All you should need to do is use Unicode character types (NCHAR, NVARCHAR, and NTEXT) and SQL Server should handle the Chinese characters gracefully. Getting SQL Server to provide Chinese error messages is a bit more complicted, but let's deal with one problem at a time!
Something between your keyboard and the web page control is only passing 8 bit values. Since you can successfully enter Unicode (the Chinese characters) via SQL Enterprise Mangler, I think we can assume everything is Ok as far as your OS, which leads me to think the problem is either in your browser or beyond (probably the web page).
Just for the jolly factor, let's cut SQL Server out of the problem altogether for a moment. try to construct a "hello world" sort of web page with just a text control and a button. Have it simply allow the user to enter text into a control, then create a popup that contains that text when the user presses the button. This will allow you to get the web page working first, then we can concentrate on getting the data back and forth from the database.
Private Sub btnTemp_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles btnTemp.Click
Private Sub ShowMessage(ByVal message As String)
scriptString += "alert('" + message + "');"
scriptString += "</script>"
Good! Now, can you modify that same sample script so that a new button simply sticks the Textbox contents into a column in the database? If you can do that, then modify the existing button to pull the data from the column and display that data from the column.
Try running the statement in Query Analyzer like this
Insert into TestTable(Comments, Language) values(N'彰','chinese')
and see if that works.
How are you constructing your INSERT statement in your asp.net code?
Is it something like
string cmd="Insert into your_table(Comments) values('"+someTextBox.Text"')";
// Execute cmd
Edit: I see from your post that that is the case. You should really consider using a parameterized query for this. It ought to solve your unicode issue AND help protect you against SQL injection attacks.