The bike started a few times during earlier attempts with my ignition system, so the bike hardware seem to do its part. However it didn’t run for more than a few seconds each time it started on that occasion.
During that time I timed the spark in relation to the zero crossings from the AC generator, so no external sensor was needed. However, phase shift from the generator was occurring especially during the critical start RPM’s which moved the zero crossing nonlinearly. That made it hard to actually know the position of the engine.
When I got it to start I was just having luck with the values I choose on an ignition curve in the software.
Therefore I decided it would be better to first get it to start and run cleanly with a more simple approach and then continue later with the first sensorless approach.
For some reason though, the spark is now a lot weaker than on my previous system and today I will try to figure out why. Updates will keep coming.