If I understand correctly, biasing a single-ended Class A amp "hot" doesn't change its gain at all, it just lowers its headroom (clips earlier).

What about a Class AB push-pull amp? Besides increasing heat and lowering component life, does running an AB amp "hot" have any effect on its gain? Doesn't increasing the bias level on an AB amp increase the time where both active elements (transistors or valves) are conducting at the same time? What occurs to the output signal when both elements are conducting and pulling current through opposite ends of a center-tapped output transformer? Don't they cancel each other out? Is the idea with eliminating crossover distortion to have one element slowly "fade out" (end its conduction cycle) just as the other one is beginning? Except at zero, the two should never be conducting the exact same amount of current, correct?

Thanks.

 

I am not sure what your saying here .. But I do know that a class A amp gives the best sound quality of all the classes and almost a perfect reproduction of original music hence why people love them .. The AB class is slightly worse than class A but not by much in sound quality . The problems comes where the class A amps run hot and usually pulls the most power and AB class was hybrid of the A class and B class for slightly worse sound quality but alot more power friendly..

 

What do you mean by "hot"? Excess current?

 

What do you mean by "hot"? Excess current?

Yes, exactly. Biasing "hot" is surprisingly called under-biasing.

Anyways, if we take a single-ended Class A amp with a supply voltage of +10 vdc, then we normally want to bias it so the output is around the mid-point between the rails or approximately +5 vdc. This way, we get maximum headroom or the most space for voltage gain before clipping, roughly +/- 4 volts if using a silicon based transistor. If we were to raise the bias above this mid-point, for example, to +7 vdc, then the most the signal can increase is 3 volts before clipping. Therefore, raising or lowering the bias (q-point) from the ideal mid-point only serves to decrease the amount of amplification possible before the circuit distorts (less headroom), but bias has nothing to do with gain, correct?

Now, consider a Class AB push-pull amp. If the bias is decreased to zero, the amp will operate as a Class B amp, complete with crossover distortion. If we slightly bias this amp to remove this crossover distortion, we have a Class AB amp again, but what if we continue to increase the bias? The active elements (BJT's, FET's, valves) will conduct more and more beyond 180°. Does this have any effect on the gain of the PP amp? Will it effect headroom? Other than less efficiency and shorter component life, is there a downside to operating an AB amp with excessive bias?

Thanks again.

 

Does this have any effect on the gain of the PP amp? Will it effect headroom? Other than less efficiency and shorter component life, is there a downside to operating an AB amp with excessive bias?

Nice explanation, BTW.

Basically, no, no , and no and possibly a benefit if you have golden ears. The part of a traditional audio power amplifier that sets the static current through the output stack and the part that sets the overall circuit voltage gain are independent. The output stack usually has a voltage gain of 1 and lotsa current gain. Before that is the voltage amplifier gain stage (duh) and before that is the input stage which might also have some voltage gain. There are a zillion variations, but that's the basic layout of most designs. Super ultra audiophile amps can be very different.

If you subscribe to the theory that an AB amp *always* has crossover distortion no matter how high the bias current is relative to the output current, then one school of thought is to increase the bias current so the crossover happens significantly away from 0 V. In this way the amp is class A for larger range of quiet passages where crossover distortion would be the most perceivable, and buried in the louder audio that moves the amp into true class AB operation; a class AAB operation, sorta.

If you are going to do this, like run 1 A of bias current in an amp that has 5 A of peak current capability, maintaining a constant bias current can be tricky because of device heating. A DC servo is more complex than standard diode or active-zener biasing, but more stable over all operating conditions.

ak

 

If you over bias push-pull you will enter into Class A zone. So you have a Class A amplifier with push-pull output stage.
The voltage of this gain push-pull is closer to one. And your output stage will enter into Class B (only one transistor conduct current) if Iload > 2* idle bias current. For example for Iq = 0.5A for ILoad > 1A your output stage enters into Class B (only one of the two transistors is active) and the voltage gain will slightly drop .

 

Ah, then in that case the effect of running it "hot" is reduced distortion.

 

More like shifted-to-a-less-noticeable-place distortion. Note that crossover distortion modifying like this is a peak or instantaneous current artifact, not RMS current. Also, the transition from class A to B can be smooth or abrupt depending on the types of output transistors and the circuit design around them, and probably is not reduced (or reduced very little) by hot biasing.

ak