Oracle UTF8 character set data displays incorrectly in Excel 2000 SP3

[Erik Martens]

Server: AIX 4.3 / Oracle 8.1.7.2.0. Enterprise Edition
Database
Client: Microsoft Excel 2000 (9.0.7616 SP3)

A CSV report file is generated from an Oracle Portal web
application. When the CSV file contains special characters
the CSV file displays correctly in Notepad but incorrectly
in Microsoft Excel.

We use following code to generate output in a CSV file
htp.p(CONVERT('WIN:WE8MSWIN1252 - Canon España
S.A.', 'WE8MSWIN1252'));

NOTEPAD EXCEL
WIN:WE8MSWIN1252 - Canon España S.A. WIN:WE8MSWIN1252 -
Canon Españ S.A.
When opening the CSV file in MS Excel why is the a after
the ñ suddenly missing?

Opening the CSV file in Hex Editor:
ñ displays as . in Cygnus Hex Editor.
.. Corresponds to hexadecimal value: F1

We would like the same output when opening the CSV file in
Notepad and MS Excel.
Does Microsoft Excel 2000 SP3 support Oracle database data
with character set UTF8 (multi byte Unicde character set)

Could you please share any possible similar experiences or
provide advice on how to solve this issue.

Thank you in advance,

Erik Martens