Class AB vs DoubleCross vs ABBA @ Gm doubling compensation comparison

[jan.didden]

Perhaps we should first explain exactly what a WingSpread-Diagram is. I think some users don't know that.

Isn't it just the gain of the output stage at various output currents?

Jan

 

[alexcp]

Yes, the DC gain (which for EF is slightly less than 1.0) across a range of DC input voltages. Often, many diagrams, corresponding to various bias currents, are shown together:

(source)

Note that AC behavior, and in particular frequency dependent AC behavior, may be different.

 

[hbtaudio]

I completely agree with both of you.


#
my quick play with MC12 has confirmed (in my optimization view; while keeping the perhaps strange-looking parallel circuit) the most favorable result with Douglas Self's optimal bias diction: clearly Class B!


THD_1k is steadily falling and less than 0.04% at the beginning at mW.


HBt.

 

[hbtaudio]

Isn't it just the gain of the output stage at various output currents?

No Jan,
but in practice (view) it works in a very similar way 😉 .

 

[hbtaudio]

dy/dx as a function of x

vs

y/x as a function of x

 

[hbtaudio]

complete (?)

Bye,
HBt.


(As we all know, the problem is the zero crossing.)

 

[EdGr]

The key words are small-signal gain.

I had not seen small-signal modeling being applied to distortion until I joined this board. Computing actual distortion requires solving non-linear equations.
Ed

 

[hbtaudio]

You are absolutely right - and this is the key ...

Thanks Ed,


that's why I wrote: “I think, or rather suspect, that even this short fragment is not helpful in understanding the dynamic output resistance (EF with maximum local negative feedback through the load resistor).”


#
m(x, param bias)

Nevertheless, we should explain the dynamic internal resistance of “the source” - or the output resistance of “the EF”.
This brings us much closer to the term gm-doubling.


But once again to the equation systems to be solved - the zero is the problem!


greetings,
HBt.

 

[mvs0]

yes I did noticed that , but again at 1mW measured THD&N is 0,01% (still nothing special), anyway IMHO it will be much better amp if THD vs output power characteristic is flat (green line) , or is inverse to actual measured data (orange line) ,
btw , do you have any information what is the output level of generated RF signals at full output power ?

can you also plot the hearing threshold line for lets say 112db/w speakers at listening distance of 4m?

 

[TNT]

that would be a valuable additional data point what's going on with that dubious hash claim.

We must do it! :-D

//

 

[stocktrader200]

Class d will eventually win all. It has the advantage that at low power, it's accuracy is best. The tpa3255 should be capable of 0.004% thd set up correctly.

 

[mvs0]

can you also plot the hearing threshold line for lets say 112db/w speakers at listening distance of 4m?

Ok, did it myself.. 🙂

So why you ever need the green or even orange line is beyond comprehension for me.

 

[wahab]

When it comes to electromagnetical pollution this Purifi amp doesnt really deserve its name,
in some way those class D amps are a disaster, i mean, at 100W/8R there s 2.8V RMS of 500Khz residual,
so the distorsion is somewhat transfered at very high frequency where it is supposed to not matter,
set apart that it create a huge pollution of the AM radio band.


https://www.soundstagehifi.com/inde...asured-performance-of-the-eigentakt-amplifier

 

[ostripper]

That is why I keep my uCD-400 in it's metal case .
And only use it for my subwoofer where it don't matter.

I use my 50mA bias Class AB EF's , coolely running the Sanken outputs to actually
listen to my topping DAC...

OS

 

[hbtaudio]

I hope we will overcome the Class D dominance together, won't we?

Very gladly, I would love to - 'm probably a traditionalist.


Class A, B, AB, ABBA, D are these (nowadays) not all just questions of faith?


regards,
HBt.

 

[mvs0]

When it comes to electromagnetical pollution this Purifi amp doesnt really deserve its name,
in some way those class D amps are a disaster, i mean, at 100W/8R there s 2.8V RMS of 500Khz residual,
so the distorsion is somewhat transfered at very high frequency where it is supposed to not matter,
set apart that it create a huge pollution of the AM radio band.


https://www.soundstagehifi.com/inde...asured-performance-of-the-eigentakt-amplifier

Is this a measurement where the 1KHz is at max power? It does not say in the picture or review...

 

[jan.didden]

I hope we will overcome the Class D dominance together, won't we?

No chance in hell. In the real world, it's downhill for class (A)B since a few years already.

Jan

 

[banat]

That is why I keep my uCD-400 in it's metal case ...


OS

That `s not enogh for complete RF schielding since output power lines ( speaker cables) altogether with speakers crossovers and also the speakers coils alone is the part of the class-D amp constant RF emision in the medium AM band.

 

[banat]

One point to think about of those expencive class D amps is the fast&far in the field cheap and succesfull servicing in the case of the amp failure , IMHO in that case standard A(B) class amps still is the winners .

 

[wahab]

Is this a measurement where the 1KHz is at max power? It does not say in the picture or review...

The signal is on the y axys, it s at 0dB, the residual is at about -20dB in respect of whatever power.

No chance in hell. In the real world, it's downhill for class (A)B since a few years already.

Jan

That s because regulation authorities are not aware of the huge EM pollution, class D is comparable to cars
with very good efficency that would output big amounts of black smoke, if your neighbor has such an amp
you ll be forbiden to listen to an AM radio or your SW amateur radio receiver, i m surprised that you didnt spot
what is basically negating other s people s rights by what is litteraly an unlawfull appropriation of the EM
spectrum, not by using it but by rendering it unusable for others.