⚡ Transistors

Motivation This note aims to cover my mental derivation of the knowledge required all the way up to Metal Oxide Semiconductor Field Effect Transistors (MOSFET)

💡Switches

We want switches that we can control digitally, as in something that can switch between ON and OFF states

Requirements for a good switch Basic Requirements ✅ At OFF state, blocks current nearly completely ✅ At ON state, having minimal current resistance ✅ Transition almost instantly between ON and OFF states

⚙️ Semiconductors

First, we need some kind of switch that can control between conducting and insulating through applying electric potential difference

We have found materials like this in their element and compound forms for what we call semiconductors

Etymology Semi-Conductors are named so as they are not fully conductive nor resistive, sitting somewhere in between where they conduct only under specific circumstances are met, like under certain temperature or electric potential difference

ℹ️

Current State of Semiconductors Common semiconductor materials include:

• Silicon (Si) – the most widely used, abundant, and well-understood.
• Germanium (Ge) – used in early transistors and some high-speed applications.
• Gallium Arsenide (GaAs) – used in high-frequency and optoelectronic devices.
• Silicon Carbide (SiC) and Gallium Nitride (GaN) – used in power electronics for high voltage and temperature tolerance.

These materials form the backbone of modern electronics, enabling everything from logic gates to solar cells.

🔧 Doping

This is done through by combining with other elements to either

• Increase their negative charge carriers (electrons) OR
• Increase their positive charge carriers (the “carriers” here are just lack of electrons—otherwise called electron holes or just holes)

This process is called doping

Doping in Detail The term doping comes from early chemistry and metallurgy, meaning to add a small amount of a foreign substance to alter properties.

N-Type Adds atoms with extra valence electrons (e.g. Phosphorus, Arsenic) to Silicon, creating free electrons

P-Type Adds atoms with fewer valence electrons (e.g. Boron, Gallium), creating electron holes.

ℹ️

Carrier Behavior Electrons are real particles that move through the lattice.

Holes are the absence of electrons in the valence band, and behave like positive carriers because electrons move to fill them.

Conceptually, both types of carriers contribute to current flow in electricity

ℹ️

Terminology Doping-Types

• Those with elevated negative charge carriers are called
  Negative-Type Semiconductors—or N-Type for short
• Those with elevated positive charge carriers are called
  Positive-Type Semiconductors—or P-Type for short

Purity-Types

• Raw forms are labelled as Intrinsic
• Doped forms are labelled as Extrinsic

🔗 P-N Junction and Depletion Regions

Recombination leaves behind fixed ionized atoms—positive in the N-type and negative in the P-type—creating an internal electric field that opposes further diffusion

This is thus called the Depletion Zone

Flow of Current The flow of electrons will only resume when an electric potential difference is applied across the junction in the direction of natural diffusion of majority carriers, by weakening the internal barrier

🚪 Diodes

Through this structure, we have invented a one-way gate for charge carriers, called a diode

Etymology The term comes from di- for “two” and -ode for “electrode”—the two terminals where electricity flows

ℹ️

One-Way Gate Behavior Flow in the other direction would be too difficult to happen, unless:

• At low voltages where a little bit leaks the other way——because of quantum tunneling
• Ultra high voltages where the system is volatile and chain explosion of charge carrier flow can happen.

🧱 Side-by-Side P-N Junctions (NPN / PNP)

When we combine both the N-Type and P-Type Semiconductors, where one type is on both sides of the other type, same-type charge carriers cant flow from one side to the other side, as the preference would be to flow to the middle opposite-type semiconductor.

Thus, not flowing is the default. But we can make this a switch, thus we aim to find ways to allow current to flow from one end to another when we want to let it flow, and cut the connection when we don’t want it to flow.

Switch Quality NPN / PNP ☑️ Blocks current nearly completely in OFF state

• Due to opposing preferred diffusion direction

☑️ Have minimal resistance to current in ON state

• Potentially when we can connect the same-type doped semiconductor on both ends

☑️ Transition almost instantly between ON and OFF states

• Potentially if switching between on and off states is primarily controlled by potential difference and not primarily by current

🔁 Transistor

By carefully biasing the junctions in an NPN or PNP structure, we create a transistor——a device that can control current flow using a small input signal, or rather a controlled semiconductor switch

• Bipolar Junction Transistor (BJT): current-controlled via base current
• Field Effect Transistor (FET): voltage-controlled via gate voltage

We’ll begin with BJTs, then move to FETs (particularly Inversion-based FETs commonly called MOSFETs)

🔌 Bipolar Junction Transistor (BJT)

Following from the structure of the side-by-side P-N Junctions, we have the Emitter where charge carriers flow from and Collector where charge carriers flow into, and the middle block is called the Base where the flow of charge carriers are controlled The Base is defined up front as it would be too tedious to explain this from scratch

    flowchart LR
    %% Actual bias directions
    X[External Circuit] --> E
    E[Emitter] --> B[Base]
    B --> C[Collector]
    C --> X

    %% Styling for doping levels
    classDef emitter fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
    classDef base fill:#f8c2d4,color:#000,stroke:#000,stroke-width:1px
    classDef collector fill:#f38ba8,color:#000,stroke:#000,stroke-width:1px

    class E emitter
    class B base
    class C collector
  

Doping Levels The 3 blocks are doped at different amounts:

• The emitter is doped highly because it is where the charge carrier emits from
• The base is thin and doped lightly, so that there is minimal recombination when current is required to flow across its opposite-type semiconductor
• The collector is doped moderately, for the best balance between electric potential difference to the base and current from the base

⛔ Cut-Off State (Fully OFF)

The opposite-facing P-N Junctions means that even when a potential difference is applied from the Emitter to Collector (across the Base) current would not be able flow from the Emitter to the Collector under reasonable circumstances

This is thus also the default state of this structure, Logical OFF

🛠️ Initialise with Separate Potential Difference Biases (Active State)

Naturally between the emitter and base, current will easily flow if potential difference is applied in the direction of emitter to base, as that is in line with the natural diffusion direction of majority carriers

But between base and collector, the preferred direction would be collector to base. Since we need the current to flow from base to collector, the potential difference must be applied in the opposite direction of the natural diffusion direction of majority carriers

    flowchart LR
    %% Actual bias directions
    E[Emitter] -->|Forward Bias| B[Base]
    B -->|Reverse Bias| C[Collector]

    %% Preferred natural directions (dotted)
    E -.->|Preferred Flow| B
    C -.->|Preferred Flow| B

 %% Forbidden direct bias
    E -.->|⚠️ Not allowed| C

    %% Styling for doping levels
    classDef emitter fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
    classDef base fill:#f8c2d4,color:#000,stroke:#000,stroke-width:1px
    classDef collector fill:#f38ba8,color:#000,stroke:#000,stroke-width:1px

    class E emitter
    class B base
    class C collector
  

Thus, during initialisation, it is easier apply separate electric potential difference bias—emitter to base, and base to collector, to drive current

If it were to have both at forward-bias, it would waste unnecessary amounts of energy trying to oppose its natural bias during initialisation

⚠️ Current Through Opposite-Type Semiconductor

This flow is called the base-flow which will flow out to the external circuit.

    flowchart LR
 X[External Circuit] --> |Current Flow| E

    E[Emitter] -->|Current Flow| B

    %% Base-flow diversion
    B[Base] -.->|Base-Flow diverted| X

    %% Styling for doping levels
    classDef emitter fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
    classDef base fill:#f8c2d4,color:#000,stroke:#000,stroke-width:1px

    class E emitter
    class B base
  

Moreover, since base-flow maintains a healthy potential difference between the emitter and the base—allowing charge carriers to flow smoothly—the transistor’s ability to operate depends on this base-flow being sustained

To take it even further, the base-flow is the switch to turn on the entire transistor

⚡ Transition to Saturation (Fully ON)

As base current increases, more majority carriers flood into the base while the collector keeps puling them out to the external circuit.

But eventually the resistance of the external circuit will bottleneck the rate at which the collector can pull them out, causing majority carriers to build up at the base to a point where the electric potential difference is actually now higher at the base than the collector

This flips the collector–base junction from reverse to forward bias.

    flowchart LR
 E[Emitter] -->|Forward Bias| B[Base]
 B -->|Now Forward Bias| C[Collector]
 
 %% Preferred natural directions (dotted)
  E -.->|Preferred Flow| B
  B -.->|Now Preferred Flow| C
  E -->|Electron Flow| C

 classDef emitter fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
 classDef base fill:#f8c2d4,color:#000,stroke:#000,stroke-width:1px
 classDef collector fill:#f38ba8,color:#000,stroke:#000,stroke-width:1px

 class E emitter
 class B base
 class C collector
  

At this point, both junctions are forward biased—the transistor enters saturation mode and is considered fully ON, and Logically ON

🏷️ Naming Logic

This construction is called a Bipolar Junction Transistor, or BJT for short.

• Bipolar refers to both flow of majority carriers and flow of minority carriers involved

  It is commonly thought of as flow of minority carriers into the opposite-type semiconductor instead of majority carriers flowing away, although they are the same thing).

• Junction is simply because of the P-N Junction used.

🍘 Inherent Flaw of Structure

We realised that for current to flow through the opposite-type semiconductor, a base flow of current must exist to hold enough potential difference for the current to continue flowing——other methods such as boosting potential difference would be less economical

Thus, a consistent flow of current is required to keep this transistor enabled, which consumes a non-negligible amount of extra power

If there were a way to have current flow through the opposite-type semiconductor without relying on another current, it would greatly reduce the consumption of extra energy

One of the ways of achieving this is to drive current through electric fields instead, as we will explore in our next section on Field-Effect Transistors (FETs)

Switch Quality Bipolar-Junction Transistors ✅ Blocks current nearly completely in OFF state

• Due to opposing preferred diffusion direction

✅ Has low resistance in ON state

• Because charge carriers move quickly from Emiiter, through the Base, to Collector

❌ Does not transition instantly between ON and OFF states

• Because it takes time for charge carriers to build up in the base when turning ON, and to clear out when turning OFF

🔀 Inversion-Mode Semiconductors——Preface

We will focus on a dominant subtype of FETs called Inversion-Mode FETs——most commonly recognized through their dominant subtype that became the namesake for this class——the Metal Oxide Semiconductor Field-Effect Transistor (MOSFET)

⚡ Metal Oxide Semiconductor Field-Effect Transistor (MOSFET)

Similarly, we follow from the construction of the side-by-side P-N Junction we shall name the same-type semiconductors at the start and end of the semiconductor block

The start of the block will be called the Source as it where mobile charge carriers are supplied from,

And the end place of the block as Drain as it is where mobile charge carriers drain into

    graph LR
	S[Source] -..-> M(Opposite-Type Semiconductor)
    M -..-> D[Drain]
	
	classDef high fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
	classDef low fill:#f8c2d4,color:#000,stroke:#000,stroke-width:1px
	
	class S high
	class M low
	class D high
  

🌉 Bridge of Carriers

To allow current to flow, we would need to somehow connect the 2 same-types from Source to Drain, and have a way to control it

It would be like a switch that lays a conducting bridge between the Source and Drain whenever we want it to flow through the opposite-type semiconductor, and then remove the bridge whenever we don’t want it to flow

We could make a physical bridge, but what if we want something faster, and more reliable?

Physical switches after all are more prone to breakage and the switching process would be slow

    graph LR
	S[Source] ---> M(Bridge on Opposite-Type Semiconductor)
    M ---> D[Drain]
	
	classDef high fill:#d6336c,color:#fff,stroke:#000,stroke-width:1px
	classDef medium fill:#f38ba8,color:#000,stroke:#000,stroke-width:1px

	
	class S high
	class M medium
	class D high
  

🧲 Dielectric Materials

Insulators alone could block charge carriers, but it blocks electric fields as well, we want the potential difference to be felt, so that the charge carriers will move and form a bridge, while they are blocked from flowing away from where we want to lay the bridge

Luckily there are such materials that exhibit this quality, known as dielectric materials that

• Allows electric fields to pass through itself
• Not allowing current to flow through it

Detailed Information on Dielectric Materials These are materials that can still be polarised (electric-fields can pass through it) but its charge carriers are bound too tightly to the atoms to be mobile, thus conducting no electricity

Dielectric is a shorthand for dia- meaning passing through, and electric, which means electric fields in this case——di- is a common shortened variation for dia- when the start of the suffix is also a vowel

🗼 Formation of Bridge

Once this is set up, simply set a terminal at the bottom of the block, and one above the top of the block, with the dielectric material right at the surface of the top

To create the bridge of charge carriers, a potential difference is applied between the bottom of the block, and above the top of the block.

The potential difference is applied primarily through a metal piece right on top of the dielectric material

The resulting electric field then attracts the charge carriers supposed to flow between the Source and Drain from within the opposite-type semiconductor towards the top, where they are stopped by the dielectric material, and accumulate to form the conductive bridge of charge carriers

🌊 Sustained Flow of Charge Carriers

This kickstarts the flow of current between the Source and Drain, that will now primarily maintain that bridge of charge carriers instead of from the opposite-type semiconductor

The potential difference between the top and the bottom must still remain applied to keep the bridge exactly at the semiconductor-dielectric interface

🎚️ Evaluation

THUS, we have arrived at the currently most popular type of transistor, let’s evaluate its switch quality

Switch Quality Metal Oxide Semiconductor Field-Effect Transistor ✅ Blocks current nearly completely in OFF state

• Due to opposing preferred diffusion direction

✅ Has low resistance in ON state

• Because the electric field creates a dense, conductive channel between same-type Source and Drain regions

✅ Transition almost instantly between ON and OFF states

• As the bridge of charge carriers is set up immediately by the fast-forming electric field when potential difference is applied at the gate,

🏷️ Naming Logic

Now that we’ve constructed the inversion-based FET, it’s worth noting that the most popular type before the 2010s featured a gate stack where:

• The dielectric material was silicon dioxide (oxide)
• The gate electrode—used to apply a potential difference—was aluminium (metal)

This layered structure gave rise to the namesake: Metal-Oxide-Semiconductor Field-Effect Transistor (MOSFET)

While modern MOSFETs no longer use those exact materials, the gate stack still consists of:

• A metal gate electrode
• An oxide-based dielectric

So the name remains apt—even as the materials evolve.

🎁 Extras

There are some variations of MOSFETs that reduce the leakage of charge carriers flowing between Source and Drain as the transistor becomes microscopically small

Fin Field-Effect Transistors (FINFET) for example, does this by introducing a vertical wall in between the Source and Drain called a Fin

🧩 Timeline of Field-Effect Transistor Evolution