Why voltage swings to 2*Vcc when supply voltage is Vcc in class A amplifier?

[mona123]

Why peak to peak voltage swings to 2*Vcc when supply voltage is only Vcc in class A amplifier? Where does extra Vcc excursion come from? Will it depend on whether I use an inductor at collector or not? Thanks.

 

[FvM]

Will it depend on whether I use an inductor at collector or not?

Exactly. Consider two points:
- the inductor's stored energy is able to drive out above Vcc
- the average voltage across an inductor is zero, so the DC level at a class A output stage must equal Vcc. An output voltage swinging down to zero implies a maximum of 2*Vcc in return, assuming a symmetrical waveform.

 

[mona123]

Thanks FvM. Is there a book that explains this well? I would really like to understand it well. If you look at the ac load line it appears that if you bias it at Vcc quiscent if you change the load and hence ac load line slope you may be able to swing beyond 2*Vcc. Is it true? Thanks.

Exactly. Consider two points:
- the inductor's stored energy is able to drive out above Vcc
- the average voltage across an inductor is zero, so the DC level at a class A output stage must equal Vcc. An output voltage swinging down to zero implies a maximum of 2*Vcc in return, assuming a symmetrical waveform.

 

[FvM]

Assuming the output can't swing below gnd, it also can't go beyond 2*VCC for sine waveform.

 

[mona123]

Is there a book that explains mathematics behind it? Thanks.

 

[mfding]

How about this?

For example, for a BJT, collector voltage Vc=DC voltage @ collector + AC voltage @ collector

So for class A with RFC (or LC tank) at the collector:Vc=Vcc+Vb*gm*RL, where Vb is the AC voltage seen at the base. If Vb increases such that Vb*gm*RL reaches an amplitude of Vcc then Vc will be clipped.

At this point, the peak Vc is 2*Vcc and the min Vc is 0 (ideally) and the entire waveform of Vc is a sinusoidal of amplitude Vcc shifted (above) by a DC value of Vcc. (i.e. the peak is 2*Vcc and the min is 0).

 

[mona123]

Why Vb*gm*RL can only reach Vcc? ac small signal has nothing to do with DC . Thats what is confusing me.

 

[mfding]

Why Vb*gm*RL can only reach Vcc? ac small signal has nothing to do with DC . Thats what is confusing me.

ac small signal usually has nothing to do with dc but we are talking about large signal here as we are interested in the peak collector swing i.e. maximum collector voltage). so it is really "large" signal.

more specifically, because the minimum instantaneous collector voltage (which is the sum of DC and AC collector voltage) can not be below 0 volt (ideally). Otherwise, it will be clipped due to BJT device unable to conduct instantaneous current out of collector (instantaneous collector current is always into the collector under correct operation). when the min instantaneous collector voltage is 0, the amplitude of the AC voltage (for which "small signal voltage" is no longer a good term) is Vcc (this can be understood by the graphical description I mentioned in my previous post).

 

[mona123]

Hi mfding,

Thanks for the answer. I agree with what you are saying about collector voltage cannot go below 0 (more like VCEsat) and that it will be clipped but nothing prevents instantaneous voltage to rise beyond 2*Vcc. The assumption you are making is waveform is pure sinusoid. In that case too lets say a sinusoid can go to 3*Vcc, wont't upper side of waveform be unclipped and lower side will be clipped to 0 or VCEsat? I think now I understand that its us who force it to 2*Vcc peak with proper termination so that ouput waveform is sinusoidal for linearity in class A. Do you guys agree?

 

[mfding]

Hi mfding,

Thanks for the answer. I agree with what you are saying about collector voltage cannot go below 0 (more like VCEsat) and that it will be clipped but nothing prevents instantaneous voltage to rise beyond 2*Vcc. The assumption you are making is waveform is pure sinusoid. In that case too lets say a sinusoid can go to 3*Vcc, wont't upper side of waveform be unclipped and lower side will be clipped to 0 or VCEsat? I think now I understand that its us who force it to 2*Vcc peak with proper termination so that ouput waveform is sinusoidal for linearity in class A. Do you guys agree?

hi mona123,

yes, the assumption is the amplifier is still linear and the output collector waveform is a pure sinusoidal plus DC for a pure sinusoidal input. i agree with you.