Hi:

Attached is a sim for a Class AB amplifier with a single 40 volt supply. I get it that
Vcc / 2 = 40V / 2 = 20V output is not a realistic expectation. However, is an 11.7V
peak output realstic? How would a person estimate the real output?

David

 

Haven't we already discussed this?

IL_max ≈ (0.5Vcc - Vbe3 )/( RE1/(β3 + 1) + RE3 + RL) ≈ (20V - 0.7V)/(390Ω/81 + 0.5Ω + 8Ω) ≈ 1.4A hence Vout_max ≈ 1.4A*8Ω ≈ 11V.

https://forum.allaboutcircuits.com/threads/class-ab-amp-with-class-a-driver.138964/#post-1166680
http://www2.engr.arizona.edu/~brew/ece304spr07/Pdf/Class AB with VF.pdf

 

Haven't we already discussed this?

IL_max ≈ (0.5Vcc - Vbe3 )/( RE1/(β3 + 1) + RE3 + RL) ≈ (20V - 0.7V)/(390Ω/81 + 0.5Ω + 8Ω) ≈ 1.4A hence Vout_max ≈ 1.4A*8Ω ≈ 11V.

https://forum.allaboutcircuits.com/threads/class-ab-amp-with-class-a-driver.138964/#post-1166680
http://www2.engr.arizona.edu/~brew/ece304spr07/Pdf/Class AB with VF.pdf

Thanks Jony130

One last question on this topic. Can you tell me where you obtained the DC Beta for the TIP31C.

I can't see it from looking at any datasheet.

My Spice models are as follows:
.MODEL TIP31C NPN(IS=1.62181E-13 ISE=1.75416E-11 ISC=4.36516E-14 XTI=3 BF=83.6 BR=20.607 IKF=6.98433 IKR=0.997156 XTB=1.5301 +VAF=110.5 VAR=159.374 VJE=0.636 VJC=0.408 RE=0.56 RC=0.96 RB=164.793 RBM=0.100291 IRB=1.24287E-7 CJE=4.77E-10 CJC=7.29E-11 +XCJC=0.589205 FC=0.5 NF=0.9899 NR=0.989511 NE=1.95 NC=1.014 MJE=0.327 MJC=0.339 TF=2.3733E-8 TR=1.0000E-8 ITF=1 VTF=10 XTF=10 EG=1.1605 VCEO=100 ICRATING=3 MFG=TEXAS)

.MODEL TIP32C PNP(IS=6.77594E-13 ISE=1.31133E-11 ISC=1.31133E-11 XTI=3 BF=198.8 BR=25.4966 IKF=0.891251 IKR=0.410482 XTB=1.2648 +VAF=77.429 VAR=70.9603 VJE=0.59 VJC=0.5 RE=0.06 RC=0.16 RB=161.0 RBM=3.097 IRB=3.548134E-5 CJE=2.7E-10 CJC=1.07E-10 +XCJC=0.589205 +FC=0.5 NF=1.001 NR=1.004 NE=1.98 NC=1.12 MJE=0.319 MJC=0.352 TF=2.3733E-8 TR=1.0000E-8 ITF=1 VTF=10 XTF=10 +EG=1.0863 VCEO=100 ICRATING=3 MFG=TEXAS)

Is the 32C DC Beta really 198.8? The 31 and 32 are supposed to be complementary?

David

 

Can you tell me where you obtained the DC Beta for the TIP31C.

I just pick a random number around BF=83.6.

Is the 32C DC Beta really 198.8?

As you can see here:



At "low" current the beta is around 190. But it drops below 100 for Ic > 0.8A

The 31 and 32 are supposed to be complementary?

Complementary does not mean that the transistors are perfectly matched. The differences between PNP and NPN BJT's are unavoidable.


As a side note. You have chosen too weak transistors for this task. TIP31/31 are only 3A devices.

Next time try one of this BJT's
https://www.cordellaudio.com/book/spice_models.shtml

 

Your circuit has the driver transistors turning OFF the output transistors instead of turning them ON with LOTs of required current.
If you use Sziklai pairs instead then the first transistor provides plenty of current to completely turn on the output transistor.
A bootstrap capacitor can produce even more output swing and two diodes or a transistor can bias the outputs.
Negative feedback sets the voltage gain and reduces distortion.

 

Your circuit has the driver transistors turning OFF the output transistors instead of turning them ON with LOTs of required current.
If you use Sziklai pairs instead then the first transistor provides plenty of current to completely turn on the output transistor.
A bootstrap capacitor can produce even more output swing and two diodes or a transistor can bias the outputs.
Negative feedback sets the voltage gain and reduces distortion.

This circuit is out of a professor's textbook. Does he not know what he is doing?

 

This circuit is out of a professor's textbook. Does he not know what he is doing?

I know this will come as a shock to some people, but there are other options between complete right and complete wrong.

The circuit as shown will work, at least on paper, but it does not make efficient use of the available power source, is not optimized for low crossover distortion, is too dependent on the gain curves of the transistors (no negative feedback), etc. It is a unity gain buffer or current booster, not an audio power amplifier in the classic sense. National Semiconductor used to produce a power hybrid device with the same circuit topology inside, but theirs was in an essentially isothermal package, laser trimmed, etc. Look up the LH0002.

He knows what he is doing, and the circuit is useful to demonstrate certain operational characteristics of some devices. Just don't confuse a textbook example intended to teach a concept with a production ready design.

ak

 

OK. Are the driver's really turning off the outputs?

 

hi David.
AG is saying the voltage input swing at vm is driving Q1 out of conduction, so the Base current driving into Q3 is via RE1.
Its not an 'active' pull up thats driving Q3.

E

 

This circuit is out of a professor's textbook. Does he not know what he is doing?

This same horrible very old "teacher's circuit" is asked about by many students in online forums, but the circuit is never used by any half-decent amplifier. In Google the amplifier circuit is shown a few times but always from one country "over there".

One comment in Google about the horrible circuit says, "oh yes, yet another circuit copied from a magazine by another magazine, scanned by some guy in the 90s, then copied from a newsgroup to a website, where the last bit of explanation was lost, then stolen by another website, and then even that was lost. Seriously, none of this circuit makes much sense today".

 

This same horrible very old "teacher's circuit" is asked about by many students in online forums, but the circuit is never used by any half-decent amplifier. In Google the amplifier circuit is shown a few times but always from one country "over there".

One comment in Google about the horrible circuit says, "oh yes, yet another circuit copied from a magazine by another magazine, scanned by some guy in the 90s, then copied from a newsgroup to a website, where the last bit of explanation was lost, then stolen by another website, and then even that was lost. Seriously, none of this circuit makes much sense today".

It makes about as much sense as the half-assed answers I get on this web site

 

National Semi developed the circuit for a (probably military) customer, then offered it to the open market as the LH0002. Screaming-fast for its day, and pretty expensive. Later there was a power version in a TO-3 can, LH0033. I built an NTSC video distribution amplifier with one as the output stage. Flat flat, DG and DP at measurement limits.

Note that this is not an integrated circuit. LH stands for Linear Hybrid. Here is everything just about everything there is to know about the circuit:

http://www.ti.com/lit/an/snoa725a/snoa725a.pdf

ak

 

OK. Are the driver's really turning off the outputs?

You tell me. We don't give out free homework answers. Look at the circuit, *think* about how a transistor functions, tell us what you think, and we will advise.
Hint: Emitter follower.

ak

Curious about your handle. Are you a 55 year old electrical engineer, or an electrical engineer born in 1955?

 

It makes about as much sense as the half-assed answers I get on this web site

Then I suggest you go to another website that can give you full-assed answers.

 

Then I suggest you go to Electronics School like most of did. We didn't learn this stuff on a forum.

 

A long time ago designers "invented" the "best" topology for an efficient class AB amplifier. And they come up with a circuit that looks like this:

So-called voltage-feedback amplifier VFA


As you can see it is a direct coupled amplifier and it looks very similar to the op-amp based topology.
And this is the most popular topology of an audio amplifier even today (with some enhancements).
The 90% or more of an audio amplifers will use this topology.

And the second most popular topology look like this:

So-called current-feedback amplifier CFA





Hence if you are interested in the audio amplifier you should study these two topologies.

 

Note the input stage of the second schematic ...

ak

 

Note the input stage of the second schematic ...

ak

Yes, it is similar to the circuit we are talking about and is completely symmetrical. But I bet the transistors T1, T2, T3 and T4 are little and not power transistors and have fairly low currents.

 

Then I suggest you go to Electronics School like most of did. We didn't learn this stuff on a forum.

It is obvious that you didn't learn it at all.

 

e e 55: 186 messages and only 2 likes. Maybe you should pay more attention to the advice you're getting, and less time showing how little you know.

Without question, AG is a cranky old fart who loves the voice in his head. BUT, he has literally forgotten more about audio circuit design than most are even aware of. AND his rants, no matter how poorly presented, always are based in fact and experience. I don't always agree with him, but I disagree very carefully - not because of his attitude (he was banned from this forum for a while), but because of his training, experience, and insight.

Returning to post #6, yes, your instructor knows what he is doing. The problem evidenced by the way you chose to phrase your question is that you don't.

ak