Derivation of Euler lagrange for Yang Mills

I need someone to sketch the conventional steps(from variation to vanishing of arbitrary function chosen , etc, etc) of Classical Yang-Mills. If using exterior product product could you emphasize any additional Physical assumptions. I am trying to understand the effect of variational calculus on the geometry of the system. If possible can you provide any reference that is very explicit about the steps taken, perhaps an older reference, but not Sir Michael Atiyah. I like his work on the subject but I am seeking something much more elementary to build off.