150W Is it good?

[pjacobi]

Hi Sreten, All,

Originally posted by sreten
[...]
Given the balance of each LTP pair is not exact the actual VAS
current is not exact, it cannot be set as accurately / predictably
as in the single ended case.

This on its own is not a big deal, but the thermal stability of
the VAS current may be depending the biasing scheme used.
[...]

I also had this impression, but running LTSpice for temperatures from 15 to 40 degrees Celcius gave less than 15% change. Not really problematic, I assume.

I'd suggest some other points should be looked at:

- If you aren't pretty sure about your signal source, perhaps a follower should be put before the input, so that the current summing works exactly.

- Output stage: no comments on my post #69?
http://www.diyaudio.com/forums/showt...152#post326152

Regards,
Peter Jacobi

[millwood]

Originally posted by pjacobi
- Output stage: no comments on my post #69?
http://www.diyaudio.com/forums/showt...152#post326152

Regards,
Peter Jacobi

that's the famous T-circuitry. Leach used it in his design based on various simulations, but he offered no explanations.

I have simulated quite a few driver designs and found the T is always the winner. If someone can explain that I am all ears.

the cap is there to reduce switching distortion.

edit: I will offer one conjecture: the drivers in the T remain in class A throughout the cycle. they never switch off, unlike drivers in the traditional circuitry. but I am not sure how much of a big difference cross-over distortions at small signal levels will be.

[Eva]

My thoughts about the output stage :

The "T" circuit [drivers without connection to the output] produces consistent and fast turn-on/turn-off to the output devices, allows the drivers and pre-drivers to work in full class-A and reduces THD but at the expense of increasing current gain of the output stage and reducing phase margin

In the straightforward circuit [all transistors tied to the output], the drivers and the output devices are degenerated through its emitter resistors, having lower gain and providing more phase margin at the expense of class AB operation of the drivers and pre-drivers, higher THD and slower turn-on/turn-off of the output devices

[Wonders are also said about CFPs, but what happens when you drive a CFP into clipping at 20Khz?... When the output reaches the higher rail the lower devices may be still 'trying' to turn off... ]