How to find broken UTF-8 characters ?

Hi,

I regularly have to import from Excel files, that hold rather simple
text and number columns.
That works.

Occasionally there are unwanted special characters at the end of
text-columns that Exel and pgAdmin either show as a upward arrow with a
short leg on top 90° to the right or others are invisible with UTF-8
aware programs or get dispayed as ? by Access.
A text viewer shows "ÔÇÄ" or E2 80 8E in Hex for the invisible thingy.

My database is unicode so it doesn't mind those freak-chars.
The problem rises when I need to export those records to CSV with
pgAdmin. pgAdmin complains about not beeing able to store those lines in
the local charset.

How can I find those broken UTF-8 characters?
How can I get rid of them?