SOLR Snippets alternative query is needed (HELP)

Hi all

I'm using SOLR with a linear combination of n-gram stemming technique using several ngram indexes.

For that reason my queries are things like this: drameallgemeina hfttuna
resulting in this query
text:(drameallgemeina hfttuna ....)^0.5 text_5:(drame ramea ameal meall eallg allge llgem lgeme gemei emein meina hfttu fttun ttuna melde eldet ausma usmai smail maila ailan iland asdma sdmat dmatt matti attin ttino)^0.3 .....

I'm using a fulltext index with no stemming applied so I can generate the snippets with real words.
My problem relies to the fact that SOLR will use these grams of the query to find the best snippet, so will find words like "meina" or "una" in text if their exist?

In those cases it should be better to have an alternativeQuery so that SOLR can use it instead of the last one.

For example:

hl=true&hl.q=drameallgemeina hfttuna&q=t:(drameallgemeina hfttuna ....)^0.5 text_5:(drame ramea ameal meall eallg ....

Anyone could help me?