Wilson Current Mirror

In the world of analog electronics, creating a stable, unwavering flow of current is a fundamental challenge, much like ensuring a perfectly constant water flow from a garden hose regardless of pressure changes. While simple two-transistor current mirrors offer a basic solution, they falter when precision is paramount, as their output current varies with voltage. This limitation creates a knowledge gap for engineers seeking to design high-performance circuits. This article introduces an elegant and powerful solution: the Wilson current mirror. We will explore how this clever three-transistor arrangement harnesses one of engineering's most powerful concepts—negative feedback—to achieve near-perfect performance. In the following sections, you will gain a deep understanding of its design, its remarkable advantages, and its critical role in modern electronics. The "Principles and Mechanisms" section will deconstruct the circuit to reveal how its feedback loop generates massive output impedance and uncanny accuracy. Following that, "Applications and Interdisciplinary Connections" will showcase how these properties are leveraged in real-world amplifier designs to achieve exceptional gain and signal purity.

Imagine you are trying to design the world’s most reliable garden hose. Your goal is to make the flow of water out of the nozzle perfectly constant, regardless of whether you point it straight up, straight down, or kink the hose slightly. The pressure at the nozzle will change, but you want the flow—the current of water—to remain stubbornly the same. In the world of electronics, this is precisely the job of a ​​current source​​, a fundamental building block that provides a steady, unwavering flow of electrical current.

A simple ​​current mirror​​, made of just two transistors, is like a basic hose. It does a decent job, but if the "pressure" (the voltage) at its output changes, the current wavers. This imperfection is quantified by a parameter called ​​output impedance​​, or output resistance. An ideal current source would have an infinite output impedance, meaning its current wouldn't change at all, no matter how the output voltage fluctuates. Our simple mirror's output impedance is finite, limited by a characteristic of the transistor known as its intrinsic output resistance, $r_o$. So, how can we do better? How can we build a nearly perfect current source?

One intuitive strategy to improve our hose would be to add a second valve in series with the first. The first valve sets the approximate flow, and the second, more sensitive one, absorbs most of the pressure fluctuations, shielding the first valve from them. In electronics, this is the essence of the ​​cascode​​ configuration. By stacking a second transistor on top of the first, we dramatically increase the output impedance. The cascode transistor acts as a buffer, shielding the current-defining transistor from voltage variations at the output. The result is a much more constant current, with an output impedance boosted by a factor of roughly $g_m r_o$, where $g_m$ is the transistor's transconductance—a measure of its amplification ability. This factor, known as the intrinsic gain of the transistor, can be 50 or 100, representing a significant improvement.

But nature, and the brilliant minds of engineers, have found an even more elegant solution: the ​​Wilson current mirror​​. At first glance, it's just a clever arrangement of three transistors. But hidden within this simple topology is a powerful principle that elevates its performance to an entirely different level: ​​negative feedback​​.

Negative feedback is one of the most fundamental and powerful concepts in all of engineering and science. It's the principle behind the thermostat in your house. If the room gets too hot, the thermostat senses this and turns off the furnace. If it gets too cold, it turns the furnace on. It constantly works to oppose any deviation from the set temperature. The Wilson mirror does the same thing with current.

Let's see how this magic works. The Wilson mirror uses three transistors, which we can call Q1, Q2, and Q3. The output current flows through Q3. The core of the feedback mechanism is the interaction between transistors Q2 and Q3.

Imagine the voltage at the output (the collector of Q3) tries to increase for some reason. This would normally cause the current flowing through Q3 to increase slightly (this is the effect quantified by $r_o$). In the Wilson configuration, this slight increase in current through Q3 must also flow through Q2. This increased current through Q2 causes the voltage at its collector (which is the emitter of Q3) to rise.

Now, here's the clever part. The base of Q3 is held at a relatively stable DC voltage by the diode-connected transistor, Q1. When the emitter voltage of Q3 rises, the voltage difference between its base and emitter ($V_{BE3}$) decreases. A smaller $V_{BE}$ on a transistor means it conducts less current. This reduction in current directly counteracts the initial tendency for the current to increase!

This is negative feedback in its purest form. Any change at the output is sensed, and the circuit automatically adjusts itself to fight that change. The loop is this: a change at the output affects Q3, which in turn affects Q2's operating point, and the change in Q2's condition is fed back to Q3's emitter, stabilizing the whole system. The transistors Q2 and Q3 form a self-correcting team, relentlessly working to keep the output current constant.

So, what is the practical result of this elegant feedback loop? A truly colossal output impedance. While a simple mirror's output resistance is just $r_o$, and a cascode's is about $g_m r_o^2$, the Wilson mirror's output resistance is approximately $\frac{\beta}{2} r_o$ for BJT transistors, or on the order of $g_m r_o^2$ for MOSFETs. Given that the current gain $\beta$ can be 100 or more, and the intrinsic gain $g_m r_o$ is also large, the Wilson mirror can achieve an output impedance that is dozens, or even hundreds, of times larger than that of a simple mirror.

A concrete comparison tells the whole story. For typical transistors in an integrated circuit, a simple mirror might have an output resistance of $800 \text{ k}\Omega$. A Widlar source, another common design, might improve this to around $2.6 \text{ M}\Omega$. The Wilson mirror, using the very same transistors, could achieve a staggering $40 \text{ M}\Omega$. It's like turning our leaky garden hose into a precision laboratory instrument.

The Wilson mirror's cleverness doesn't stop there. In BJT current mirrors, a small error arises because the base of the output transistor needs to draw a small current to operate, "stealing" it from the reference side. This means the output current is always slightly smaller than intended. For a simple mirror, the error is about $2/\beta$. If $\beta$ is 80, the output current is about 2.5% low.

The Wilson mirror's topology includes a third transistor (Q3) that ingeniously supplies most of the required base current, effectively canceling out this error. The analysis shows that the error in a Wilson mirror is reduced to approximately $2/\beta^2$. For our transistor with $\beta=80$, the error plummets from 2.5% to a minuscule 0.03%! This makes the Wilson mirror not only incredibly "stiff" (high impedance) but also incredibly accurate.

In engineering, as in life, there is no free lunch. The spectacular performance of the Wilson mirror comes at a cost, and that cost is ​​compliance voltage​​, or "headroom." This is the minimum voltage that must be maintained at the output for all the transistors to operate correctly (i.e., not in saturation). If the output voltage drops too low, the feedback mechanism breaks down, and the circuit stops behaving like a good current source.

Compared to a standard cascode mirror, the Wilson mirror requires a higher minimum output voltage. The difference is roughly the voltage of a forward-biased diode, $V_{BE(on)}$ (about $0.7 \text{ V}$), minus the small saturation voltage $V_{CE(sat)}$ (about $0.2 \text{ V}$). This difference, about half a volt, might not seem like much, but in modern low-voltage electronics powered by batteries, every fraction of a volt counts. This reduced "headroom" can be a significant drawback, limiting the usable voltage range of the output.

This trade-off between performance and headroom didn't stop engineers. It inspired them. If the three-transistor Wilson mirror has a headroom problem, can we fix it? The answer is a resounding yes, leading to the "wide-swing" or "improved" Wilson mirror. By adding a fourth transistor, arranged cleverly to adjust the internal bias points, designers created a circuit that retains the fantastic high output impedance and accuracy of the original Wilson mirror, while significantly lowering its minimum voltage requirement.

The story of the Wilson current mirror is a perfect illustration of the engineering journey: identify a need (a better current source), apply a fundamental principle (negative feedback) in a clever way to achieve a massive leap in performance, recognize the inherent trade-offs (compliance voltage), and then innovate again to overcome those limitations. It is a testament to how simple components, arranged with deep understanding, can give rise to extraordinarily powerful and elegant functions.

Now that we have analyzed the Wilson current mirror and understood the feedback mechanism behind its remarkable properties, we can explore its practical purpose. A key principle in science and engineering is to see a tool in action, understanding how its unique characteristics solve real problems and open up new possibilities. The Wilson current mirror is not merely an academic curiosity; it is a workhorse, a fundamental building block that enables the high-performance analog circuits at the heart of our modern world. Its applications are a beautiful illustration of how a deep understanding of principles can be used to overcome the inherent limitations of physical devices.

Let us embark on a journey into two of the most fundamental challenges in amplifier design—the quest for gain and the quest for purity—and see how our three-transistor friend provides an elegant solution to both.

Every amplifier has a fundamental purpose: to make a small signal larger. The measure of this ability is its voltage gain, which we can think of as the product of two factors: the amplifier's ability to convert an input voltage into a current (its transconductance, $G_m$) and the output resistance ($R_{out}$) that this current flows through to develop an output voltage. The relationship is elegantly simple: $A_v = -G_m \times R_{out}$. To achieve a truly massive gain, we need a massive output resistance.

But here we face a conundrum of the microscopic world. In an integrated circuit, or "chip," where components are unimaginably small, how does one create a large resistance? A classical resistor is just a strip of material, and its resistance is proportional to its length. To get a large resistance would require a long, thin strip, which would consume an enormous, unacceptable amount of precious silicon real estate. It would be like trying to fit a kilometer-long wire inside a watch.

The solution is one of the most beautiful ideas in electronics: the ​​active load​​. Instead of a passive, space-hungry resistor, we use other transistors to create a circuit that behaves like a very large resistance. A simple current mirror can serve as an active load, presenting a resistance equal to its own output resistance, $r_o$. This is a good start, but we can do so much better.

This is where the Wilson current mirror enters the stage. When used as an active load for an amplifier stage, it brings its superpower—its prodigiously high output impedance—to the party. While a simple mirror offers a resistance of $r_o$, the clever feedback within the Wilson mirror boosts this to a value closer to $g_m r_o^2$, where $g_m$ is the transconductance of the transistors. Since the factor $g_m r_o$ (the intrinsic gain of a single transistor) can easily be 50 or 100, the Wilson mirror provides an output resistance that is 50 to 100 times greater than a simple mirror!

This is not just a small improvement; it is a game-changer. By simply replacing a two-transistor simple mirror load with a three-transistor Wilson mirror load, an engineer can nearly double the overall gain of a typical differential-to-single-ended amplifier stage, without any other changes. It's a testament to the power of sophisticated design: with one extra, tiny transistor, we have dramatically enhanced the performance of the entire circuit.

An amplifier's job is not only to amplify the signal we care about but also to ignore the noise and interference we don't. In the real world, our electronic signals are constantly bombarded by unwanted noise—the 60-hertz hum from power lines, interference from nearby radio signals, and other electronic chatter. The most powerful weapon against this is the ​​differential amplifier​​.

The genius of a differential amplifier is that it amplifies the difference between two input signals while rejecting anything that is common to both. Since noise often gets added to both signal wires equally, it appears as a "common-mode" signal and is ignored. The measure of an amplifier's ability to do this is its Common-Mode Rejection Ratio, or CMRR. A high CMRR is the hallmark of a precision instrument.

The secret to achieving a high CMRR lies at the "tail" of the differential pair. The two transistors of the pair are supplied by a single current source, which is meant to provide a constant, unwavering total current. If a common-mode noise voltage tries to push the current up in both transistors simultaneously, a "stiff" or high-impedance tail source will resist this change, preventing the noise from being amplified. A low-impedance source, by contrast, would be "squishy," allowing the common-mode current to change and get passed through the amplifier.

Once again, the Wilson current mirror proves to be the ideal tool for the job. Used as a tail current source, its high output impedance provides exactly the stiffness needed to reject common-mode signals. If a simple BJT current mirror with an output resistance of $r_o$ is replaced by a Wilson mirror, the tail impedance skyrockets to a value on the order of $\beta r_o / 2$, where $\beta$ is the transistor's current gain. For a typical $\beta$ of 100, this is a fifty-fold improvement! This directly translates into a fifty-fold reduction in the undesirable common-mode gain, dramatically improving the amplifier's purity and noise immunity.

We have seen that a high-impedance Wilson mirror tail source is absolutely critical for rejecting common-mode noise. A natural question arises: does this magnificent impedance also help to boost the differential-mode gain, the amplification of the signal we actually want?

The answer, which might surprise you, is a resounding ​​no​​.

This is one of those beautiful moments in science where understanding what something doesn't do is as insightful as understanding what it does. The reason lies in the perfect symmetry of differential operation. When a pure differential signal is applied, one input goes up by a small voltage while the other goes down by the exact same amount. The currents in the two transistors change by equal and opposite amounts. One increases, the other decreases, and the total current flowing into the tail source remains perfectly constant.

Because the total current doesn't change, the voltage at the common point where the transistors meet their tail source doesn't move at all. It becomes a "virtual ground"—a point that is as stable as if it were physically wired to the ground reference. And if that point doesn't move, the impedance of whatever is connected to it becomes irrelevant. The tail source could have an impedance of a thousand ohms or a billion ohms; for a pure differential signal, it makes no difference to the gain.

This is the profound elegance of the design. The circuit intrinsically separates its response to the two types of signals. The Wilson mirror's high impedance is 'on call,' standing guard to fight off any common-mode disturbances, but it gracefully steps aside and does not interfere with the amplification of the desired differential signal.

The Wilson current mirror, then, is more than just a circuit. It is a physical manifestation of a powerful idea: using feedback to craft near-perfection from imperfect parts. Whether it is used as an active load to create immense gain or as a tail source to ensure signal purity, it demonstrates how a simple, elegant arrangement of a few transistors can solve some of the most fundamental problems in engineering, enabling the precision electronics that shape our technological landscape.