low crossover-distortion class AB output stage topologies

Well, as I said, nothing comprehensive. A couple of allusions by Nelson to his non-switching patent which I was going to discuss here. A reference the the D. Self bias circuit which in my eyes is just an extraordinarily stable static bias generator, i.e. in its published form it does nothing to provide dynamic bias. The rest is all bits and pieces with passing reference taken to this or that design which is why I thought a dedicated thread might be justified.

Greetings,

Eric
 

gromanswe

Disabled Account
2002-07-08 8:42 am
North
www.google.com
That is an interesting Challenge!

I have, with pen and papper,
tried to find a satisfactory solution,
that never lets one transistor's Current go below a certain point,
while the other is conducting output current.

Of cource this is possible, but I have tried
to find a so easy and simple way as possible.

Not found it yet.
I'll tell you when I have.
Haven't dealt with that problem for a long time.
 
Cross over noise is absolutly unacceptable by my standards.
I will post it later.:rolleyes: :devilr:

Crossover distortion is not just noise, but also linear distortion that sounds so disgusting if not reduced to acceptable scale.
Solution: Feedback Error Eorrection, Feedforward EC, and many other.
Currently I am using 30-70mA feedforward with crossing control, and Negative Feedback for main output stage (its still drift 40mA when cool 100mA when hot, I am going to fix it again sometime, I don't like thermal drift).

May be he forgot to post it.:confused:
 
Some in EW & WW :

Margan (non-passing branch becomes a CCS )
Jager (output using common emitters)
http://francis.audio2.pagesperso-orange.fr/C31_Jager.gif
Renardson (feed forward)
Audio Amplifier Circuit Design
Van de Gevel (common mode feedback loop)
Jager (output using common emitters)
http://francis.audio2.pagesperso-orange.fr/C31_Jager.gif
Was that topology from the article
"A new Class AB design"
from the authors Wim de Jager,
van Tuy & van der Ven (University of Twente)
Electronics World, December 1999, Page 982 ??
If yes, I am looking for this article because I don't find this magazine, that I bought 1999 at my local railway station book-shop (maybe misplaced or borrowed and not get back).

Additional questions:
1) Are this topology to find in commercial audio amplifiers?
2) What is the colloquial term of this Class AB topology?

By post #32 about
http://www.diyaudio.com/forums/solid-state/92412-why-collector-drain-output-stages-so-rare-4.html
I have this found:
More information on the above amp :
Wim de Jager, Erik van der Ven, Ed van Tuyl
ELECTRONICS WORLD December 1999 pp. 982 – 987
"This class-AB common-emitter power amplifier incorporates a new current-mode class-AB driver circuit to obtain good thermal stability of the quiescent current in the output stage. It also guarantees non-zero currents in the output transistor that is conducting the residual current, avoiding HF switching distortion. Maximum output voltage is near to the rail to rail limit. Saturation of the power transistors is avoided, resulting in fast recovering from clipping. The circuit has an excellent stability due to a phase margin of 85 ° with a b 1/34"

Perhaps one of the members knows an URL for download this article as PDF file or can post a scan of this article.
 
Last edited:

Ian Finch

Member
Paid Member
2010-04-11 4:22 am
Coffs Harbour, NSW
Thank you for posting this article reference Edmond, very interesting.
One is tempted to think that Jager's scheme works very well at reducing crossover distortion.
So, as I don't see wide use or even adaptation, what is the catch - or perhaps difficulty that deters us?

Would anyone familiar with this design care to comment?
 
Last edited:
Thank you for posting this article reference Edmond, very interesting.


You're welcome.

One is tempted to think that Jager's scheme works very well at reducing crossover distortion.
So, as I don't see wide use or even adaptation, what is the catch - or perhaps difficulty that deters us?
..........

Years ago, I've simulated this amp and as far as I can remember, results were not that spectacular. At the high end of the audio spectrum distortion AND cross-conduction rises sharply.

I'm not aware of any commercial power amp using this topology, but according to the references, it seems that this technique has found application in IC based op-amps.

Cheers,
E.
 
[FONT=Arial, sans-serif]Francis Brooke's simulations of Jager's amp :
[/FONT]

[FONT=Arial, sans-serif]53 Weff/5.5ohm[/FONT]

[FONT=Arial, sans-serif]H2/H1=-81dB and H3/H1=-88dB for Vout=4Vc à 1kHz for a load of R=5.5ohm[/FONT]

[FONT=Arial, sans-serif]Vd out : 3mV at 27°C (Iq = 100mA) and 2mV at 75°C (Iq = 110mA)[/FONT]
[FONT=Arial, sans-serif]PSRR : 59dB[/FONT]

[FONT=Arial, sans-serif]Square wave 1kHz : overshoot 8% for a load of 8ohm//0.1µF and 4% for 8ohm//1µF[/FONT]

[FONT=Arial, sans-serif]Zout : 2,7 mOhm [/FONT][FONT=Arial, sans-serif]for a load of 5,5 Ohm[/FONT]

[FONT=Arial, sans-serif]Good thermal behaviour but low maximum slew-rate and output transistors need to be fast.
[/FONT]

[FONT=Arial, sans-serif]
[/FONT]
 
You're welcome.
Years ago, I've simulated this amp and as far as I can remember, results were not that spectacular. At the high end of the audio spectrum distortion AND cross-conduction rises sharply.Cheers, E.
The disadvantage of BjT power devices in that topology is the smaller value of ft. Therefore the differences between class AB and pure class-A in the upper range are even greater than in normal power follower circuit.
by post #42 (URL's about Abacus and LEF) about
http://www.diyaudio.com/forums/solid-state/92412-why-collector-drain-output-stages-so-rare-5.html
there are other advantages mentioned, but this advantages more useful in the low frequency aera (causes impedances with inductive character and thus phase shifting effects between voltage and current).
For me is the conclusion, that this topology is to prefer for "high power" subwoofer power amps with low efficiency speakers like JL audio's "13W7AE", where the desire of a high quiescent current stability must be present and no unwanted audible effects occur through phase shifts arround the resonance frequency from the drived bass loudspeaker system (by use of power follower output stages often a problem without impedance compensation network).
What about Linn's solution, realized by the model "Klout" ? - have a look to the attachement by post #42 about
http://www.diyaudio.com/forums/solid-state/92412-why-collector-drain-output-stages-so-rare-5.html
I'm not aware of any commercial power amp using this topology, but according to the references, it seems that this technique has found application in IC based op-amps.
Do you know any types?
 
Last edited: