Fixed "mistakes" on great sch :)

Somebody here posted sch for single supply stereo d class amp.I am new here and very long in this problematic so i saw instantly that he didnt post sch honestly but he cut 4 connections to make ti to not to work.However this is corrected sch.This Amp is really powerfull and high quality one.
 

Attachments

  • Single suply stereo class D.JPG
    Single suply stereo class D.JPG
    150.3 KB · Views: 355
Guys , its nice to comment about D -class .... but AFTER u build and try something.I have behind me at least 10 differents concepts and all r working great......on each concept i made minimum 30 to 50 power amplifiers.So i will try explain a few things.The only hum and noise can exist on output on this schematic only if u put them on input.ON and OFF concept is right one and good starting point to think.....I can tell u that this is good design and very powerful,easy to make,try and play with.Point is that somebody posted this here and posted it with mistakes which r now corrected,so u r welcome :) .
 
As your schematic shows no NFB path, any noise on the PSU lines will get straight to the speaker. If you added a proper frequency-compensated op-amp input stage with NFB taken from the switching side of your output inductor it'll make a reasonable amp.
I would be very wary of making foolish comments about "after you build something". There are quite a few of us here who are employed to design class-D amplifiers and know a great deal more about the subject than you will for a few years if you think the schematic you have posted is anything like the state of the art!
 
I can tell u that this is good design and very powerful,easy to make,try and play with.Point is that somebody posted this here and posted it with mistakes which r now corrected,so u r welcome :) .

I can tell you from building something very similar that it will have hum problems with no input signal.
Any hum on the power supplies will modulate the output signal.

It can also have problems when overdriven like the VS power supply dropping out and possibly blowing up the output mosfets.

Been there and done it.

You are better sticking to something like an IRS2092 with its built in fault tolerance.
 
Last edited:
I will not agree ..... srry :) .... this design is good one .... i didnt say best ever ...... This is single supply design if u didnt notice ..... On 24V to 50V it gives great power.....hum is discutable here ..... u can have hum becouse u made bad input regulation....Volume pot should be at the inputs on the sch and in front of them u should put fixed fg. x10 opamp.When pot is on max only hum which come to it will appear on output......if u short inoput on schematic there is no way that u can have anything on output.
 
Despite my favor for postfilter feedback designs instead of no feedback at all...

I think this amp will not have much of hum with shorted inputs.
It is a full bridge. Without signal both halves will run at 50% duty cycle and generate identical hum vs. GND. But the speaker is not vs GND, it is differential between both ends of the bridge and hum cancels out. That's one of the inherent advantages of full bridges.
Hum will enter the speaker signal at larger modulations, but in that situation most people won't care, because then the music signal is by far larger than the hum.
It is likely to become visibile in detailed measurements with fourier analysis, but I can easily believe that it is no headache under normal use conditions.
 
Hence why fullbridge is suitable for openloop designs wheres halfbridge is not.

Because:
Without signal both halves will run at 50% duty cycle and generate identical hum vs. GND. But the speaker is not vs GND, it is differential between both ends of the bridge and hum cancels out.

In a half bridge the speaker is between half bridge and GND ==> no differential connection of the speaker between of two identical hums ==> no canceling of the hum.
 
I will try to explain how i see hum ....

First of all i will not agree that full bridge create less hum than half bridge.As audio signals r amplified by 2 times or more in full bridge topolgy also is hum.Couse hums of the chanels r in oposite phases.

Second , and most important......hum is modulation which is transfered to speakers.....Modulation can exist only if it happens on input comparator.It can happen in normal conditions on few ways.

1. If triangle voltage has floating while input is zero......
2. If voltage which is setting the 50% duty cycle is not super stabilised
3. If comparator has ulta super small input current (most of the cases).Then even smallest variations can make hum (which is high frequency modulation)
4. If input audio pre amp is not well compensated
5. If voltage for triangle generator and comparators (read PWM section) is not well regulated.
 
Second , and most important......hum is modulation which is transfered to speakers.....Modulation can exist only if it happens on input comparator.It can happen in normal conditions on few ways.

Definitely no.
The concerns in this thread are not about hum which enters the speaker signal through the modulator.
People are trying to point to the hum which enters the signal through the switches directly from the power rail.

Just do the most basic calculations:
1. Set a constant and perfect idle duty cycle of 50%.
2. Assume some portion of AC voltage on the power rail.
3. Apply Kirchhoff's voltage law to calculate the voltage across the speaker.
==> You will see the advantage of the full bridge.
 
O yes , :) (I like this conversation) ..... We r discussing about most important things here....not related to the schematic that i posted.....And i like the discussion.I am sure that we all will learn.

Full bridge is doubling output voltage and making 4 times higher power on output.....mathematicaly.....in real world that is power boost of 2.If power is boosted then hum and noise r boosted too.U didnt read slowly what i wrote in my post.If u have AC on power rail.....and this is real case......then if u alow to that AC to be present on your DC which is on low side of amp u will have triangle signal floating and most important thing DC voltage which is setting 50% will float too.Full bridge is huge advantage to get super high power boost but only if DC on low side is perfect.Also any input preamp stage should be low impedance input (audio side) and not high imepdance (that combined with gain of 10 fg. will create hum all the time.

This tehnique (D-Class) is state of art of progressing in electronics,possibilities r unlimited.I am looking forward to hear what u guys think about this.