So your problem is that you reduce coating concentration to levels that are not normally used for coating (too low) and the result is the same, is that right?

I think you should also coat with a non-relevant antigen to see if the signal is related to specific recognition of your coating antigen or just non-specific background? In the latter case you should try to improve blocking on non-specific interactions by diluting the standard/samples in a proper blocking solution.

I would advise you to run a SDS PAAGe first to see if any other (abundant) protein(s) are present in the vaccine. Consider using a capture antibody  (sandwich ELISA)

You can also use a monoclonal (anti HBSag) to check for specificity (most likely you would need a anti mouse conjugate)

Thank you. I will try this out.