OK, I tried the triac out of the circuit, using two power supply units. The first one was set for 22V and max 3A and the second one was set for 5 V and max 0.2 A. The first one was connected between A1 and A2 pins of the triac and the second one between A1 and G pins.
With all possible polarity combinations, I'm not able to trigger the triac in quadrant IV (A2 being positive to A1, and positive trigger pulse).
It seems that the combination of voltage and current on the gate pin that triggers the triac in quadrant IV is V > 10 V and I > 0.1 A. Even V = 7 V and I = 1 A does nothing (the triac conducts only 0,1 A momentarily, once the trigger voltage is removed it stops conducting).
As I already wrote the triac is BTA41 600B. I can't see any text related to my issue in its datasheet. Any idea? Is that just low quality triac out of specs?