F test is a test of joint distribution, while T test is univariate. F statistic should be amenable to F distribution under null hypothesis, while T statistic Student T distribution. Be careful, both tests should be used under strict exogeneity! That's a strong assumption.