Parse Invalid Text
Works kind of like CAVX's Unicode Stripper (and is very much inspired by that), except this is built-in to the page as a button next to submit.
Copy/Paste your invalid text into the textbox which you normally type into, hit "Parse Text", wait for the "Done!" alert box.
It does convert some characters, and other will be completely replaced with a space (eg. the single byte ellipse).
Be on the lookout for characters such as the black diamond with question mark, or the unicode "block" with hexadecimal digits (although neither of them should show up... I think :P)
You can try the page which runs the same functions, but the formatting won't be what you expect because it's only really designed for deployment via XML (ie. only use it for singular paragraphs).