# Differential BJT Class AB amplifier DC analysis

I have a few questions regarding analysis of the amplifier schematic below. Bear in mind that the resistor values have not been decided yet and are just copy-paste for now. The only requirements of the amp are that it drives an 8 ohm speaker with a supply rail of +16V and a total voltage gain of 1500 minimum (I know I am likely to have to add stages).

1. How is the best way to work out biasing the current mirror to provide the correct current? Previously we were given equal values of 1mA for each branch off the supply rail (so as not to cause current drive between stages I believe?), with 2mA coming from the current mirror. Should I use this arbitrary 1mA as a starting point to find resistor values?

2. In the amp's quiescent state, I understand that the voltage at the collector of Q8 should be 8V ideally, to allow for maximum output swing. How can I calculate the voltage at the base of Q8? I feel like I'm missing something obvious related to Q8 itself, and maybe assumptions on its Vbc?.

3. We have been told previously that as a rule of thumb the load resistance (R3 in this case??) should be a tenth of the circuit's output resistance. Is that true for this configuration? What about R2? That rule was given on a circuit without an R2.

I hope these questions aren't too simple, but we have never studied a circuit quite like this, and I am finding it hard to find a similar example online. Of course I don't want the answers outright, I'm just struggling with where to start. I am not confident enough that the addition of a couple of resistors or removal of negative supply doesn't throw me completely!

• What you're missing is the use of negative feedback to stabilise the DC working point.
– user16324
Dec 6, 2017 at 14:26

1:How is the best way to work out biasing the current mirror to provide the correct current?

Your 1mA-per-side guideline for Q3,Q4 is a good place to start, making the current mirror 2mA. Such low current won't overheat Q3,Q4.

2: How can I calculate the voltage at the base of Q8?

Voltage here must be quite high (close to the supply voltage). Otherwise, when Q8 collector voltage swings high, into saturation, output voltage going to your speaker will be limited by Q8. This means that voltage drop across R4 should also be small. Q8 provides a great deal of this amplifier's gain. R4 limits this gain, so should be quite small. So Q8's base and emitter voltages will hover quite close to the + supply rail.

3: Guidelines for R2 & R3.

Output of Q4 goes to drive input of Q8. So ask yourself, what equivalent impedance does Q8 present as a load to Q4? This will appear in parallel with Q4's own collector output impedance. (We're considering AC impedance here, not DC). So one-tenth of this parallel impedance might be in the low-kilohm ballpark for R3.
But you do have to consider DC biasing. DC voltage drop across R3 must be at least 0.7V, to bias Q8's base....likely a bit more, because there will be a DC voltage drop across R4. From guideline #1, a little less than 1mA will flow through R3 (the rest into Q8's base). So a minimum value for R3 will be 700 ohms. Because of DC voltage drop across R4, the value of R3 will be larger. Part of your design dilemma is justifying your one-tenth "rule" with the requirement of getting DC biasing right.

Your DC bias design should be checked this way: When Q8's collector voltage is sitting at mid-point (near 8V), look at Q3, Q4 collector currents. They should be nearly equal.

• Thanks for your help, I think from your answer I need to go back and look at the fundamentals of differential amplifiers. None of it is making sense to me.. Dec 18, 2017 at 18:24