An updated test of AMBER force fields and implicit solvent models in predicting the secondary structure of helical, beta-hairpin, and intrinsically disordered peptides