I would say it's not 'completely' unexpected.
Could you elaborate on what you mean with 'complete axon' exactly ? Did you just remove the code that replaces the axon with a stub ? Did you change anything to change the code that puts the channels in the axon. If you didn't do the latter it means the (fairly high concentration) Na channels will be spread all over the axon. This will not be biophysically correct.
One issue with the current axon stub (which we are fixing in a new version of the modes) is that there are boundary effects due to it's limited size. To solve this we will add a long section after the stub. But in your case the behaviour of the model might change just because the boundary effect disappears.
Is there a way you can share you code ? I could have a look at it if you want if you're scared of a bug.