Ohhh, I'm happy!
but I tried writing the conservation of energy, without using special frames, between t=0 and the position in which the rod has come back in vertical position:
##1/2 I \omega_0^2 = 1/2 I \omega_{\theta=0}^2 +1/2 M U^2+1/2 (2M) v_{cm}^2##/
but if ##\omega_0=-\omega_{\theta=0}##...