Auto-Biasing Circuits for Complementary Followers

[cbdb]

Dosnt matter how high the bias starts or how big the caps are. If the signal stays on long enough, a full cd for example, the caps will increase there voltage and increase the bias. The louder it plays the higher the bias goes. What other tricks?

 

[indra1]

Go to the Official M2 thread and try to learn from the first few pages.

 

[Nelson Pass]

Dosnt matter how high the bias starts or how big the caps are. If the signal stays on long enough, a full cd for example, the caps will increase there voltage and increase the bias. The louder it plays the higher the bias goes. What other tricks?

What circuit does that?

 

[Zen Mod]

aha

that's why I'm enjoying just first side of CD , with my N2

second side is almost unlistenable

 

[Mark Johnson]

And let's not forget that the nice people at Linear Technology sell an integrated circuit called LT1166, which auto-biases the output stage of an audio power amplifier. (link to document). It can be used with BJT output transistors, or MOSFET output transistors.

 

[indra1]

Thank you Mark, great tip! Affordable and available in easier to see DIP8. Specified for class AB bias but I see no reason why it can not be used for class A. Typical PassDiy bias of 1.3A will need R sense of 15.38 mΩ which can be 6 parallel of 0.1 Ω resistors or a 0.5 inch length of AWG#26 magnet wire.

 

[Zen Mod]

naah

far from elegance and wit , we are having with opto and hall chips

 

[indra1]

Totally agree, but a tip on other method is always great nonetheless.

 

[EUVL]

In terms of circuitry the LT1166 is more complex.
But it has one big advantage, namely the sense voltage is +/-20mV, instead of ~1.2V for the opto.
Once can argue this is not important for certain designs, such as those with sufficient NFB.
But if one chooses to design an amp with no NFB, then even a bias of 2A would mean Zout > 0.15R, or damping factor < 48.

If we accept +/-0.6V as sense voltage, then the JLH biasing circuit is IMHO just as effective.
It essentially operates on a similar principle.


My personal opinion, of course,
Patrick

 

[indra1]

In terms of circuitry the LT1166 is more complex.
But it has one big advantage, namely the sense voltage is +/-20mV, ...

Not against 1.3mV sense at 2A bias required by the ACS723 hall chip.

 

[Zen Mod]

of course , even if I don't like some solution in entirety , there is always something to learn from

 

[Nelson Pass]

And let's not forget that the nice people at Linear Technology sell an integrated circuit called LT1166, which auto-biases the output stage of an audio power amplifier. (link to document). It can be used with BJT output transistors, or MOSFET output transistors.

I found it very difficult to make it work without oscillation. I invite anyone with
a working example to share it, saving others needless frustration.

 

[EUVL]

Maybe this is of interest to you :
Problem with Bob Cordell's LME49830 and LT1166

Or maybe Bob has some experience also ?


Patrick

 

[indra1]

Many thanks Papa and Patrick.

 

[Nelson Pass]

Bob is certainly the guy who would have made it work. I screwed around
with it for a while, and then decided it was too much trouble.

 

[EUVL]

Amazing that the engineering world keep reinventing itself.

Essentially the JLH auto-bias from 1985 :
RMAF Afterthoughts


Patrick

 

[franzm]

What about the Allison circuit? Page 15/577 from here.

 

[Nelson Pass]

And you might enjoy the Aleph 0 from 1991.

 

[Circlomanen]

This is a neat auto-bias from TubeCAD that I like. (very simplified for clarity)

Class A only, but can have a very low distortion and be very stable and nice behaved. It is easy to add positive current feedback for a negative output impedance if wanted.

 

[Nelson Pass]

That's very nice, consistent with the quality of the stuff on TubeCAD.