Flyback + Secondary Synchronous Rectification question

[edaenrico]

Hi,

I'm trying to understand how the Synchronous rectification works on the secondary side of a Flyback converter.


I'm working on the following SMPS:

Topology = Isolated Flyback
Vin = 230 Vrms
Vout = 12 VDC
Iout = 1 A
SR MOSFET = On the low side

Question
---------
I browsed some schematics.
I noticed that the current flows in MOSFET from Source to Drain. Am I right or I am missing something?

Thanks,
Enrico Migliore

 

[mtwieg]

In a synchronous rectifier, current will flow from source to drain, correct. There's nothing wrong with that.

You should keep in mind that using synch rectification for a flyback will probably not give any significant benefit, especially if your output voltage isn't very very low. However, it will likely be expensive to implement (since it requires gate drive with isolation). Synch rectification should be reserved for lower output voltages, and more efficient, higher power topologies.

 

[FvM]

Yes, 12V/1A is a small wall wart SMPS format and would use preferably a schottky rectifier. But if you are required to implement as synchronous rectifier, why not? For optimal control, you'll usually refer to a dedicated synchronous switch ontroller.

 

[schmitt trigger]

Just remember to provide some deadtime between the main Mosfet drive and the synchronous Mosfet drive.

 

[edaenrico]

Hi guys,

thanks for your answers that helped me clarify the picture.

> it will likely be expensive to implement (since it requires gate drive with isolation)
what do you mean by that?

I got a prototype working with 2 fast diodes (no Schottky) in parallel.

On the secondary side I got a Freescale 8 bit microcontroller that measures the load current and other analog values.

I'd like to use the micro as a controller.


QUESTION 1
------------
When do I have to switch the MOSFET on?


QUESTION 2
------------
When do I have to switch the MOSFET off?


QUESTION 3
------------
I noticed that the body diode of the MOSFET do partecipate to the rectification in the first few microseconds of the period.
Am I right?


Thanks,
Enrico Migliore

 

[FvM]

I noticed that the body diode of the MOSFET do partecipate to the rectification in the first few microseconds of the period.

Not necessarily. The tranistor might be switched on early enough to not allow diode conduction. But you'll possibly use the forward voltage drop as an indicator when to switch the transistor on.

It's much more important not to allow diode current during switch-off because it will cause considerable commutation losses. MOSFET substrate diodes don't usually have good switching behaviour, except for some special types.

Slightly idealized, the transistor should be switched on when the transistor voltage crosses zero and switched off at current zero crossing.