I've been in this position many times. The general problem with N-debenzylation is that aliphatic amines can poison Pd catalysts (the effect is well documented).
A solution that I have found 100% successful so far is to acidify the reaction medium with HCl, which protonates the amines and prevents poisoning. In my experience, reaction mixtures containing HCl at a total concentration of 0.1-0.3 M have always successfully debenzylated in 24 h with 10% Pd/C under hydrogen at atmospheric pressure, even in cases where <10% conversion was observed over several days with various more active catalysts in the absence of acid. My solvents of choice are ethanol, water, aqueous ethanol or aqueous 1,4-dioxane. It is important to properly purge the vessel and flush with hydrogen (I normally do 3 cycles).
The slight acidity of ammonium formate may be the reason it works for you. If you have acid sensitive groups in your molecule you can use weaker acids, like acetic acid.