UTF-8 file encoding issues within Java?

Question

I'm working on an application that takes data from an IBM mainframe(z/OS), converts it from IBM-1047 encoding to UTF-8(via iconv utility) and binary FTP's it to a Unix box where we process the file with our Java app and return the processed file.
Within our Java app on the Unix platform we stream the file into a byte array and then create a new String from the byte array specifying "UTF-8" as the encoding parameter.
The problem is that Java appears to be taking certain 2 byte UTF-8 characters and converting them to a single char.
E.g. I have a \uC3A6 char in the input file, I can view the bytes in the byte array that's read in, and it's still a \uC3A6, but as soon as I create the new String with UTF-8 encoding and view the bytes, those 2 bytes are now shown as a single byte(0xE6). The code I have that's looking for the char \uC3A6 then fails.
Can anyone explain what's happening here?? Sorry for the long message.

Answer 1

The encodings which convert the character (char)0xC3A6 to the 2-entry byte array {0xC3, 0xA6} (unsigned) are "UTF-16BE", "UnicodeBigUnmarked", and "UnicodeBig." These are essentially identical except for the use of byte-order mark. As was said above UTF-8 converts (char)0xC3A6 to the 3-entry byte array {0xEC, 0x8E, 0xA6} (unsigned).
http://java.sun.com/j2se/1.4.1/docs/guide/intl/encoding.doc.html