I assume you are having at least one more table ("Zirkelbezüge" needs at least 3 tables in your data model).
Without knowing your complete "Zirkelbezug", it's hard to suggest something.
There is a whole chapter in the reference manual, search for "Zirkelbezüge vermeiden".
A solution to your problem probably involves renaming some fields or re-modelling your data model (e.g. using 1 single fact table instead of multiple, using common date fields).
indeed i have three different tables with data formats which seem to cause the Zirkelbezüge.
The problem is:
1. one order table which provides all relevant data to the order. -> matching through date
2. one table which delivers traffic data from google -> matching through date
3. one table which contains costs for adwords -> matching throug date
When i rename the date fields from google, everything works fine. otherwise its screwing up my systems. With the datefields its actually working when i just use two sources from the one database, its only google that is messing with the data.
you'll need to decide how you want to interact with your data, how you want to filter your data.
For example, do you want to select a date and force QV to limit the original fields dim_date as well as date_added to that selected date, or are these dates semantically different and you may want to select different dates in these fields.
There are a lot of threads here in the forum (and in the technical blog) covering the concept of a so called master calendar, and how to link a master calendar to several date fields (using a link table) or how to build separate master calendar tables per date field.
For example, please check: