UCD 25 watts to 1200 watts using 2 mosfets

better use transformer, because you will dont have power supply pumping problems and dont need overvoltage protection
If you use enough caps, then you wont have pumping problems.
You can calculate sufficient capacitor values with: deltaU=Ubus/(8*Pi*f*Rload*Cbus)

The music power content under 70Hz is rapidly decreasing. Few kick drums are tuned/miked/eqd to sound under 70Hz, and bass guitar pickups dont get the first harmonic... so unless you design a sub amp for home cinema, and use sufficient caps (caps still cheap) supply bus pumping is not an issue.
Simple short circuit prot can be achieved by putting a shunt in and using a opto-coupler to supply Vn+12V to SD pin of IR2110, but the shunt dissipation will be higher than the main FETs dissipation, so this usual shunt method for Class B should not be used, unless you dont care about this small dissipation.
 
If you use enough caps, then you wont have pumping problems.
You can calculate sufficient capacitor values with: deltaU=Ubus/(8*Pi*f*Rload*Cbus)

The music power content under 70Hz is rapidly decreasing. Few kick drums are tuned/miked/eqd to sound under 70Hz, and bass guitar pickups dont get the first harmonic... so unless you design a sub amp for home cinema, and use sufficient caps (caps still cheap) supply bus pumping is not an issue.
Simple short circuit prot can be achieved by putting a shunt in and using a opto-coupler to supply Vn+12V to SD pin of IR2110, but the shunt dissipation will be higher than the main FETs dissipation, so this usual shunt method for Class B should not be used, unless you dont care about this small dissipation.


so please can you offer schematic for SD-Pin with IR 2110
 
You don't need low resistance R... and yes, you would get transformer, just like current sense trafo, but on output core
yea, I think that was one of the options when I asked

Luka, thanks you hints. I read the posts of Workhorse about this in topic "D AMP is back !!!". I put the schematic into Multisim, and tried to simulate it with no success.
An externally hosted image should be here but it was not working when we last tested it.

The voltage after diode bridge reamins the same 3,8V , does not change with signal or load.
 
Pafi, what do you think of my idea, as using current-transformer for current-sensing ?

In the output? I don't like this method, because the signal frequency can be very low, so it needs a very high Al, and very high number of turns.

With a small change (series schottky and external FRED) you can measure current in the drain with small current transformer (plus demagnetisating diode on secunder), but only as long as the oscillation is on.

If (for example) slow overcurrent is detected and limited in power supply, then the fast overcurrent can be detected with current transformer.

The best overcurrent sensing method based on Vdson IMO. (However nowadays I returned to simple shunt resistors (10...25 mOhm) and BJT. What a shame! ;-) ) The difficult part of the Vdson sensing is proper blanking of the switching transients.