After 3 years of fiddling around with it I finally finished the "self programming" version of the EC3 software.   Working well on the bench and will soon try it on the engine.

It has two closed loop modes, one where the controller monitors the mixture and trims the injection to keep mixture at stoic. (14.7 : 1 A/F)  as long as the basic programming is within trim range which is pretty wide.

The other closed loop mode works the same way but uses a very narrow trim range and updates the map table if necessary to keep the mixture at 14.7.   Since the loop is all internal to the EC3 it happens MUCH faster than the EM2/3 auto program mode in earlier versions and of course does not require the EM2/3 to work.   Should eliminate a lot of tuning headaches for some builders.

It does require the O2 sensor input of the EC3 to be connected to the sensor of course.   This input was unused in the past.

The open loop mode is still there for use after the map table has been programmed.   Failures of the O2 sensor, wiring or use of 100LL won't cause a problem in this mode.

Coil Failure:

Just an FYI data point.   At about 85 hours on the 20B I had a failure of an LS-1 D-580 type ignition coil, the first one I've ever heard of.   This one was a low cost aftermarket version, not AC Delco so that may be a factor or it may have been a random failure.   As usual, the in-flight symptom was a rise in EGT on the affected rotor.  Engine continued to run normally.   Disabling the not-affected coil set (trailing in this case) caused the rotor to drop out during the next pre-flight engine checks.
   
Tracy