Discrete class D amplifier outputs low volume

[Togomori]

Hello everyone,
I have a discrete class D project that I need help with but since I'm not an analog guy I got a problem. So far I got everything set up and the discrete class D works perfectly. But only at maximum of 6 volts otherwise the Transistor M3 gets really hot and burned, how do I solve this problem? I want to increase the voltage in order to output a higher volume sound from the speaker. (Please find the circuit on the attachment)

I tried taking off the cap and resistor across the two transistors, it didn't work.

Any suggestions?

 

[Svein]

I have several problems with your schematic. Is it something you have designed, or have you found it somewhere?

By the way a class D amplifier is digital, not analog.

 

[Svein]

An analog amplifier is much simpler. Take a look at http://www.learningelectronics.net/circuits/60-watt-audio-power-amplifier-circuit.html or http://www.ecircuitcenter.com/Circuits_Audio_Amp/Advanced_Amplifier/Advanced_Audio_Amplifier.htm.

 

[davenn]

By the way a class D amplifier is digital, not analog.

That's misleading and a common misunderstanding. Tho they can be used for digital use, PWM etc
they are also commonly used as audio amplifiers with efficiencies better than Class A, B or AB amplifier configurationsDave

 

[Baluncore]

Welcome to PF.
The C and R snubber across the MOSFETS are needed. They are not the problem.

For M3 to get hot it must be conducting while M2 is also conducting. M3 gets the current at the same time as voltage because M2, being N'chan, is usually stronger than the P'chan M3.

This is probably a bias problem, but it could possibly be an ultrasonic oscillation of the MOSFET gate(s). You should try between 4R7 and 15R between the complementary emitter follower and the gate of the MOSFETS. That will usually quieten parasitic gate oscillations during transition.

If the circuit diagram was better quality it would be easier to analyse the situation and refer to parts by designation.

After anti-parasitic Rs are installed, test it then if not improved, the next step would be to test bias voltage levels.
Ground the junction of R15? (27k4) and D3 to kill all base drive so both high and low side MOSFETs turn off.
Does M3 still get hot ?

 

[donpacino]

do you have a datasheet for the mosfet?