The news D-Amp DLS3000....

[Tekko]

Fredos: very nice amp indeed

mgm2000ro: Thats no class D amp, thats a regular mono MOSFET amp. I don´t see any oscillator, comparator or integrator, just a regular class AB/B MOSFET amp.

[fredos]

The output configuartion is standard, 2 output in half bridge, but the drive circuit is different. No pulse or a very narrow pulse with no input signal, and when input become more positive, positive pulse is feed to the output filter with no negative pulse, except the reverse enegy feeded by the inductor. Same for negative input. The ESD1 in anti parrallel with the STW50Nb20 over come the slow recovery of the body diode. Just remember in my case that output did not work in complementary fashion, so recovery could be very soft. That's what I call B-D, like a class B, nothing happen with no input signal. Less EMI, higher effiency, faster feedback recovery! Yes the common mode is done with yellow-with iron poweder coil, with 2 filter at 22uH, but I also use a IEC power line filter that not show in picture. The 2 common mode is just to lower emission between board and line filter!

And before you ask me, a NE555 in charge pump feed the power for the hight side mosfet when no signal is feeded to power amp!

Fredos

www.d-amp.com

[Workhorse]

Hi Fredos,

This is just like what you use as charge pump for your High side switch....

regards,
K a n w a r

[Pafi]

except the reverse enegy feeded by the inductor.

Wich means some 10 Amperes, and can easily hold diode opened till the next positive pulse. Forward voltage of an ES1D (I think you meant this) is much more than MOSFET's, so how could it reduce recovery loss?

Eva!

These MOSFETs are not suitable at all for classic class D topologies,

Despite of this Tripath recommends the very similar STW34NB20. I don't really understand!

[fredos]

Yes Workhorse, similar, but with some 2N7000 as buffer to provide all current needed, directly referenced to V--.

Pafi. ES1D, sorry for this, yes, but they only work in the first fews nano second of switching, after that, the body diode of mosfet take care of the current. That why I use a higher voltage drop diode, just for the speed of them, after that, the body diode take care of all the rest!

Tekko: sorry, but it was a free running class d amplifier...The first op amp is the error amplifier, and all transistor was configured as a hight voltage comparator...Like the UCD, but at lower frequency.

I wish that this tread give a lot of idea to everyone!

Fredos

www.d-amp.com

[Eva]

Originally posted by fredos
The output configuartion is standard, 2 output in half bridge, but the drive circuit is different. No pulse or a very narrow pulse with no input signal, and when input become more positive, positive pulse is feed to the output filter with no negative pulse, except the reverse enegy feeded by the inductor. Same for negative input.

There is no diode recovery phenomena at idle and low output levels because inductor current crosses zero during each cycle causing the switches to turn on at zero voltage, so that "B-D" mode of operation does not have any impact over EMI, but does increase cross over and low level distortion due to the inherent discontinuity. It's preferable to turn-on the switches in a straight complementary way with minimum dead time as that shunts the body diodes during most of the cycle, thus almost preventing them from conducting. That in turn improves efficiency (particularly with low Rds-on devices) and reduces diode recovery issues because much less charge is stored in the junctions. Remember that stored charge depends on the amount of time that the diode has been previously conducting, and that reverse recovery times and currents are dependent on the amount of stored charge. In practice, it means that a plain body diode recovers much faster if the MOSFET was shunting it just 50ns before it's required to turn off.

The ESD1 in anti parrallel with the STW50Nb20 over come the slow recovery of the body diode. Just remember in my case that output did not work in complementary fashion, so recovery could be very soft. That's what I call B-D, like a class B, nothing happen with no input signal. Less EMI, higher effiency, faster feedback recovery! Yes the common mode is done with yellow-with iron poweder coil, with 2 filter at 22uH, but I also use a IEC power line filter that not show in picture. The 2 common mode is just to lower emission between board and line filter!

And before you ask me, a NE555 in charge pump feed the power for the hight side mosfet when no signal is feeded to power amp!

Fredos

These cores are not suitable. The resulting filters will show very low inductance and RF impedance. There are special materials intended to show high impedance in the 10Mhz to 100Mhz critical range. Also, the required turn counts are somewhar higher (maybe a dozen on each side).

[fredos]

Eva

That's why I use these core, just for radiated emission. Conducted emission are take by the IEC line filter...I think you should revise your last post about recovery phenomena, because inductor create a big reverse voltage swing with verey narrow pulse....So? And sorry, no crossover distortion with this topology. Hysterisis depend of input signal, so lower is it, lower is the hysterisis of the comparator, fews 1/100 mv at input make change at output...A trick here!

Fredos

www.d-amp.com

[mgm2000ro]

Escuse me fredos for using your topic with my question.

I'm looking a schematic class D amplifier with a high power output up to 1kW for my car amplifier (i intend to built). I'm intrested in schematics already built and works. I'm a student and my knowledge of class D amp. are very poor.

I notice that are a lot of users with very good knowledge of cass D amplifiers, but they don't show their designs/projects, even olders or abandonated.

Please share some of old project (no longer needed) or schematic built by companies, for helping us (unexpirienced users) in imaginate how to build an amplifier D. It's like u teach an child how to play football, but the child never see the ball and any play...only with words. It's only my opinion and i hope u don't feel upset or be angry with me.

[fredos]

Just start a new tread, this will give you more visibility...I will try to check old design I have built, but every body will need circad98 to see them...

Fred

www.d-amp.com

[Tekko]

An amp that large are not a good beginner project.

Im planning a class d amp myself but i sight for just earbud power to begin with.

I suggest you check out these "complete class d amp in a chip" ICs that just needs a few external components, or a kit. if it has to be a class d amp. Tho none of those ar even close to a kilowatt but you will never need more than 100w anyway.