I'm using a user function to simulate incoming CAN messages and it works fine. One message contains a float (32 bit IEEE 754) which is only passed to the serial port. Up to now I used 4 pre-calculated fixed numbers but it would be great to change this float during simulation. Of course a union would do the job, but a simulation script is not "C". Is there a chance to avoid coding a float-to-IEEE754 user function? By the way: writing to the CAN memory area via _WFLOAT does not work.