Assemblage Power Amp

[Joachim Gerhard]

I mentioned paralleling before as a means of distortion reduction.
You did it more massive though like in some american amps.
What about the JBL triple ?
Some think it is not that stable.

 

[Joachim Gerhard]

Here is the Bryston.
Q5, Q6 can also be MosFets.

 

[Joachim Gerhard]

Here is the Roender.

 

[RCruz]

Must learn how to sim this wingspread....

I belive it is better now, could you sim again using different emitter resistor values ?

0.22 0.33 0.47 0.8 1ohm ?

 

[smms73]

I mentioned paralleling before as a means of distortion reduction.
You did it more massive though like in some american amps.
What about the JBL triple ?
Some think it is not that stable.

Normally big amplifiers sound better, and there is a technical explanation for that, paralleling power transistors gives lower output impedance less distortion and a more robust amplifier.

I don't know JBL triple, is the same of Locanthi T circuit ? If yes, then you right, adding the prédriver transistor makes the circuit less stable, but that is easily fix by using some capacitors and base resistors.

The Bryston , I never heard, do you know any circuit that has values in the resistors?

 

[Joachim Gerhard]

Yes, the Locanthi T.
I think it is used by Aire without overall NFB and the distortion is ok ( 0.1% ? ).
Somewhere i found a complete Bryston circuit. I have to search.

 

[Joachim Gerhard]

http://bryston.com/PDF/Schematics/3B-8BST_SCHEMATICS.pdf

 

[smms73]

Must learn how to sim this wingspread....

I belive it is better now, could you sim again using different emitter resistor values ?

0.22 0.33 0.47 0.8 1ohm ?

You can download the example circuit from Bob cordell site in figure 19.4

CordellAudio.com - Tutorial Simulations


In my personal opinion the great feature of CFP output stage is the good thermal stability, but to have that, you have to use a circuit like Bigun uses.

Adding emitter resistors makes the crossover distortion worse.

 

[FdW]

No problem Frans, I understand. Hope the best for your project.

The more linear output stage without error correction or feedback that I know is the parallel of a large number of power transistors in a emitter follower configuration like this image, the best bias is at 56mA per device or 0.39A total.

Doesn't that look good 🙂 now with halve the number of devices. I dream of these things.

 

[vzaichenko]

Here is the Bryston.
Q5, Q6 can also be MosFets.

I used this kind of topology in a number of designs with Sanken transistors (2SC2922/A1216) - like it a lot!
Low distortion, solid-sounding...

Attached is a fragment of my design with the values. I used it with +/- 52V rails.

Cheers,
Valery

 

[smms73]

Have to take a better look at the bryston.

just for curiosity this is the wing from the output error correction amplifier that I will publish in diyaudio this week. It will use error correction and TMC.

 

[FdW]

Have to take a better look at the bryston.

just for curiosity this is the wing from the output error correction amplifier that I will publish in diyaudio this week. It will use error correction and TMC.

That looks great, be sure that I will read your article with great attention.

 

[vzaichenko]

Just attached my implementation of bryston topology at the previous page 😉

[HERE]

 

[smms73]

That looks great, be sure that I will read your article with great attention.

Damir have start a new thread with a amplifier that uses my error correction circuit.

http://www.diyaudio.com/forums/solid-state/260308-gainwire-ngnfb-classb-poweramp.html

 

[Joachim Gerhard]

Valery, is that really a*2mH coil ?
I asume it is a 2uH.

 

[FdW]

Damir have start a new thread with a amplifier that uses my error correction circuit.

http://www.diyaudio.com/forums/solid-state/260308-gainwire-ngnfb-classb-poweramp.html

And follow that thread 🙂

 

[vzaichenko]

Valery, is that really a*2mH coil ?
I asume it is a 2uH.

You're right, it'a typo of course 🙂

 

[RCruz]

In my personal opinion the great feature of CFP output stage is the good thermal stability, but to have that, you have to use a circuit like Bigun uses.

Adding emitter resistors makes the crossover distortion worse.

So you mean I should use only collector resistors (to output) instead of emitter resistors ?

 

[Joachim Gerhard]

Bias changes, how much change is acceptable ?

 

[RCruz]

Working into class a at low listening levels 🙂