Up and an update
Thanks to Omer of CatsLab (Krifartida), I got schematics for ATL2 head. But the design is a complete departure from ATL-1, and the reset circuit is totally different.
However, seeing the schematics put me on the way : the unknown IC is finally a NE556 (dual 555 in a 14 pins DIL package), connected in monostable multivibrator, as I suspected.
With this hint I was able to make enough tests to exonerate the reset circuit in the problem. It is somewhere else, I haven't found it where for the moment, but I am progressing.