Basically, your simulink model is solving differential equations. If you make the step-size too large as you can do with fixed-step solvers, the differential equations cannot be solved within the error tolerances. This results in model errors. You can use rate transition blocks to your output blocks to limit the logged data points. But large models will generally use a lot of CPU power and can run slowly. You can try to optimize the model or to adjust the simulation tolerances, but this is a more advanced function.
Second point, in the simulink scopes, you can remove the data history limit in the configuration window. This will show you all the information in the scopes.
Hope this helps,
Dennie
