My idea for Flying Cap or Never-Connected PSU

[roender]

Guys,

I need some help to develop this type of PSU.
The idea is to isolate the noise from transformer and rectifier.
The 50Hz hum noise could be removed with a super shunt or linear PSU inserted after the C2.
Could be this an first step for a battery like PSU?

Mihai

[roender]

Oooops ... a little mistake which can bring magic smoke out of MOSFET.
Was corrected.

[analog_sa]

I wonder how serious the switching pulses of the fet are and what can be done to reduce them. There is clearly more to worry about than 50Hz.

[roender]

I don't know if MOSFET switching noise is more harmful than main c**p plus diode noise.
Maybe some switch on/off control mechanism can make the current slope through the MOSFET less abruptly
What do you think?

[dantwomey]

Ecdesigns posted detailed schematics and a description of his 'never connected' circuit here. Never Connected?

Regards,
Dan

[analog_sa]

Originally posted by roender
What do you think?

It will probably depend on the power transformer inductance too. You've come up with a very simple circuit, not many excuses not to build it.

Originally posted by dantwomey
Ecdesigns posted detailed schematics and a description of his 'never connected' circuit here.

Thanks for that. Will definitely check it out.

[roender]

This is much better.
Take a look at the output ripple!

[Rick Miller]

Shows what a simple RC can do. Good work!

[roender]

Any thoughts regarding latest schematic? It simulate to good to be true!
I need a second opinion.

Thanks,
Mihai

[Nico Ras]

Something is wrong. Even in the other thread, there will be ripple because you are trying to charge the output cap from the input cap as well as draw current through a load, so you need to transfer twice the current.

There is another problem the output cap will only charge until the voltage on both caps are the same, which is lower than the start-up therefore there would still be ripple at 50 Hz.

Increase the load to something more than 1 k and you will be less happy with the results.

There is a final problem as I see it, as the voltage rises on the output cap and falls on the input cap the FET will turn of. Thus one would need a charge pump to ensure that the gate voltage remains high.