Common mode canseling in symetric amp

[Konrad]

I have been strugling for a while designing a symetric amp.
I find that "slone's" innput amp with dual mirrors suffer bias stability in the VAS stage.
Comparing the slone with Robert r. Cordell (?) s amp that has mirrored VAS stage and some form of VAS bias feedback to 2 current sorses loading the differensial innput-stage.
This comparison then tells me: Super symetri neads some form of common mode feedback to cansel the differensial in the circuits halves. To do this i then figure the output is best when bridged. Then the common mode can bee differensialy ( and symetric ) feed to the symetric (doble differensially connected) VAS stage.

Common mode feedback to a VAS stage may sound strange as a expression, but common mode feedback to a true differensial symetric VAS is possible with "BRIDGE" output. The symetry then becommes real symetry.
Adding the from now on called the common mode canseling stage
Comparing to a more known circuit from Rotell or Borbely-Lender, all symmetrical types, we ned to add Vas transistors, bias, driver and output. The sum of the to outputs is then the common mode signal whitch can bee feedback to the VAS stage like adjusting the bias point continuisly, by a symetric differensial stage ( the common mode canseling stage )

The same way it can bee fedback to 4 current sorses (that loads the innput stage).

The true/real symetric amp (thus bridged) with added common mode canseling, and current sorce loading for the input stage. Someone out there figured the circuit yet ?
It's already on the paper. It's not a simple amp!
Then using so many parts i see no reason not add output stage distortion nulling/canseling as well
Advantages can bee lowering the suply voltage for the same output. Then we can select components from a wider amount of devices, j- fets and sucth for low noise input stage without kaskoding, and still make a farly high power amp.

[Konrad]

The atacced shows a bit and a half more than intended but will hopfully clearly show the common mode rejektion stage, (the one not so obvius at first glance).

[thanh]

Can you give me a better photos?

[cunningham]

Question...Do you think it is better to have two differential amps and two VAS stages that operate in class AB, or just one differential amp and one VAS operating in class A, biased with current source?

I have not been able to achieve as good results and stability from a completly symetric class AB VAS then from a class A, current source bias. You can still make a large power amp from the Class A biased VAS.

[Konrad]

Originally posted by cunningham
Question...Do you think it is better to have two differential amps and two VAS stages that operate in class AB, or just one differential amp and one VAS operating in class A, biased with current source?

I have not been able to achieve as good results and stability from a completly symetric class AB VAS then from a class A, current source bias. You can still make a large power amp from the Class A biased VAS.

With two differensial amps that operates with two VAS, add on two output stages ( it becoms two amp's). As i see it: If bridged the difference between a fully symetric amp is the common mode feedback wich can bee redused with ease when designed fully symetric. Common mode feedback is implemented in the opa1632 (? / op 1632 ?) fully symetric opamp.


Common mode im two bridged compared to common mode in a fully symetric amp with common mode feedback, and where i can choose common mode feedback ratio the choise is easy.

This common mode doesnt do any harm to the speaker (idealy) but it will move the operating point for the input stages and therby 'indusing' distortion.
AB vvs A vas... well i think it's not the place to look for great power savings in a 'classic' semiconductor amp, and like yuo say there is some drawbacks, and somtimes requier moore components as well.