Crossover distortion

[lumanauw]

When I read about theories of output stages, they say that classB (and AB) power amp suffers a "Crossover distortion". With my limited knowledge, I accept it as happening when NPN and PNP output stages will be 1=on and the other =off. The transition between on and off causing this crossover distortion.

I assume this happens because each transistor will be off and on. If we can make all transistor never off (like classA), maybe this crossover distortion will dissappear. (Is this right?)

I'm looking for a way to always have both transistor on. Someone said there is a patent on "never turning off" in class (A)B, but I couldn't find it. Does anyone know where it is?

I think about this drawing that I attached. This is complementary EF output stage. If I put D1 and D2 after emitors, before output, and put 10mA CCS to rails, this way both T1 and T2 will never turn off, minimum will always have to pass 10mA in every condition.

I have question about this idea.

1. T1 and T2 will never turn off. But can this idea eliminate crossover distortion in class(A)B amp?

2. But it makes the D1 and D2 conducting/not conducting. Does diodes in this mode gives another kind of on/off distortion? If it gives "diode distortion" that is worse than "transistor distortion" then this idea is not good.

3. If this idea is working, what kind of diode is suitable for D1 and D2? Ordinary diode/Schottky/Fast recovery/Soft recovery?

[Eva]

Diode reverse recovery and turn-on delay, by themselves, are much worse than crossover distortion. Schottky diodes may be an interesting option

[Steven]

Originally posted by lumanauw

I'm looking for a way to always have both transistor on. Someone said there is a patent on "never turning off" in class (A)B, but I couldn't find it. Does anyone know where it is?

http://www.diyaudio.com/forums/showt...339#post282339

S.Tanaka, "New Biasing Circuit for Class B Operation," JAES, 1981, Vol. 29, No 3, p. 148 (patented)

Nelson Pass: patent # 3,995,228


Steven

[Steven]

Before I forget, also the Hawksford error correction circuit does not switch off the output transistors.

Steven

[lumanauw]

Hi, Steven,

Thanks for the link. You also think the same too (but for classA?)

Nelson Pass: patent # 3,995,228

Do you have a small JPG on this patent?

[peranders]

Originally posted by Eva
Diode reverse recovery and turn-on delay, by themselves, are much worse than crossover distortion. Schottky diodes may be an interesting option

I have done similar experiments and I got real nasty crossover distortion => not a good idea

[uli]

Originally posted by lumanauw

Do you have a small JPG on this patent?

Source: www.passdiy.com
www.passlabs.com


Uli

[lumanauw]

=> not a good idea

Thanks for the input.
I see passdiy and passlab website, but couldn't find patents by Mr.Pass.
Where can I see all his audio patents?

Threshold series amps are marvelous in this area. Many kind of amp, with different cct, but same purpose. NP incorporated EC like cct, but he says its not for EC. Now I know what they was intended. To eliminate Xover distortion.

[darkfenriz]

What is so-called class AA?
I heard that Technics used it

[Steven]

Originally posted by lumanauw
Hi, Steven,

Thanks for the link. You also think the same too (but for classA?)

Do you have a small JPG on this patent?

Type the patent number in here: http://patft.uspto.gov/netahtml/srchnum.htm

or look here: http://l2.espacenet.com/espacenet/vi...b&LG=en&DB=EPD

Steven