thank you for your answers, very usefull, but I have now the following question:
I would like to avoid a dedicated High Side power source for the Vs of the module, so I would like to set up a "bootstrap" process with the DC/DC (375Vin to 5Vout) I have in cascade with the swith based on the LTM9100.
I am wondering if I could power the LTM9100 from the Bus with a Current limiter as described in the data sheet, for a power-up starting phase and then, once the DC/DC has started to generate the 5V (isolated from the input) use this to provide the VCC input power to the LTM9100.
In other words there will be an initial phase for which the MOSFET switch will not be completely enhanced, but it will be enough to let the DC/DC starting. After this phase the 5V generated by the DC/DC will power the VCC of the LTM9100 which, in turns, will generate the additional 10V to the isolated section requested to complitely enhance the MOSFET.
Could it work?
I think the success (or not) of this scheme will depend on the characteristics of the dc-dc converter and its coordination with the ramp rate for the load. Assuming that the dc-dc converter starts up and provides the isolated 5V output while the FET's Vd is still greater than Vs(min), I think it should work.
Note that the by using the current limiter to power Vs from the input bus in the high-side configuration, the load will start charging immediately upon the input being applied, so there's no way for the load to be "off". I don't think this is an issue with your application, but am mentioning just in case.