The author says that the sentence in a) is impossible to translate into the language of logic, but it is possible for the sentence in c). And I can not understand why, the second sentence is about the same causality effect: the first sentence connects not setting an alarm clock with oversleeping, and the second sentence connects studying hard with getting an A.
What could prevent me from claiming that the causal connection in c) is beyond the truth value of the components?

These look like Paul Teller's suggested on-line answers to questions on p. 56 of his book.
I agree with you: (i) 'You will sleep too late unless you set your alarm' and (ii) 'You won't get an A in this course unless you study hard' look pretty much on a par to me.
Now, you could arguably take two lines:
Compare: "Either you set the alarm, or you don't set the alarm and will oversleep": your reason for saying that may be a casual belief, but arguably what you say, the literal content of the message, is just truth-functional: the suggestion in (B) is that it's like that with "unless" claims. And I think that there's something to be said for taking line (B).
But I honestly, with due respect to Teller, can't see any good reason for confidently treating (i) and (ii) differently.