Lab 5: Differential amplifier

The goal of this lab is to get familiar with differential signals and the differential amplifier.

Note that the lab assignment matches part of your homework assignment, so you may want to do the homework before the lab.

There is no pre-lab assignment, other than noting in your ELog that there is no pre-lab assignment.

Note that in this lab, and all of the ones using amplifiers, it is increasingly important to keep your circuits "clean". As discussed in the lab 1 instructions, it is best to trim the legs off components so that they sit flush with the surface of the breadboard. This will reduce the contributions from parasitic capacitance and inductance, which can make the amplifiers liable to have a high frequency oscillation noise. You should also connect the scope probes to separate wires rather than the legs of components, otherwise the increasingly complicated circuits can easily be bumped out of connection. An added benefit of cleanliness is that it makes it easier to see your circuit and identify any misconnections that could otherwise take a long time to debug.

You can also reduce the noise in your amplifiers with "by-pass" capacitors. These are capacitors placed between V_CC and ground and then also between V_EE and ground. Use a capacitor value of at least 1 μF and connect it as close to the transistors' (or later op-amps') power connection as possible. These capacitors will act as local energy storage so that any fast switching of current out of the power rails is supplied by the capacitors without having to generate current transients in the wires coming from the power supplies.

Generate a differential signal

Make a unity-gain phase splitter by taking one output from an emitter-follower and another from a common emitter amplifier. They then form a differential signal pair. (Hint: while you could do this with two different amplifiers, you could also make it with just one transistor and some AC decoupling on the two outputs.)

Use this to produce a pair of differential signals that are sine waves with about ±100 mV amplitude and about 100 kHz frequency.

They won't be perfect opposites of each other. Measure how close to opposites the two signals are. You only need to do this with about 10% precision, which you could get using the cursor functions. But, you could do it more precisely by saving the waveforms as CSV files and adding them. (In research, it is a good idea to save such "extra precision" information whenever you can. It will sometimes turn out to be useful to answer some detailed questions that come up later, so it is worth the ~30 seconds to save it in your log even if you don't need, or use, that extra precision at the time.)

Differential amplifier

Build a differential amplifier to receive these two signals. Choose the R_E and R_C values to have a gain of one for differential signals. Choose R_EE to suppress the common mode component by at least a factor of 10.

Note that you will need to AC couple the inputs to remove the DC bias from the previous stage. Then you need to set the DC bias of the two inputs somehow in order to have the 0.6 V base-to-emitter drop on two transistors of the differential amplifier. Although the AC coupling leaves them without a DC bias, there needs to be something to reference the signals' to ground. You could do that with a resistor divider between V_CC and V_EE, but getting both inputs at the same bias value could be imprecise. Another option is to just put a large resistor (~1M) between each base and ground. That will set the base quiescent point to V_B = 0 V for both transistors.

Choose the component values wisely! You need to make sure that the quiescent voltages are all far enough away from clipping or the circuit won't work. So, calculate the quiescent voltages you expect for your chosen component values, and then measure the quiescent voltages to make sure you see what you expect. You need to adjust the component values to tweak the quiescent voltages to be comfortably far away from clipping either at V_CC, V_EE, or with the V_CE>0.2 V rule. (Note that you do have the freedom to change the power supply set values for V_CC and V_EE.)

Confirm that the circuit behaves as expected for the differential component and common mode components. If you don't get exactly the gains (G_diff and G_CM) that you expect, think about potential sources of imbalance that might cause a break down in the derivation we used to obtain the differential gain formula.

Check the time dependence of point A in the circuit and see if it matches your expectation.

Now make the two signals that you input to the differential amplifier be the same rather than negatives of each other. Check the common-mode rejection behaves as expected.

Again, you may find less than optimal behavior; don't worry about deviations of less than 10%. (Ideally, we would use exactly matched transistors that are fabricated together on the same substrate. Here you are just using two un-matched transistors.)