Biasing Whithout VBE multiplier

[parsecaudio]

Hello to all DIY members.
A little ideas for biasing the output stage in power amp based on the base-emitter voltage across Q1 and Q2.
Any suggestion are welcome.

[parsecaudio]

a principle schematic

[deleveld]

Whats wrong with a Vbe multiplier? Your circuit looks much more complex. What advantage does your approach have over a Vbe multiplier?

Doug

[parsecaudio]

Thank you doug for your response.
The main problem is the difficult to compensate both PNP and NPN output transistors with a single npn (or,rarely, pnp).
In this manner,a better temperature compensation and more accurate biasing are made.
I remember that a circuit 1 is a complete output stage that can delivery(with paralleling transistor in ouput stage) more than 250w.

[darkfenriz]

Such a circuit is sometimes called 'diamond buffer'.
On the cons side it has bad clipping behaviour and of course bias is hardly adjustable.
Of course the principle is OK and often used for example in opamps.

P.S. If you really want complementary thermal compensation you can use complementary Vbe multiplier. I use it too and it works well for three stages of output followers, with diamond buffer you can use rather two or four, not odd number.

[parsecaudio]

Another little advantage is for mosfet biasing.
Some people says that mosfets have negative temperature coefficient(bias decrease whith increase of temperature).Looking at the datasheet of the major brandt of mosfet,this is not true ,or it is true for a restricted current area.
Whit this circuit the biasing for the output mosfets will be more accurate despite a conventional biasing circuit.

[parsecaudio]

Yes,darkfenriz.The bias regulation is more difficult for this circuit,but if the VBE (or VGS)on the first transistors are well selected the circuit work very good and it has no need to bias adjustements.For my experience,it is very difficult to obtain a temperature coefficient exactely tailored on output transistor.The greater part of the times,the circuits with vbe multipliers must be overcompensated to avoid the current avalanche.
Moreover a little variation in the biasing it can be made varying the current in two "current mirror" stages.

[mrshow4u]

Darfenriz say's "P.S. If you really want complementary thermal compensation you can use complementary Vbe multiplier. I use it too and it works well for three stages of output followers " I agree with that. I've seen the comlementary Vbe multiplier used to great effect on things like Parasound. I think John Curl liked to use the comp Vbe multiplier on all of his stuff. (Parasound, Levinson, etc)

[parsecaudio]

Ok mrshow4you.
You have a better ideas for biasing the mosfets?

[parsecaudio]

A little technical consideration over this circuit.
The main goal of this circuit is not the biasing control(this is only a secondary effect) but the fact that the gain of the amplifier is the same in the class A and B zone .Whit this the transfer functyon is linear and not crossover distortion are made.Referring to the attached image,if the output current is:
io=i3+i4=B1 B3 i1+B2 B4 i2.
If the gain of Q1 and Q3 is the same on Q2-Q4 we obtain
io=B id.
This is valid on entire user power range.The current amplification in class A zone is the same that in class B.
Another thing is that the bias current it does not take part to the circuit function.
In other words,varying the bias current it not vary the linearity of the circuit.
This result is valid,surely,only for this simplifyed model.
In the truth there are some negative effects,but the bias distortion are more low than a different schematic or a voltage follower.
I remember you that this circuit is a current driving(especially on the mosfet power amp i have attached)