I have a CLOB that we are converting to a BLOB using DBMS_LOB.CONVERTTOBLOB and then serving it up as a file using WPG_DOCLOAD.DOWNLOAD_FILE. When you open the file, it is replacing the spaces with a Â.
Most likely, the spaces in the original CLOB were not regular spaces (0x20) but rather non-breaking spaces (0xA0).
What's the target encoding used to convert to BLOB? I suppose UTF-8?
UTF-8 encodes non-breaking spaces as 0xC2A0, where 0xC2 is specifically the codepoint for "Â". So I think the tool you're using doesn't support UTF-8, or doesn't recognize the file as being properly encoded.