Now the gain of the input set for a PID is considered, such that the set value is the physical voltage that goes into the ADwin.
For that I also changed how we keep track of the AIn of PID channel index for the module arrays (moved storage from Attributes in devices to connection_table_properties).

Because the TiCo is started later than the Adwin main process,
an error is raised if we want to change the output too early again after t=0

Change TiCo rate to 50MHz, collect analog outputs not via PseudoClock but directly from outputs (such that we can use different rates for ramps in the same clockline).