Globalization :: Non-English Character In Oracle 10g Express Edition

I have a table. It's name is INSTITUTION. It has a NUMBER INS_ID and NVARCHAR2(50) INS_NAME . INS_NAME can contain Turkish characters, such as "ğ,ü,ş,ç,ö". According to business logic, there can not be a repetition on the INS_NAME.User will enter institution name from a textbox in ASP.NET , and I check this name in database from c sharp code, if there is no repetition, we will add this record.

The problem is; when user enter a instition name that contains Turkish character, there is a duplication. If there is a instition name is *"su işleri"* , the both query; SELECT * FROM INSTITUTION WHERE INS_NAME = *'su işleri'*; and SELECT * FROM INSTITUTION WHERE INS_NAME = *'su isleri'*; returns no result, even though there it is.But if instition name is "oracle corporation" (there is no Turkish character) it query successfully. I have the same problem in Toad for Oracle 11.5.1.2. When I query database from toad SELECT * FROM INSTITUTION, the phrase *"su işleri"* has appeared. But when I query SELECT * FROM INSTITUTION WHERE INS_NAME = *'su işleri'*; , there is again no result.When I connect oracle database directly and perform the query SELECT * FROM INSTITUTION , the phrase *"su isleri"* (not *"su işleri"* ) has appeared.

Here are the language settings of the database:

National Language Support
National Language Parameter Value
NLS_CALENDAR______________GREGORIAN
NLS_CHARACTERSET__________WE8MSWIN1252
NLS_COMP__________________BINARY
NLS_CURRENCY______________TL
NLS_DATE_FORMAT__________DD/MM/RRRR
NLS_DATE_LANGUAGE________TURKISH
NLS_DUAL_CURRENCY_________YTL
[code]....