solid state amp question

I've seen this circuit in a lot of solid state amps and I don't understand it.

The input to the power amp is a long-tail differential amp. The left input is tied to the signal source. The right input is tied to the output of the power amp.

What's going on here? Attached is a schem for an amp I'm working on with that setup.

It's the same thing I explained to you in the other thread. Most power amplifiers are just bigger opamps. Higher voltages, more output, hence a more elaborate circuit.

Q901/903 form the differential input pair. Left one positive input, right one negative input. That's why it is connected to the output through  R923, because the output gets fed back to the negative input. The ratio of R923/R917 determines how much signal gets fed back, and defines the overall gain of the amplifier. It's basically just a big non-inverting amplifier.

Q911 is the second gain stage, in the three-transistor-circuit you posted this would be the third one. Q913 and the connected resistors/trimmer form a bias spreader, that sets the quiescent current in the output transistors Q927/Q929. Q923/925 buffer the output transistors, Q907 buffers the second stage, both Q905/909 form current sources to improve linearity of this stage. The other four transistors limit the current in the output transistors so they don't get destroyed in the case of a fault or short on the output.

> I've seen this circuit in a lot of

Because it is THE basic opamp.

The Tandberg is fancified. 15 transistors!

Look at a 3-transistor opamp:
https://www.passdiy.com/project/amplifiers/diy-op-amps  --- figure 10, halfway down

The diff-pair compares the input to a sample of the output. The result is usually fed to a second stage for level-shifting and more gain. The Fig 10 plan stops there; it is conceptually complete but not the best performer. You can complicate the 2nd stage for better performance. A little opamp ('741, '5532, 990) will have a 3rd stage as an output buffer. (Fig 13 shows a simple buffer added.) The beefy Tandberg wants to drive a speaker with high current so the "buffer" is itself a 2-stage (Darlington) affair.

They all work the same. You can build the Fig 10 amp in a few moments out of odd parts and poke it to see what it does.

samgraysound said:

I've seen this circuit in a lot of solid state amps and I don't understand it....What's going on here?

Click to expand...

May I suggest some literature for further reading?

Duncan - High Performance Audio Power Amplifiers
Slone - High Power Audio Amplifier Construction Manual
Cordell - Designing Audio Power Amplifiers

samgraysound said:

I've seen this circuit in a lot of solid state amps and I don't understand it.

The input to the power amp is a long-tail differential amp. The left input is tied to the signal source. The right input is tied to the output of the power amp.

What's going on here? Attached is a schem for an amp I'm working on with that setup.

Click to expand...

I even talk about the basic opamp topology/behavior in my very old article http://www.johnhroberts.com/des_art_1.pdf

The input long tail pair, converts + and - inputs into a single net up/down, the middle stage references this to one rail and provides voltage gain to increase the swing, the last stage buffers this with current gain. This can be as simple as 3 transistors (without buffer), or hundreds of devices/components. 

JR

Thanks everyone. Looks like I got my reading cut out for me.

Sam