Hi Heimo
The reason I ask for energy measurements is to understand why you think inductive case is erroneous. From above data, it is clear that the sign of the error changes when you switch from capacitive to inductive case. This does not necessarily mean that the error calculation is faulty. It only means that you are compensating the system in the opposite direction for getting accurate results in that particular case. Looking at the energy register values you get now, in both cases, you can discern that by the end of calibration cycle, the register readings in inductive and capacitive cases will be closer to each other. After the calibration, if you do energy measurements at various signal levels, you can verify that the results are still accurate even in the inductive case. Kindly try that and let me know.
Regards
hmani