SMPS syncronous rectifier question
MindSpring User
I am trying to find an info about syncronous rectifier (I am not sure it is
right name for it) - very low Rds MOSFET used instead of the diode in the
SMPS rectifier circuit. Where can I read more about it? There are many
things that are not clear to me. I
have never seen power MOSFET without embedded diode between source and
drain for example....
I just need a hint where to begin.
Thank you in advance
Michael
PS please reply to '(ko...@mindspring.com)'
Bill sloman
MindSpring User wrote:
>
> Hi to everybody.
> I am trying to find an info about syncronous rectifier (I am not sure it is
> right name for it) - very low Rds MOSFET used instead of the diode in the
> SMPS rectifier circuit. Where can I read more about it? There are many
> things that are not clear to me. I
> have never seen power MOSFET without embedded diode between source and
> drain for example....
The MOSFET itself is a perfectly ordinary power device, and you have to
drive the gate just as you would with any other MOSFET.
When you use it as a "synchronous rectifier" you put it more or less
where
the rectifier would go in a convnetional circuit, and make yourself an
extra gate drive waveform to turn the MOSFET on when the rectifier
would have been conducting.
The advantage is that if the Rds is low enough, the voltage drop across
the
MOSFET when it is turned on is less - usually much less - than the 0.5V
to
1V across a high current rectifier.
Hope this helps
Bill Sloman, Nijmegen
Allan Herriman
I built one of these just recently (buck configuration). The
controller chip allowed a few tens of ns of simultaneous "off" time in
both mosfets. This is needed to avoid "cross conduction," which would
cause the mosfets to short out the input supply.
During this "off" time, the body diode of the synchronous rectifier
fet conducted the full output current.
This isn't a problem in itself, as the body diode points the right
way.
I figured that I could improve the efficiency slightly by putting a
schottky diode in parallel with the fet. I would save because the
voltage drop would be lower during the "off" time, but I would lose
because of the slightly increased capacitance.
Test results for 2A, 500kHz, 5V -> 2.5V buck converter.
92.4% efficient without schottky diode.
93.2% efficient with schottky diode.
So, for my case the extra diode improved the efficiency by a
worthwhile amount.
Another point about synchronous rectifier converters: they don't
support a discontinuous current mode at light loads. You will always
have a certain current flowing in the output inductor, leading to
lower efficiency at light loads.
Certain controller chips can sense this condition and turn off both
fets to save power.
Regards,
Allan.
On Tue, 06 Apr 1999 11:16:32 +0100, Bill sloman <bill....@ieee.org>
wrote: