Can't think why not. You've found three roots with the trig identity that you correctly got through deMoivre. They all work. Cubic equations have a maximum of three roots. We've disposed of the idea t=1 is one of them. Can't think of what else could go wrong.