Discovering non-ascii codes in text files

I am currently generating Word and PDF files based on the contents of a wide range of standard TXT files. The process is working well but some codes in the text files translate to unknown characters in the generated Word file. These characters are are mainly the " and ', but there are others as well.
I really want to be able to replace these unknown codes with their equivalent in Word. I am currently using Regex to replace American spellings but I would be grateful if anybody could point a way to handle these characters at the TXT source so that they produce meaningful characters in the resulting Word file.

Thanks