Maintaining character entities

Hello,
I'm new to the list - I've checked through lots of archive material and Dave
Pawson's stuff which was helpful, but I can't actually solve my problem,
so...

I've got XML documents, marked up to a DTD, and calling character entity
sets. When I run through the XSLT processor (xalan) to output another XML
file I find the entities have been converted to something different, and
fairly inconsistently. 

What I would like to achieve is having “ ü in my input xml, and
these entities still being untouched in my output. Can anyone advise how I
achieve this please?

What I'm getting are (“, ü), or (“ and ü), or (“
and ü), depending on character encoding settings and entity sets used. Am I
missing something?

Thanks in advance

Chris

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list