Tested with MSSQL, PostgreSQL, Oracle and DB2 connectors, inserted Hebrew characters to text fields (depends on the DB), but got those malformed, assuming the issue is related to the server, for example:

update employee set firstnme = 'י'  where empno = '000010';

Got:

FIRSTNME:\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd\xef\xbf\xbd

Is it a known issue or some configuration related issue?