Engine RPM wasn’t Indicated.  If we assume 40 volts peak to peak and without the MOSFT cycling (as in your drawing,) the half wave may charge the capacitor to 20 volts. If  we also assume the ignition coil has a ratio 100:1, we generate a 2000 volt spark. 

We need 210 volts at the ignition capacitor to drive the ignition coil’s primary winding in order to generate a sufficient spark. The capacitor voltage must remain consistent across the full RPM range, from cranking speed to beyond red line. That same 100:1 ratio coil now provides 21,000 volts. 

Bench testing showed that MOSFET cycling caused the capacitor to reach the change limit of 210 volts, even at the very low voltages expected at cranking speeds. 

Here are some waveforms that I recorded from my engine because I was curious about the voltages from the various coils. I've recorded the ignition charging coils (loaded and unloaded), the trigger coils and the primary of the ignition coil. I have attached some waveforms and a list of AC voltages from a multimeter provided by Rob Seaton in an old thread.

https://www.rotax-owner.com/en/912-914-technical-questions/6704-unable-to-start-no-ignition?start=5

Multimeter set to AC volts, connect one probe to one red wire and other probe to ground.

For start-up RPM: ~300rpm-7V AC (all values approx)

If you have the engine running,

500rpm-12V.

1000rpm-20V,

2000rpm-40V,

3000rpm-60V,

4000rpm-80V,

5000rpm-100V.

Hi Kevin.

I have found a similar V / RPM table in a post of M. Kyle in recreationalflying forum 

https://www.recreationalflying.com/topic/9632-rotax-912-ignition-problems/page/4/#comment-496841

RPM       Meter     Oscilloscope 
               V(RMS) V(P2P)

300         5,4                20

520         9,2                40

800         14                 60

1000      18,5               80

2000      34,5             150

3000      50                200

4000      66                300

Values are slightly lower that R. Seaton's.
M.Kyle  is using a good quality Fluke meter so I suppose that values are true RMS internally computed from high frecuency samples.

Hi Mike. 

You're probably right, but I find this way of controlling the MOSFET to raise the voltage a bit strange.
Thank you, for your detailed explanations.