Let $L, K$ be complete lattices. A lattice homomorphism $f: L\to K$ is said to be incomplete if there is an infinite set $S \subseteq L$ such that $f(\bigvee_L S) \neq \bigvee_K f(S).$
Suppose that $L$ contains a prime ideal $P$ such that there is $S\subseteq P$ with $\bigvee_L S \notin P$. Then we can define $f:L\to K$ by sending $P$ to the smallest element of $K$ and $L\setminus P$ to the largest element of $K$ and easily see that $f$ is an incomplete lattice homomorphism.
Does the converse of the above hold? More exactly: Assume $L, K$ are complete lattices and $f:L\to K$ is an incomplete lattice homomorphism. Does this imply that $L$ contains a prime ideal $P$ such that there is $S\subseteq P$ with $\bigvee_L S \notin P$?