in reply to Re^5: XML::Simple parser error : Input is not proper UTF-8, indicate encoding
in thread XML::Simple parser error : Input is not proper UTF-8, indicate encoding

nvivek arrived at this weird notation with angles and ^ followed by a letter by displaying it in vi. Your example program is wrong, it has a literal ^ and N as you neglected to substitute this notation for the original character. When the is corrected, the program predictably bombs out with:
$ perl pm986755.pl Entity: line 9: parser error : PCDATA invalid Char value 14 <CALLERID>&#xA0;jW&#xB7;h&#xAE;&#xF5;&#xBF;&#x8A;7a&#xB7;&#xD8 +;T&#xD9;< + ^
When the character is substituted with the character reference &#x0e;, it also bombs out:
$ perl pm986755.pl Entity: line 9: parser error : xmlParseCharRef: invalid xmlChar value +14 <CALLERID>&#xA0;jW&#xB7;h&#xAE;&#xF5;&#xBF;&#x8A;7a&#xB7;&#xD8;T&# +xD9;&#x0e; + ^
Upgrading the version in the PI to 1.1 does not help. XML-Simple respectively its underlying modules XML::Parser/expat and XML::LibXML/libxml2 cannot deal with XML 1.1!

Your advice was flawed from the beginning, it simply cannot work in the general case. Whatever puts control characters there is apt to also put a chr(0) character. No matter whether plain character or character reference, it's illegal in all versions of XML.