I don't fully understand all the details of your question (screenshots and step by step instructions to reproduce the issue really help us!), but I think I understand the general gist - some accented character is not displaying properly in your finding aid.
Like all other parts of AtoM, we use UTF-8 as the character encoding in our finding aid generation. UTF-8 is a very flexible and widely adopted encoding standard (it's become the default standard for how characters are displayed on the web in and in emails for example), and so it should definitely be able to support whatever characters you wish to display.
In creating your description, did you type all the contents into AtoM directly? Or did you perhaps copy and paste from another document (like a Word document or Excel spreadsheet, for example)? Or, was this description created via an import, such as a CSV import or an EAD XML import?
I suspect that the problem has to do with a mix of character encodings in your content - this can sometimes happen if you've been cutting and pasting from non-UTF-8 sources, such as a text document or spreadsheet. Similarly, if you created a CSV for import in Excel, it's possible it wasn't set to properly use UTF-8 encoding, and while most characters might have come in okay, some might in fact not be properly encoded - meaning AtoM's finding aid can't properly display them.
Have you tried deleting the problem character in the description, manually typing in the correct ő character into your description, and then regenerating the finding aid to see if that helps?