Readily exchangeable protons such as amines, hydroxyls, and carboxylic acids can be tough to find in a proton NMR. First question, though, you aren't using deutero-methanol for a solvent, are you? That would definitely exchange your amine protons for deuterium, making them disappear from your proton spectra and (very slightly) increasing the amount of residual CD3OH you see from your CD3OD spectra. Even in d6-DMSO, however, small amounts of water or very high concentrations of your sample can lead to lots of proton transfer, and your peak may broaden out so far that your baseline-leveling software obscures it altogether. Hydrogen bonding and resulting hydrogen exchange processes are very fast on an NMR timescale and may blur your peak into background.