Hi there, all you Class D experts.
I need some advice on a Class D audio amplifier I'm designing for my car. I've build the switch-mode power supply already, which outputs +-35V rails. I'm gonna use an H-bridge for my output drive
What Class-D modulation is best suited to car audio? The reason why I ask, is that you get your so called "standard" Class D design, which is basically a triangle/ sawtooth wave that goes through a comparator with the audio signal. The audio signal would go to your non-inverting input of your one comparator and the inverting input of the other comparator to give you the 180 deg phase shift (hope you still with me). This will drive the positive and negative side FETs respectively.
Is this an effective way of driving an H-bridge? I'm talking in terms of eficiency only, as I'm not too worried about audio fidelity.
What I'm thinking of, is driving only the positive side FETs if the audio signal is positive and the negative side FETs if the audio signal is negative. That would increase my efficiency and maybe my fidelity as well, as the switching is alot less than conventional.
Thanx🙂
I need some advice on a Class D audio amplifier I'm designing for my car. I've build the switch-mode power supply already, which outputs +-35V rails. I'm gonna use an H-bridge for my output drive
What Class-D modulation is best suited to car audio? The reason why I ask, is that you get your so called "standard" Class D design, which is basically a triangle/ sawtooth wave that goes through a comparator with the audio signal. The audio signal would go to your non-inverting input of your one comparator and the inverting input of the other comparator to give you the 180 deg phase shift (hope you still with me). This will drive the positive and negative side FETs respectively.
Is this an effective way of driving an H-bridge? I'm talking in terms of eficiency only, as I'm not too worried about audio fidelity.
What I'm thinking of, is driving only the positive side FETs if the audio signal is positive and the negative side FETs if the audio signal is negative. That would increase my efficiency and maybe my fidelity as well, as the switching is alot less than conventional.
Thanx🙂
I wouldn't do this because the small increase in efficiency isn't worth the signal degradation incorporated. Keep in mind that doing this is almost the same as class AB with conventional amplifiers.
Regards
Charles
Thanks for your reply Charles.
If a class D amp is powered to about 30-40% of its maximum power, your efficiency will decrease considerably, because you still switching maximum current through your load. That's why I think that if you switch the one side of your H-bridge only for the positive side of your audio input signal, and the other side for your negative input, you actually double your efficiency.
I'm not really too fussy about my audio fidelity as it is for car use. I want to be able to play my music for long periods without having to run the car in idle, so ifficiency is more important.
Thanks
If a class D amp is powered to about 30-40% of its maximum power, your efficiency will decrease considerably, because you still switching maximum current through your load. That's why I think that if you switch the one side of your H-bridge only for the positive side of your audio input signal, and the other side for your negative input, you actually double your efficiency.
I'm not really too fussy about my audio fidelity as it is for car use. I want to be able to play my music for long periods without having to run the car in idle, so ifficiency is more important.
Thanks
Efficiency is only slightly increased with such a modulation scheme due to not needing to drive the "opposite FET's" gates. You will still have some conduction losses because the filter inductor's freewheeling current will have to flow somewhere while your FETs are not conducting.
Regards
Charles
Regards
Charles
Isn't that what the internal reverse diodes are for in the FETs i.e. to clamp the freewheeling current from my load? So even if my FET's not conducting it will still clamp my reverse current. I'll have losses, because I don't use my inductive load's reverse current, but I gain wrt power supply drain.
Thanks
Thanks
No, the intrinsic diode in a FET clamps reverse current so there aren't large, damaging spikes on the FET when it is driving inductive loads.
What Charles is talking about is the current that is flowing through your output filter inductor - it will still be stored in the L even when the FETs are all turned off. It will have to go somewhere. Think about where current is flowing - it goes back and forth through the inductors every cycle.
What Charles is talking about is the current that is flowing through your output filter inductor - it will still be stored in the L even when the FETs are all turned off. It will have to go somewhere. Think about where current is flowing - it goes back and forth through the inductors every cycle.
I want to design something similar to the Class BD design by Rockford Fosgate. It uses a suppressed carrier modulation technique. So at low power or if there's no audio input signal, it gets rid of the close to 50% Duty Cycle squarewaves flowing through your filter.
What do you think, is a design like this out of my league?
Thanks
What do you think, is a design like this out of my league?
Thanks
Evan Shultz said:
Well, that current actually goes through the intrinsic diode, but in the **other** Mosfet (not the one that has just stopped). Problem with this scheme (I built an amp some years back) is the pulse width -> recovered voltage transfer is not linear for duty cycles below 50%. It can be worked around, but at the cost of loss of efficiency, just what Pokka doesn't want.
Rodolfo
- Status
- Not open for further replies.