Let $\mathcal M$ be a premouse. $\mathcal M$ is said to be $\omega$-small if and only if whenever we have that $\kappa=crit(E)$ for some extender $E$ on the sequence of $\mathcal M$, then $\mathcal J^{\mathcal M}_{\kappa} \not \models$ There are $\omega$-many Woodins. $\mathcal M$ is said to be properly small if and only if $\mathcal M$ is $\omega$-small and $\mathcal M \models$ "there is a largest cardinal and $ZF^-$ (i.e without the power set axiom) and there are no Woodins".
Now if $\mathcal M$ is $\omega$-small then so is any of the $\mathcal J^{\mathcal M}{\beta}$ for any $\beta$ but if $\mathcal M$ is properly small then it could be that there is some $\beta$ for which we have that $\mathcal J^{\mathcal M}_{\beta} \models$ there are $\omega$ Woodins.
I am confused about how that can be, since in the definition of properly small we also have that the premouse has to be $\omega$-small. Can someone show me an example? Thx