Also known as the source follower, the common-drain amplifier is notable for its low output impedance. This article introduces the basic common-drain configuration and examines its large-signal characteristics.
Ideally, the output resistance of an op amp should be zero. This is because the op amp output resistance is in series with the load resistance, and so the two resistances will form a voltage divider. For a non-zero output resistance, this will result in a reduced output voltage at the load.
Despite its widespread use, the common-source amplifier is far from ideal in this regard. Its output impedance is on the same order of magnitude as an MOS transistor’s output resistance, which can easily be several kΩ. Because of that, it’s useful to buffer the output of a common-source amplifier with another amplifier that has a comparatively low output impedance.
Enter the common-drain amplifier. While not used nearly as often as its common-source counterpart, when placed at the output of an op amp it makes a great voltage buffer: its small-signal gain is approximately unity, and its output resistance is relatively small. This amplifier’s input and output impedances are also useful when creating oscillators or two-terminal active inductors.
Figure 1 shows a basic common-drain amplifier with an ideal current source.
 |
|
| Figure 1. | The basic common-drain amplifier configuration. |
As you can see in the Figure 1, the common-drain amplifier uses the gate as its input and the source as its output. Because the output voltage at the source “follows” the input voltage at the gate, this configuration is also known as the source follower. The drain is tied to a DC voltage. For NMOS, the DC voltage is assumed to be VDD; for PMOS, it’s ground.
In this article, we’ll learn about this source follower’s large-signal characteristics. A subsequent article will cover the amplifier’s small-signal behavior.
Large-signal behavior with an ideal current source
To understand the large-signal characteristics of the circuit above, let’s see what happens to VOUT when we sweep VIN from 0 to VDD.
At the start of the sweep, both VIN and VOUT are equal to 0. As long as VIN is less than the threshold voltage of the NMOS transistor (VTH), the transistor will remain off and VOUT will continue to be 0. Once VIN reaches the threshold voltage, however, the transistor turns on.
When it turns on, the transistor is in saturation. An NMOS transistor reaches saturation when VDS = VGS – VTH. Since VDS = VDD in this configuration, and VDD will always be greater than VGS – VTH, the transistor will remain in saturation for the rest of the sweep.
The current source forces the bias current (IBIAS) through the transistor once it’s turned on. Based on that, we can calculate the gate-to-source voltage as follows:
 |
(1) |
where:
μn is the mobility of the NMOS transistor,
COX is the oxide capacitance,
W is the transistor width,
L is the transistor length.
Since VGS in this case is equal to VIN – VOUT, we can rearrange this equation to solve for VOUT:
 |
(2) |
We can see from Equation 2 that VOUT follows VIN at a lower voltage. This points to one of the common applications of the source follower – namely, that it can be used as a level shifter.
Figure 2 graphs the large-signal characteristic of the common-drain amplifier in Figure 1. The output voltage on the graph goes negative because an ideal current source is being used in the circuit.
 |
|
| Figure 2. | DC characteristic of the common-drain amplifier in Figure 1. |
The proper operating region for the common-drain amplifier’s input voltage is given by:
 |
(3) |
A more realistic implementation
Because it uses an ideal current source, Figure 1 isn’t representative of how we’d actually implement a common-drain amplifier. In real life, the current source would be implemented via a load transistor, as shown in Figure 3. The load transistor is labeled as M2.
M2 must be in saturation to work properly, and so the large-signal characteristic of this amplifier is slightly different. Saturation occurs when VOUT ≥ VBIAS – VTH. M2 operates in the linear region if VOUT is lower than VBIAS – VTH, degrading the performance of the source follower.
 |
|
| Figure 3. | Common-drain amplifier with MOS current source (M2). |
The DC characteristic for Figure 3 is shown in Figure 4.
 |
|
| Figure 4. | DC characteristic of common-drain amplifier with MOS current source. |
Because VBIAS must be greater than VTH, our operating range is more limited than it was with an ideal current source. Using the relationship of VOUT to VIN in Equation 2, we can define the range of VIN in which this version of the common-drain amplifier works properly:
 |
(4) |
Wrapping up
From our large-signal analysis, we can see that the source follower outputs a level-shifted voltage of the input. Because the output tracks changes in the input, the common-drain amplifier can be used as a buffer or level-shifter. We’ll learn about more uses for the source follower when we examine its small-signal behavior in the next article.
