It would be a good idea to define varTypes within sqlSave. Here's how I do it. :-)
columnTypes <- list(ColumnName="varchar(255)", NumberColumn="float", datecolumn="date")
This defines the column types for you data so you can add the argument within sqlSave. So lets do the sqlSave now.
sqlSave(channel, YourTable, rownames = FALSE, varTypes = columnTypes)
and if the table has already been added then I would do an sqlDrop before that.
Why is this an answer?
Well, I had the same problem. I was getting an error saying it couldn't execute the sqlSave but it was putting in my columns in the database. So I had to fix the columnTypes. R tries to do it manually but it's not perfect. :-)