Transistors

A transistor is a three-terminal semiconductor component in which a small signal controls a much larger current. As a switch, it lets a microcontroller pin run a motor, a relay or a strip of LEDs. As an amplifier, it turns a faint signal into a strong one. Billions of them make up the processor in every computer and phone. This article explains how the two main kinds work, how to wire them, and how to test and handle them safely.

1. A short recap

• Doping adds impurity atoms to silicon: phosphorus makes n-type, with spare electrons; boron makes p-type, with holes (missing electrons that act as positive charges).^[1] The plentiful carrier in a region is its majority carrier; the scarce one is the minority carrier. (Recap: doping; majority and minority carriers.)
• Where p-type meets n-type is a PN junction, discovered at Bell Labs in 1940,^[2] with a thin depletion region empty of free carriers. Forward bias narrows it and current flows; reverse bias widens it and blocks current.^[3] (Recap: the PN junction, in article 6, Diodes.)
• A resistor sets current by Ohm's law, V = I × R; two in series make a voltage divider. (Recap: resistors in circuits; the voltage divider.)
• A capacitor blocks DC but passes AC, so it can couple signals between stages, and one beside a chip supplies sudden bursts of current. (Recap: coupling; decoupling.)
• A coil makes a voltage spike when its current is switched off; a flyback diode across it gives that current a safe path. (Recap: inductive kickback; the flyback diode.)

2. What a transistor is

A transistor is a semiconductor device that can amplify or switch electrical signals.^[3] A small current or voltage at one of its three terminals controls a much larger current between the other two. The energy in that larger current comes from the power supply, not from the control signal; the transistor only regulates it.^[4]

Two families matter most. The bipolar junction transistor (BJT) is controlled by a small current into its middle terminal. "Bipolar" means both electrons and holes carry its current. The field-effect transistor (FET), most often a MOSFET, is controlled by a voltage on an insulated terminal that draws almost no steady current, and uses only one kind of carrier.^[5]

3. A short history

Before transistors, amplifiers and switches used vacuum tubes and electromechanical relays. After World War II, William Shockley organized a group at Bell Telephone Laboratories to find semiconductor replacements for them.^[6] On December 16, 1947, John Bardeen and Walter Brattain succeeded with the point-contact transistor: two gold contacts on a slab of germanium, where the voltage on one contact controlled the current through the other, amplifying a signal up to 100 times. Bell Labs announced the "transistor" on June 30, 1948.^[6]

On January 23, 1948, Shockley conceived the sturdier junction transistor, a sandwich of doped layers that works by injecting minority carriers across a junction, the principle behind all bipolar transistors today.^[7] The first working n-p-n junction transistors were announced in 1951.^[8] Shockley, Bardeen and Brattain shared the 1956 Nobel Prize in Physics "for their researches on semiconductors and their discovery of the transistor effect."^[9]

Early transistors were germanium; the Regency TR-1 pocket radio of 1954 used four.^[10] Germanium leaked current when "off" and worked only from about 0 to 70 °C, while silicon worked from −55 to 125 °C, and by the end of the 1950s silicon was the preferred material.^[11] In 1959 Mohamed Atalla and Dawon Kahng at Bell Labs built the first successful MOS (metal-oxide-semiconductor) field-effect transistor.^[12] In 1963 Frank Wanlass and Chih-Tang Sah showed that pairing p-channel and n-channel MOS transistors drew close to zero power while idle, the idea now called CMOS.^[13] Today over 99 percent of microchips use MOS transistors.^[12]

Gordon Moore predicted in 1965 that the components on a chip would double every year, revised in 1975 to every two years: Moore's law.^[14] Intel's first microprocessor, the 4004 of 1971, had 2,300 transistors.^[15] A 1995 Pentium held nearly 5 million.^[14] Apple's M3 chip of 2023 has 25 billion, and the M3 Max 92 billion.^[16]

4. How a bipolar junction transistor works

A BJT is three doped regions in a row, making two PN junctions back to back. An NPN transistor has a thin p-type layer between two n-type layers; a PNP has the reverse.^[3]​^[17] The regions are:

• The emitter (E), heavily doped, which "emits" carriers into the middle layer.^[18]
• The base (B), the middle layer, very thin and lightly doped.^[19]
• The collector (C), which collects the carriers. It is physically larger because it must shed more heat.^[19]

In normal use, the base-emitter junction of an NPN transistor is forward biased: the base sits about 0.6 to 0.7 V above the emitter, like a silicon diode.^[17]​^[19] This voltage is V_BE. Electrons, the majority carriers in the n-type emitter, flood into the base, where they are minority carriers. The base is so thin and has so few holes that almost none recombine. Instead they diffuse (spread by random motion) across to the collector-base junction.^[19]​^[7]

That junction is reverse biased, with the collector more positive than the base. A reverse-biased junction blocks majority carriers but sweeps minority carriers straight across, so the arriving electrons are pulled into the collector. Typically 95 to 98 percent of the emitter current reaches the collector and 2 to 5 percent leaves through the base, so I_E = I_B + I_C.^[19]

So a small base current controls a large collector current. The ratio is the current gain, β (beta), written h_FE on datasheets: I_C = β × I_B.^[3] With β = 100, 1 mA into the base allows 100 mA through the collector. Typical values run from 50 to 200, some up to 2,000.^[17] Gain also varies between parts and with current: the 2N3904 guarantees 100 to 300 at 10 mA but only 30 or more at 100 mA.^[20] Good designs never depend on an exact beta.

A PNP works the same way with every polarity reversed: holes are the carriers, current flows from emitter to collector, and you turn it on by pulling the base about 0.6 V below the emitter.^[17] Emitter and collector are not interchangeable, because they are doped differently.^[18]

5. How a MOSFET works

A MOSFET (metal-oxide-semiconductor field-effect transistor) controls current with an electric field. Carriers enter at the source (S) and leave at the drain (D), and the gate (G) controls the flow. The gate is a metal or polysilicon layer on a very thin film of silicon dioxide, an insulator, so it never touches the silicon.^[5]

In the common N-channel enhancement MOSFET, the source and drain are heavily doped n-type islands (N+) in a block of p-type silicon, the body. With no gate voltage, the path from source to drain runs n-p-n, two junctions back to back, so no current flows. Make the gate positive relative to the source and the gate, oxide and body act like a small capacitor: the gate's positive charge pulls electrons to the surface just beneath the oxide. Once enough gather, they form a thin layer that behaves like n-type silicon, an inversion channel, joining source to drain.^[5]

The gate-to-source voltage, V_GS, at which the channel just starts to form is the threshold voltage, V_GS(th). Higher gate voltage builds a stronger channel, and a fully on MOSFET behaves like a small resistor, R_DS(on). Two datasheet traps: the threshold is measured at a tiny current, 250 µA for the IRLZ44N, so a gate at threshold is barely on; and R_DS(on) is guaranteed only at the gate voltages listed.^[21] A logic-level MOSFET is rated at gate voltages a logic chip can supply, such as 5 V.^[21] R_DS(on) also rises as the part heats up.^[22]

The insulated gate draws no steady current, but it needs a brief surge to charge its capacitance at each switch.^[5] Datasheets give the charge needed as total gate charge, Q_g: 48 nC for the IRLZ44N.^[21]

A P-channel MOSFET is the mirror image and turns on when the gate is pulled below the source.^[23] A depletion-mode MOSFET has a built-in channel, conducts at zero gate voltage and is turned off by the gate; enhancement mode, off until turned on, is the most common.^[5]​^[19] In a discrete MOSFET (one transistor per package) the body is tied to the source, which leaves a PN junction from source to drain called the body diode.^[5]​^[22]

The JFET, briefly

The junction field-effect transistor (JFET) has a gate that forms a PN junction with the channel instead of being insulated. Reverse biasing that junction widens its depletion region and squeezes the channel until it is "pinched off."^[24]​^[19] A JFET conducts at zero gate voltage, and its gate draws almost no current, so it barely loads the signal feeding it.^[19]

6. Symbols, packages and pinouts

On a schematic, a BJT's emitter arrow shows the direction of conventional current: out of an NPN, into a PNP. SparkFun's memory aid: NPN means "not pointing in."^[17] A MOSFET symbol draws the gate as a separate line, because it is insulated; a broken channel line marks an enhancement part, and the body arrow points in for N-channel.^[5] The physical case is the package:

• TO-92: a small plastic body with one flat face and three leads, for low-power parts such as the 2N3904 and 2N7000.^[20]​^[25]
• TO-220: a larger body with a metal tab for bolting to a heat sink (finned metal that carries heat away). The tab is wired to a lead: the collector on a TIP120, the drain on an IRLZ44N.^[26]​^[21]
• SOT-23: a tiny surface-mount package soldered flat to a board. The MMBT3904 is a 2N3904 in this package.^[27]

The key rule: there is no standard pin order. The base is often not the middle lead, and the only sure guides are the datasheet and a meter test.^[18]​^[19] The 2N3904 and BC547 are both small NPN transistors in TO-92 cases, yet their pins run in opposite orders.^[20]​^[28] Packages differ too: the 2N3904 runs E-B-C, while the MMBT3904 numbers its pins base, emitter, collector.^[20]​^[27] Check the datasheet for the exact part and package in your hand.

7. Common parts and their key specs

Datasheets list absolute maximum ratings, limits never to be exceeded, then characteristics under stated test conditions. For a BJT, look at maximum voltage V_CEO, current I_C, power P_tot, gain h_FE, and V_CE(sat), the small voltage left across it when fully on.^[20] For a MOSFET, look at V_DS, I_D, R_DS(on), V_GS(th) and the maximum gate voltage, beyond which the oxide fails.^[21]

8. Operating regions

A BJT has three useful regions.^[17] In cutoff, V_BE is below about 0.6 V, no base current flows, and it is an open switch. In the active region, the base-emitter junction is forward biased, the collector-base junction reverse biased, and I_C = β × I_B; amplifiers work here. In saturation, both junctions are forward biased and it acts like a closed switch with only about 0.05 to 0.2 V across it; extra base current no longer raises the collector current, which the load now limits.

A MOSFET also has three regions, with clashing names.^[22] In cutoff, V_GS is below threshold. In the ohmic region (also called linear or triode), the drain-source voltage is small and the channel acts like a resistor; a MOSFET switch sits here when on. In the saturation region, the drain-source voltage is large and drain current levels off at a value set by V_GS; this is the amplifying region, the counterpart of the BJT's active region.

So "saturated" means fully on for a BJT but amplifying for a MOSFET. Infineon adds that the linear region differs from linear mode: "operating in linear mode means in the saturation region, not the ohmic region."^[22] Check which convention a document uses.

9. Transistor circuits

A BJT switching a relay (worked example)

An Arduino Uno's ATmega328P chip can source only a little current per pin:^[33] 40 mA absolute maximum, with its output voltage guaranteed only up to 20 mA.^[34] A small relay, a switch moved by an electromagnet, needs more. The 5 V Omron G5V-1 has a 167 Ω coil drawing 30 mA.^[35] A transistor solves this: the pin supplies a few milliamperes to the base, and the transistor carries the coil current. Sitting between load and ground, it is a low-side switch.^[17]

The steps (values calculated from the cited datasheets):

1. Check the transistor. The 2N3904's 40 V and 200 mA ratings comfortably exceed 5 V and 30 mA.^[20]
2. Choose the base current. A switch must saturate, so give the base far more than I_C ÷ β. The datasheet specifies saturation with base current one tenth of collector current (10 mA and 1 mA).^[20] Using that ratio, I_B ≈ 30 mA ÷ 10 = 3 mA.
3. Size the base resistor. Saturated, V_BE is 0.65 to 0.85 V; allow 0.8 V.^[20] Then R_B = (5 − 0.8) V ÷ 0.003 A ≈ 1,400 Ω. The next lower standard value, 1.2 kΩ, gives about 3.5 mA. Even if the pin reaches only its guaranteed 4.2 V, the base gets about 2.8 mA.^[34]
4. Check the heat. With about 0.2 V across the transistor, the coil draws about 29 mA.^[20]​^[35] The transistor dissipates 0.2 V × 29 mA plus 0.8 V × 3.5 mA, under 10 mW against its 625 mW rating.
5. Add the flyback diode. At switch-off the coil's spike would hit a collector rated for only 40 V. A diode across the coil, band toward +5 V, gives the current a safe loop until it dies away.^[37]​^[20] A 1N4148, rated 300 mA, easily handles 30 mA.^[36] (Recap: inductive kickback; the flyback diode.)

On a 3.3 V board, (3.3 − 0.8) ÷ 0.003 ≈ 833 Ω, so use 820 Ω (calculated). A PNP used as a high-side switch, between supply and load, turns on when its base is pulled low.^[17]

A logic-level MOSFET as a low-side switch (worked example)

For bigger loads a MOSFET is usually better: it needs no steady gate current and wastes little power when fully on. Suppose a 5 V pin must switch a 12 V LED strip drawing 3 A, using an IRLZ44N.

1. Check the ratings. The IRLZ44N handles 55 V, and with 5 V on the gate its R_DS(on) is at most 0.025 Ω.^[21]
2. Work out the heat (calculated). Power in a resistance is I² × R: 3² × 0.025 = 0.225 W. With a junction-to-air thermal resistance of 62 °C per watt, the chip runs about 14 °C above the air.^[21] No heat sink is needed.
3. Add a gate resistor. The gate draws a surge at each switch.^[5] A 330 Ω resistor limits it to about 5 V ÷ 330 Ω ≈ 15 mA, safe for the pin.^[34] Delivering the 48 nC gate charge then takes roughly 3 µs (calculated), negligible for on-off switching.^[21]
4. Add a pull-down. During reset the pins are tri-stated, neither high nor low.^[34] A floating gate can drift to any voltage, the same reason CMOS inputs must never float; a 10 kΩ resistor to ground is the usual fix.^[23] It holds the MOSFET off until the pin drives it.
5. Join the grounds. Gate voltage only matters relative to the source, so the 12 V supply's negative must connect to the microcontroller's ground.

Two cautions. The IRLZ44N's R_DS(on) is specified only at 4, 5 and 10 V on the gate, and its 1 to 2 V threshold is where it barely conducts, so a 3.3 V pin is not guaranteed to turn it fully on.^[21] For 3.3 V logic, pick a part specified lower, such as the AO3400A at 2.5 V, and redo the heat check.^[31] And a motor, solenoid or relay still needs a flyback diode: switching an inductive load without a clamp can drive the drain past its rating into destructive avalanche breakdown.^[22]

The common-emitter amplifier

In a common-emitter amplifier the signal enters at the base and leaves at the collector, with the emitter shared by both.^[4] Because a transistor conducts one way only, a bare AC signal would cut it off for half of each cycle. The cure is bias: a steady DC voltage on the base that keeps the transistor active, with the signal riding on top. For faithful reproduction (class A), the resting point sits roughly halfway between cutoff and saturation.^[4]​^[32]

The usual bias is a voltage divider, R1 and R2, plus a resistor RE in the emitter lead, one of the most widely used arrangements.^[19] (Recap: the voltage divider.) For figure 7 (all calculated):

• Base: 12 V × 10 kΩ ÷ 57 kΩ ≈ 2.1 V.
• Emitter: one V_BE lower, about 1.45 V, so 1.45 V ÷ 1.5 kΩ ≈ 0.97 mA flows. The collector current is nearly the same.
• Collector: 12 V minus 0.96 mA × 4.7 kΩ (about 4.5 V) leaves about 7.5 V, with about 6 V across the transistor: room to swing both ways.
• Divider: it carries about 0.21 mA, some twenty times the 10 µA base current. Base current does pull a divider down,^[32] here by about 0.08 V, so real readings land slightly lower.

RE makes the design stable. Simple single-resistor bias is thermally unstable: as the transistor warms, collector current rises and the operating point drifts.^[19] With RE, any rise in current raises the emitter voltage, which reduces V_BE and pulls the current back. A bypass capacitor CE across RE keeps that DC correction without cutting the AC gain.^[19] Coupling capacitors Cin and Cout pass the signal while blocking DC, so connected equipment cannot upset the bias.^[32]​^[19] (Recap: coupling capacitors.)

When the input rises, more collector current flows, more voltage drops across RC, and the collector voltage falls. The output is an enlarged, upside-down copy of the input, so this is an inverting amplifier, powered by the supply.^[4]

The Darlington pair

A Darlington pair feeds one BJT's emitter into another's base. Their gains multiply, up to about 10,000, but their base-emitter drops add, so the output loses 1.2 to 1.4 V.^[17] The TIP120 packages one in a TO-220 with a gain of at least 1,000. At 3 A its V_CE(sat) can reach 2 V, so it would dissipate about 6 W (calculated), three times its 2 W rating without a heat sink.^[26] The IRLZ44N above did the same job at about 0.23 W.

CMOS logic: how billions of switches compute

In digital circuits a transistor is simply an on-off switch, and integrated circuits pack vast numbers of them onto one piece of silicon to process ones and zeros.^[3] The basic building block is the CMOS inverter: a P-channel MOSFET from the supply to the output and an N-channel MOSFET from the output to ground, with gates joined as the input. A low input turns the P-channel on and the N-channel off, so the output is high; a high input gives a low output.^[23]

In either steady state one transistor is off, so almost no current flows from supply to ground.^[23]​^[13] Current does flow briefly at every switch, so power rises with switching speed;^[23] those bursts are what the decoupling capacitors beside every chip supply. (Recap: decoupling.) Four transistors in series and parallel make NAND and NOR gates.^[23] Gates combine into circuits that add, compare and remember, and billions of transistors make a processor like the M3.^[16] The Microcontrollers section shows what such chips do.

10. Testing transistors

A multimeter's diode-test mode, which shows a junction's forward voltage, checks most BJTs, which read as two diodes joined at the base.^[18] (Recap: testing a diode.)

• For an NPN, red (+) probe on the base and black on the emitter, then the collector: each should read a silicon junction, roughly 0.6 to 0.7 V (All About Circuits' example: 0.655 and 0.621 V). For a PNP, put black on the base.^[18]
• Reversed, each junction reads OL (open), and collector to emitter reads OL both ways. Exactly two of the six probe combinations should conduct; otherwise the part is faulty or not a BJT.^[18]
• The emitter-base reading is slightly higher than collector-base, because the emitter is more heavily doped, which helps identify unmarked parts.^[18]

Avoid resistance ranges that push more than 1 mA through the part.^[19] A meter's h_FE socket gives only an "approximate" gain, measured on one typical meter at just 10 µA of base current.^[38] Gain changes with current, so treat it as a rough check.^[20]

A MOSFET can be checked with a diode test that applies about 3 to 4 V; 1.5 V meters cannot turn the gate on.^[39] For an N-channel part, put black on the source and touch red to the gate, charging it. Move red to the drain: the reading should be low, because the MOSFET is on. A finger across gate and source discharges it, and the reading should go high.^[39] Red on the source and black on the drain should show the body diode.^[22] Failed MOSFETs often short drain to gate.^[39] Handle them with the static precautions below.

11. Everyday uses

• Computers and phones: processors and memory are built from MOS transistors, tens of billions per chip.^[12]​^[16]
• Amplifiers: a transistor can amplify a microphone's signal to drive a loudspeaker.^[3] Audio power stages often use a push-pull pair, one transistor per half-wave.^[32]
• Radios and hearing aids: the 1952 Sonotone hearing aid was America's first transistorized consumer product, and the TR-1 radio followed in 1954.^[40]​^[10]
• Power supplies and solar inverters: switching power supplies and motor-drive inverters use pairs of power MOSFETs,^[22] and a solar inverter uses transistor switches to turn DC into AC.^[41]
• Motor drivers: the DRV8833 chip contains two H-bridges of N-channel MOSFETs to run small motors in either direction.^[42]

12. Safety and handling

Static electricity. Electrostatic discharge (ESD) from the human body is one of the most frequent causes of transistor damage.^[19] MOSFETs are especially vulnerable: a discharge can arc through the thin gate oxide and burn a hole in it.^[22] CMOS chips share the weakness.^[23] Keep MOSFETs in anti-static bags or conductive foam, ground yourself with a wrist strap, handle parts by the package, and use a grounded soldering iron.^[22]​^[19]

Heat and thermal runaway. A transistor turns the voltage across it times the current through it into heat. A 2N3904 in free air handles 625 mW;^[20] a TIP120 handles 2 W alone or up to 65 W with ideal cooling.^[26] Check the dissipation, as in the worked examples, and add a heat sink when needed. Heat can feed on itself: it frees more carriers, allowing more current and more heat, a cycle called thermal runaway.^[19] A BJT's collector current rises with temperature, which is why RE matters in the amplifier above.^[19] A fully on MOSFET resists runaway because R_DS(on) rises with heat, but one run partly on can become thermally unstable.^[22] A TO-220 tab carries the collector or drain voltage, and so does any heat sink bolted to it.^[26]​^[21]

Switching mains loads. Never wire a hobby transistor circuit directly to household mains. Use a relay or module rated for mains, which keeps the high-voltage side separated from your circuit. The little G5V-1 is rated for only 0.5 A at 125 VAC.^[35] A power relay such as the Omron G5LE handles 10 A at 120 VAC, with insulation tested to 2,000 VAC between coil and contacts.^[43] OSHA notes that 50 to 150 mA through the body can stop breathing and may kill,^[44] only a few times a relay coil's 30 mA. Unplug before touching, and keep mains wiring enclosed.

Coming up in this section: 8 Transformers.

The whole story in brief

• A transistor lets a small signal control a large current from the supply, as a switch or an amplifier.
• Invented at Bell Labs in 1947, it moved from germanium to silicon and from bipolar to MOS; chips now hold tens of billions.
• In a BJT, about 0.6 to 0.7 V on the base-emitter junction lets a base current control a collector current about β times larger.
• A MOSFET's insulated gate builds a channel above V_GS(th); fully on, it is a small resistance, R_DS(on), at the specified gate voltage.
• Switches use cutoff and saturation, amplifiers the active region. Overdrive the base about tenfold, pull gates down, add flyback diodes and check the heat.
• Pin orders vary, so check the datasheet. Test BJTs as two diodes and MOSFETs with the gate-charge trick. Guard against static; keep mains behind a rated relay.

Sources

All links were checked in October 2026. External links open in a new tab.

1. Computer History Museum, The Silicon Engine, "1941: Semiconductor diode rectifiers serve in WW II": https://www.computerhistory.org/siliconengine/semiconductor-diode-rectifiers-serve-in-ww-ii/
2. Computer History Museum, The Silicon Engine, "1940: Discovery of the p-n Junction": https://www.computerhistory.org/siliconengine/discovery-of-the-p-n-junction/
3. OpenStax, University Physics Volume 3, 9.7 "Semiconductor Devices": https://openstax.org/books/university-physics-volume-3/pages/9-7-semiconductor-devices
4. All About Circuits, Lessons in Electric Circuits, Vol. III, ch. 4, "The Common-Emitter Amplifier": https://www.allaboutcircuits.com/textbook/semiconductors/chpt-4/common-emitter-amplifier/
5. All About Circuits, Lessons in Electric Circuits, Vol. III, ch. 2, "Insulated-Gate Field-Effect Transistors (MOSFET)": https://www.allaboutcircuits.com/textbook/semiconductors/chpt-2/insulated-gate-field-effect-transistors-mosfet/
6. Computer History Museum, The Silicon Engine, "1947: Invention of the Point-Contact Transistor": https://www.computerhistory.org/siliconengine/invention-of-the-point-contact-transistor/
7. Computer History Museum, The Silicon Engine, "1948: Conception of the Junction Transistor": https://www.computerhistory.org/siliconengine/conception-of-the-junction-transistor/
8. Computer History Museum, The Silicon Engine, "1951: First Grown-Junction Transistors Fabricated": https://www.computerhistory.org/siliconengine/first-grown-junction-transistors-fabricated/
9. NobelPrize.org, "The Nobel Prize in Physics 1956": https://www.nobelprize.org/prizes/physics/1956/summary/
10. Computer History Museum, Revolution, "Regency TR-1 transistor radio": https://www.computerhistory.org/revolution/digital-logic/12/273/1373
11. Computer History Museum, The Silicon Engine, "1954: Silicon Transistors Offer Superior Operating Characteristics": https://www.computerhistory.org/siliconengine/silicon-transistors-offer-superior-operating-characteristics/
12. Computer History Museum, The Silicon Engine, "1960: Metal Oxide Semiconductor (MOS) Transistor Demonstrated": https://www.computerhistory.org/siliconengine/metal-oxide-semiconductor-mos-transistor-demonstrated/
13. Computer History Museum, The Silicon Engine, "1963: Complementary MOS Circuit Configuration is Invented": https://www.computerhistory.org/siliconengine/complementary-mos-circuit-configuration-is-invented/
14. Computer History Museum, The Silicon Engine, "1965: “Moore's Law” Predicts the Future of Integrated Circuits": https://www.computerhistory.org/siliconengine/moores-law-predicts-the-future-of-integrated-circuits/
15. Computer History Museum, The Silicon Engine, "1971: Microprocessor Integrates CPU Function onto a Single Chip": https://www.computerhistory.org/siliconengine/microprocessor-integrates-cpu-function-onto-a-single-chip/
16. Apple Newsroom, "Apple unveils M3, M3 Pro, and M3 Max, the most advanced chips for a personal computer" (October 30, 2023): https://www.apple.com/newsroom/2023/10/apple-unveils-m3-m3-pro-and-m3-max-the-most-advanced-chips-for-a-personal-computer/
17. SparkFun Learn, "Transistors": https://learn.sparkfun.com/tutorials/transistors/all
18. All About Circuits, Lessons in Electric Circuits, Vol. III, ch. 4, "Meter Check of a Transistor (BJT)": https://www.allaboutcircuits.com/textbook/semiconductors/chpt-4/meter-check-transistor-bjt/
19. U.S. Navy, NEETS Module 7: Introduction to Solid-State Devices and Power Supplies (PDF): https://legacy.maritime.org/doc/neets/mod07.pdf
20. Diotec Semiconductor, 2N3904 NPN small signal transistor data sheet (PDF): https://diotec.com/request/datasheet/2n3904.pdf
21. Infineon (International Rectifier), IRLZ44NPbF HEXFET power MOSFET data sheet (PDF): https://www.infineon.com/dgdl/irlz44npbf.pdf?fileId=5546d462533600a40153567217c32725
22. P. B. Green, Infineon Technologies, "Designing with power MOSFETs" (application note, PDF): https://www.infineon.com/assets/row/public/documents/24/42/infineon-designing-with-power-mosfets-applicationnotes-en.pdf
23. All About Circuits, Lessons in Electric Circuits, Vol. IV, ch. 3, "CMOS Gate Circuitry": https://www.allaboutcircuits.com/textbook/digital/chpt-3/cmos-gate-circuitry/
24. All About Circuits, Lessons in Electric Circuits, Vol. III, ch. 6, "Introduction to Insulated-Gate Field-Effect Transistors": https://www.allaboutcircuits.com/textbook/semiconductors/chpt-6/insulated-gate-field-effect-transistors-igfet/
25. Diotec Semiconductor, 2N7000 N-channel enhancement MOSFET data sheet (PDF): https://diotec.com/request/datasheet/2n7000.pdf
26. Diotec Semiconductor, TIP120 NPN Darlington power transistor data sheet (PDF): https://diotec.com/request/datasheet/tip120.pdf
27. Nexperia, MMBT3904 NPN switching transistor data sheet (PDF): https://assets.nexperia.com/documents/data-sheet/MMBT3904.pdf
28. Diotec Semiconductor, BC546 to BC549 NPN small signal transistors data sheet (PDF): https://diotec.com/request/datasheet/bc546.pdf
29. Diotec Semiconductor, 2N3906 PNP small signal transistor data sheet (PDF): https://diotec.com/request/datasheet/2n3906.pdf
30. Diotec Semiconductor, 2N2222A NPN transistor data sheet (PDF): https://diotec.com/request/datasheet/2n2222a.pdf
31. Alpha & Omega Semiconductor, AO3400A 30V N-channel MOSFET data sheet (PDF): https://www.aosmd.com/res/datasheets/AO3400A.pdf
32. All About Circuits, Lessons in Electric Circuits, Vol. III, ch. 4, "Biasing Techniques (BJT)": https://www.allaboutcircuits.com/textbook/semiconductors/chpt-4/biasing-techniques-bjt/
33. Arduino, Arduino UNO R3 product reference manual (PDF): https://docs.arduino.cc/resources/datasheets/A000066-datasheet.pdf
34. Microchip Technology, ATmega48A/PA/88A/PA/168A/PA/328/P data sheet (PDF): https://ww1.microchip.com/downloads/aemDocuments/documents/MCU08/ProductDocuments/DataSheets/ATmega48A-PA-88A-PA-168A-PA-328-P-DS-DS40002061B.pdf
35. Omron, G5V-1 low signal relay data sheet (PDF): https://omronfs.omron.com/en_US/ecb/products/pdf/en-g5v_1.pdf
36. Vishay, 1N4148 Small Signal Fast Switching Diode data sheet (PDF): https://www.vishay.com/docs/81857/1n4148.pdf
37. SparkFun Learn, "Diodes": https://learn.sparkfun.com/tutorials/diodes/all
38. Velleman, DVM66 digital multimeter user manual (PDF): https://cdn.velleman.eu/downloads/1/dvm66gb.pdf
39. 4QD, "Testing MOSFETs": https://www.4qdtec.com/mostest.html
40. Computer History Museum, Revolution, "Inventing the Transistor": https://www.computerhistory.org/revolution/digital-logic/12/273
41. U.S. Department of Energy, "Solar Integration: Inverters and Grid Services Basics": https://www.energy.gov/cmei/systems/solar-integration-inverters-and-grid-services-basics
42. Texas Instruments, DRV8833 dual H-bridge motor driver data sheet (PDF): https://www.ti.com/lit/ds/symlink/drv8833.pdf
43. Omron, G5LE PCB power relay data sheet (PDF): https://omronfs.omron.com/en_US/ecb/products/pdf/en-g5le.pdf
44. U.S. Occupational Safety and Health Administration, Controlling Electrical Hazards (OSHA 3075, PDF): https://www.osha.gov/sites/default/files/publications/OSHA3075.pdf