Failures in wind turbines are often attribute to faults in the power electronics. To investigate the fault mechanisms, a test bench was set up in which entire converters up to a power of 10 MW can be tested under changing climatic conditions and electrical loads. In order to generate realistic loads, one of the converters has to behave like a drive train of a wind turbine. Except from a few differences, the test bench which is used here and has a total power of 300 kW, is a scaled copy of the 10 MW test bench. This small test bench is used to perform preliminary developments of control concepts and test scenarios that can later be transferred to the large test bench. It consists mainly of two back-to-back converters and three transformers. The goal of this work is to force the behaviour of a drive train of a wind turbine on one of the converters. The second converter should be able to perform a generator current control, as is common in wind turbines. This paper shows the implementation of a wind rotor model, the connection with the existing generator model and the implementation of a speed-dependent generator load curve on the side of the device under test (DUT). The results demonstrate the functionality of the overall test bench. For this purpose, different curves of calculated and measured values such as wind speed, pitch angle, rotor speed, phase current and voltage under different conditions are shown. In summary, it can be shown that one of the converters behaves indeed like a drive train of a wind turbine and that realistic scenarios can be created based on measured wind speed curves.