.net - Use A Nchar Column - Store Czech Text In An IBM819 Database?

switching our database over to UTF-8 (multi-byte char, unicode) encoding from IBM819 (single-byte char, Latin1 subset), because of a new requirement to support text entry of Czech characters.I've just been reminded by a colleague of the nchar and nvarchar column types, which neither of us has actually ever used.

If I were to define the relevant columns as nchar and nvarchar instead of char and varchar, what would I have to do to allow my currently-IBM819 database to work with this new IBM819-incompatible text coming in from my .net application?Would all conversion be automatic? Would I need to write some scary conversion code?

EDIT: Right, I defined an "nchartest" table containing a single column, "ntest" of type nchar(500).I connected to that via .net code and confirmed that 500 chars of lorem ipsum could be entered:

Dim insertString As String = "INSERT INTO nchartest VALUES (?)"
Dim insertCommand As New IfxCommand(insertString, cConnection, tTransaction)
insertCommand.Parameters.Add("incident_recno", IfxType.NChar).Value = "[lipsum]"
Return insertCommand.ExecuteNonQuery

I then replaced the lipsum string with 100 or so chars of Czech text, and attempting to insert that caused an IfxException to be thrown by the .net Informix interface, with the message ERROR [HY000] [Informix .NET provider]Unspecified System Error = -21005.

Following up on some research I did yesterday, I tried setting my CLIENT_LOCALE setting in my connection string to en_US to match that reported by Threading.Thread.CurrentThread.CurrentUICulture, but that connection could not be opened due to ERROR [HY000] [Informix .NET provider][Informix]Unable to load locale categories.

Following that, I tried setting CLIENT_LOCALE=en_US.unicode, which produced an ERROR [HY000] [Informix .NET provider][Informix]Unspecified System Error = -31142.