is there a way to run a Simulink model in floating point single precision without adding several converter blocks? Does anyone know a simpler possibility, i.e. by changing some general settings?
Typically the data type is propagated from the Source blocks in the model - if you configure these blocks to produce a single type, the rest of your model might become single-precision too.
Another option that I can think of becomes available only if you have a Simulink Fixed-Point license. I believe setting the Data Type Override option to 'Single' and Data Type Override applies to 'All numeric types' on the Fixed-Point Tool should force all data types to become single-precision.