Mario...the system is designed such that you need to have the controller and the APs running the same version of code. Once the versions match, the controller will be able to push out configuration updates to the APs.
If the versions do not match, you will normally still see the AP on the controller, but any configuration changes you make to the AP on the controller will not be implemented on the AP.
This is intentional. This will prevent potential problems that might happen if a controller were to push out a configuration to an AP that it's not compatible with.