If you say you expose a plate on the ctp and process in another plate processor and it comes out fine, then why is the platesetter suspect ? I think you have to reconsider everything and start over again. Forget about focus and beam shape, come back to simple things : what do you see on plate ...