New Output Stage?

[bimo]

I sim Baxandall super pair as driver on output stage.
It have good clipping, but slew rate reduce about a half compare to CFP driver.
THD is almost same.

How to improve this output stage? Is it stable?

 

[bimo]

Reduce THD at 20kHz and increase slew rate.

 

[keantoken]

The KSA/KSC predrivers don't do well at low Vce. I suggest the BC5xx or BC3x7 instead. They are not in any danger of overvoltage.

 

[bimo]

The KSA/KSC predrivers don't do well at low Vce. I suggest the BC5xx or BC3x7 instead. They are not in any danger of overvoltage.

OK, i will try it. You think it can work in reality?

 

[keantoken]

I'm sure it can work. I've done a lot of stuff in simulation with this kind of OPS, but I haven't built anything with it yet.

 

[bimo]

Change pre driver using BC550/560.
It reduce gain margin to 12 but reduce THD too, slightly.

 

[bimo]

I never seen PSRR like this before 😀

 

[TonyTecson]

Crown and Bryston used something similar some years back....

 

[Mihkus]

Whats so fancy about it? Its nothing new but definitely good performance.

 

[bimo]

Whats so fancy about it? Its nothing new but definitely good performance.

It is cool and new for me as newbie 😎😀
So, I ask who have more experiences than me.

 

[AndrewT]

The two peaks (58Hz & 40kHz) probably indicate the need for some compensation of a local loop are two.

I don't know where to point you to find the errant loop, but an RC for pole and zero is probably needed for each resonance.

 

[UltimateX86]

Why not using this super pair on output transistor ?

 

[bimo]

The two peaks (58Hz & 40kHz) probably indicate the need for some compensation of a local loop are two.

I don't know where to point you to find the errant loop, but an RC for pole and zero is probably needed for each resonance.

This is not the frequency response but PSRR.
There are no peaks at the frequency response and it have good phase margin and gain margin.

 

[bimo]

Add R C to smooth high frequency (just in case). This reduce PM and GM slightly but THD and slew rate is not changed.

 

[bimo]

Why not using this super pair on output transistor ?

It difficult to set the bias.

 

[bimo]

Schematic.....

 

[AndrewT]

Add R C to smooth high frequency (just in case). This reduce PM and GM slightly but THD and slew rate is not changed.

the closed loop frequency response and phase tells us nearly nothing about stability.
Go back to your PSRR.

 

[AndrewT]

Schematic.....

15.38Mpixels is a bit much for my screen.
How about ½Mpixel to 1Mpixel pic of the sch, depending on what you think is legible?

 

[bimo]

15.38Mpixels is a bit much for my screen.
How about ½Mpixel to 1Mpixel pic of the sch, depending on what you think is legible?

Reduce the resolution, but you can "zoom out/in"

 

[bimo]

May be this is tell you about the stability.