ODBC translation codeset issue

Hi! We have a strange problem connecting from a windows client through ODBC to an Informix SE on a SCO Openserver 6 server. Situation: Server: SCO Openserver 6 with Maintance Pack 2 installed INFORMIX-SE Version 7.25.UC6 The locales of the database are en_us.819 Client: Windows Xp Service Pack2 IBM Informix Client Software Development's Kit 2.90 TC6 In order to access the testing database on server we configure a System DSN using ODBC driver supplied with Informix CSDK. We configure successful the "Connection" tab and we are able to test successful the database. Then we tried to configure the "Environment" tab in order to make a translation codeset between ISO8859-1 codeset of the database and CP1252 Windows' standard codeset. We inserted "en_US.CP1252" in client locale, "en_US.819" in database locale, and we left unchecked "Use Server Database Locale". Also, as we found on official IBM documentation, we inserted "igo4n304.dll" in Translation Libray and "0" in Translation Option. We created an empty Access database and we tried to link a table from the Informix's testing database. The result was that we weren't able to display correctly specific Italian letters ,like accented letters "ù","à","ò","è","é","ì" on Access(Access displayed "usual" small squares) . These characters were inserted before directly on server with Informix tools like "isql". And if we try to insert a new record with some accented characters we are able to disppplay them just in Access but if try to display them on server with Informix tools we can just see odds characters. We tried to google it but every information found didn't help us... Any ideas or suggestions are appreciated! Thanks in advance, Giacomo