I would want to build a class GH.5 amplifier with Class D Subwoofer output.

Multiple rails that have variable level depending on total power output (Volume) needed, and a PWM Powersaver for Bass.

Stacked Amps over a Raspberry.

 

Hello,

What you just did is hijacking an existing thread:
http://forum.allaboutcircuits.com/showthread.php?t=40398

Now you have a thread of your own, with all attention for you.

Here is an article from the ESP website on class G amplifiers:
http://sound.westhost.com/articles/class-g.htm

Bertus

 

Sorry about that. For a long time I have been thinking of an audio amp project and felt sure I am not informed enough to get really fussy high fidelity, but I thought I might be able to do LOUD.

Ever since I saw switching mode Amp in a Circuits book about 20 years ago I knew I would not want to go with Class A.

In the original thread I was just looking at the link to classes of Amps and from there the old ideas started to lurch around like Zombies and I thought out loud - by typing in a reply.

So I wanted to build about 450 watts of Amplifier plus a 450 SubWoofer Amp. I doubt I will actually build it but I have this urge.

 

Just to be clear - I have nothing against High Fidelity - but I do have a pretty much tin ear. If I could build something with 1% less Distortion it would just be a number on a meter to me. Loud is something I can notice.

 

There are many Class D kits, which I would strongly recommend as a first step towards learning how to build, troubleshoot, and later design your own class-D amp.
I have built kits from www.41hz.com ; they are good quality, although the documentation is a little lacking. But you can google and find many other kits.

 

A super simple Class D amp is one of my future projects for the 555 projects. They can be as simple or complex as you want them to be.

 

These are two different amps that I am thinking about building.

The Class D Amps are not my main focus. I wanted them to work as a complement to the other amp and unload the Bass Power. A kit would be a great idea for them. I am almost ashamed to say I was looking at International Rectifier Reference Amps. Buying a Reference Amp would be absolutely cheating but I was considering it.

https://ec.irf.com/v6/en/US/adirect/ir?cmd=eneNavigation&N=0+4294841614

I think working out one amp will be headache enough. I am going to drive myself crazy enough with the title GH.5 amp.

 

Thinking about something like this is fun.

I have a bad ear for tone but I can appreciate Low Noise and Dynamic Range.

I don't want it to hiss or pop.
About range - Loud I already mentioned but I want a system that whispers even better than it screams.

I even thought of making the systems Center Amp class A. The class G would allow a Class A in the center I think and reduce the inefficiency by keeping it closer to a rail.

I probably won't do that, but it was fun to think about. It will be a half bridge. I am thinking I want the center rails to be ±7VDC.

I am thinking of going nuts with the other rails. I might have jump rails above the ±7V center amp rails at +14, +28, +42, and then cut all the way to +98V and use those other rails as dropping rails on the way to the 105 Volts Peak. Having so many rails means the switching speed will need to be high. I am thinking somewhere between 800kHz and 4MHz. Around 20 to 100 times higher than the top frequency that I care to amplify because I need to switch so many times during each half cycle. I need to think about whether that High Speed trouble is going to cost me too much sanity.

I should work out what my power output would look like. I am thinking also that I want to build for a 2 Ohm to ground output load, so I can build an opposite phase amp to work full bridge and I can use 8 Ohms, or 4 Ohms as long as the signal levels don't average too high. That would be 420 Volts peak to peak. Peak Power would theoretically be 11Kw and average power would be 5.5 Kw.

That is not realistic for any kind of long term output. Derating output power average to around 40 Volts RMS would still allow the 400W that I want. The extra power would be there for any need to hammer out a few beats. I am old enough to know that I am unlikely to want more than a 2 seconds of ear bleed power. After that it is just noise, time to buy a hearing aid, and blown speakers. But that 2 seconds is enough to make your heart beat faster.

400W would be massively loud as a signal average.

I am wondering about a speaker project too.