Electrical – Differential Amplifier with Global Negative Feedback

It may be an old design, but it is still in use today, especially in car stereo amplifiers. What has changed is the quality and performance of the transistors. You asked a few questions, so I will just go stage-by-stage.

1) The differential input is much like that of op-amps. It allows for some gain and a balanced wide-band response and DC stability.

2) T401 is the signal input while T403 is the feedback from the output. This sets the gain of the amplifier to about 10 or less. It stabilizes the amplifier so with no signal in there is zero volts at the output. T401 feeds the signal to T407, which is a level shifter and adds some gain as well. T407 drive the output stage, through T405 which sets the bias(idle) current for the output transistors.

3) The more bias the less distortion, but the hotter T417 and T419 become. R418 (bias current adj)has to be trimmed with great care so the distortion is low but the outputs will not overheat. C410 creates a 'boot-strap' current fed back from the outputs that boost the drive current to the output stage, by just a small amount, increasing the maximum volume that can be put out to the speakers.

4) T13 and T417 form a NPN darlington and T415 and 419 at as a PNP darlington to act as a buffer for the signal at the collector of T407, which is already at the maximum voltage swing possible (at full volume). These darlington's allow the output to drive a low impedance load such as a speaker.

5) T409 and T411 protect the output stage from excessive currents by clamping the drive signal. This can cause distortion at high volume, but it is a warning sign that the volume needs to be turned down until the distortion stops.

6) R440 and C424 act as a small load at certain frequencies where the amplifier is unstable, and keep the amplifier from oscillating at those frequencies, regardless of the speaker load.

The "current source" which is connected between +50V and the collector of Q2 is not for protection altough it limits the output current sourcing capablity, too. Its main purpose is to generate high voltage gain. I am sure that term VA means just "voltage amplification".

Voltage at point VA jumps up if Q2 sinks less than the "current source" allows. If Q2 tries to sink more than the current source allows, the voltage at VA drops. The transition slope is very steep. Pair Q1Q2+R2 is voltage controlled emitter follower type current source. Q3 pulls the control voltage at node "from input stage" onto its knees if the voltage over R2 tries to exceed 0,7V.

The "current source" has no internal active pushing capablity, it's a current limiter which probably has exactly the same principle as how Q3 prevents the current of Q2 to be too high.

If you plan to add another "current source" in series with Q2 and to remove R2,R3 and Q3, you would gain nothing. There would still be the same parts and mechanism + an extra transistor in series with Q2. R2, R3 and Q3 would limit the current of that extra transistor.

What is the idea to have thislike protection in an internal stage of an amplifier. Nobody has a possiblity to short its output because there's the actual output stage (=high current capable voltage follower) between this circuit and the speaker connectors? Answer: Some clipping can occur. => The VA stage can get saturated. The saturation decays slowly due the diffusion capacitance in the transistors. Saturated amplifier causes severe distortion longer than a nonsaturated. This is the reason why cheap transistor amplifiers do not allways sound crisp and clear with transient rich music. here the deepness of the saturation is tried to keep limited.