is hard for me.
But every now and again, I learn something cool about my field when I write exams. Not complicated things, but instead neat, small facts about how the tools are applied. It doesn’t happen every time, but often enough to make the effort worthwhile intellectually. And the good students get the questions.
The hardest part is figuring out how to write the question clearly enough so that the students can do it, if they understand the material deeply. Making it simple enough so that the computations are not too hard, either.