Diodes

A diode is a component that lets current flow through it in one direction and blocks it in the other. That habit turns wall-outlet AC into the DC electronics need, protects circuits from backward batteries, tames relay-coil spikes and, in the LED, makes light. This article covers how diodes work, how to read them, the main types and circuits, and how to test and handle them safely. No background is assumed.

1. A short recap: doping, resistors, capacitors and coils

• Doping adds a few impurity atoms to pure silicon. Phosphorus gives n-type silicon, with spare free electrons; boron gives p-type, with holes (missing electrons that act as positive charges).^[1] (Recap: doping; majority and minority carriers.) This article picks up where article 2 stopped: the PN junction.
• The band gap, the energy needed to free an electron, will set an LED's color. (Recap: energy bands.)
• A resistor limits current by Ohm's law, V = I × R. (Recap: resistors in circuits, including the LED resistor; power ratings.)
• A capacitor stores charge and can smooth a bumpy voltage. (Recap: filtering and smoothing.)
• A coil (inductor) makes a voltage spike when its current is switched off, which a diode can absorb. (Recap: inductive kickback, in article 5, Coils (Inductors).)

2. What a diode is

A diode is a two-terminal component that conducts readily in one direction and very little in the other.^[2]​^[3] Conventional current (from + to −) passes from its anode to its cathode.^[4] When the anode is more positive than the cathode, the diode is forward biased and conducts. When the cathode is more positive, it is reverse biased and blocks.^[5] (Bias means a voltage applied to set how a component behaves.)

Most diodes are a piece of silicon with a p-type region and an n-type region. The boundary between them, the PN junction, does the work. Since its discovery in 1940 it has been the most common form of rectifier,^[6] a device that converts alternating current (AC) to direct current (DC).^[4]

3. How it works: the PN junction

On its own, each piece of doped silicon is electrically neutral: every free electron on the n side is balanced by the fixed positive ion its donor atom became, and likewise on the p side.^[5]

Now join the two. The n side is crowded with free electrons and the p side has almost none, so electrons spread across the junction into the p side, and holes spread the other way. Spreading from where something is concentrated to where it is scarce is called diffusion.^[7]​^[5] Near the junction the arriving electrons fill holes, and both disappear. This leaves the fixed ions behind with nothing to balance them: positive donor ions on the n side and negative acceptor ions on the p side. The ions are locked into the crystal and cannot move.^[5]​^[7]​^[3]

The strip around the junction now has almost no free carriers, so it is called the depletion region.^[5]​^[7] Its two layers of opposite charge create an electric field pointing from the positive ions to the negative ones. That field pushes back on any more electrons trying to cross from the n side, and diffusion stops almost instantly, at a balance point.^[5]​^[2] The voltage across the depletion region is called the built-in potential, or simply the barrier.^[7]​^[5] Figure 1 shows the arrangement.

Forward bias

Connect a battery with positive to the p side and negative to the n side. This opposes the junction's field: holes and electrons are pushed toward the junction, the depletion region narrows and the barrier drops. Electrons cross and combine with holes while the battery supplies more at each end, so current flows continuously.^[5]​^[8] Too much voltage means so much current that the heat destroys the junction.^[5]

Reverse bias

Reverse the battery. Now the applied voltage adds to the junction's field, pulling carriers away from the junction: the depletion region widens and the barrier rises. After a brief pulse, current stops except for a tiny leakage^[8]​^[5] carried by minority carriers, the few electrons in the p side and holes in the n side that heat and light keep creating.^[5]

4. The current-voltage curve

The clearest way to see what a diode does is a graph of the current through it against the voltage across it, called its current-voltage (I-V) curve or characteristic curve. For an ideal PN junction, OpenStax gives the current as I = I₀(e^eV/k_BT − 1). Here V is the voltage, T the temperature, e and k_B physical constants, and I₀ a tiny current, which is about all that flows in reverse.^[3]

Because of the exponential, each small step in forward voltage multiplies the current: the curve hugs zero, then turns sharply upward at the knee (Figure 2). For a silicon diode the current becomes useful at around 0.6 to 0.7 V. Germanium diodes conduct at about 0.3 V and Schottky diodes at about 0.15 to 0.45 V.^[4]​^[9] Past the knee, the voltage rises only slowly with current: a 1N4148 typically drops about 0.72 V,^[4] and no more than 1 V at 10 mA.^[10] Temperature matters too. Forward current is temperature-sensitive,^[3] and the 1N4148's maximum leakage rises 2,000-fold, from 25 nA at 25 °C to 50 µA at 150 °C.^[10]

In reverse, an ordinary diode passes only a leakage current, typically nanoamperes, until the voltage reaches its breakdown voltage, often 50 to 100 V or more.^[4] The 1N4148 is specified to break down at no less than 100 V.^[10] There, carriers in the strong field knock more electrons loose, which free still more: an avalanche.^[3]​^[11] Current soars, limited only by the rest of the circuit, but breakdown does no harm so long as the power rating is not exceeded.^[5]

The steep forward curve means a diode, especially an LED, must never be connected straight across a voltage source: a little extra voltage means a great deal of extra current. A series resistor sets the current instead^[12] (recap: the LED resistor).

5. Identifying a diode

The schematic symbol is a triangle pointing at a bar. The triangle points in the direction of conventional current, and the bar is the cathode.^[4] Special types change the bar or add arrows (Figure 3). On the part itself, the cathode is usually marked with a band, though some makers use a "k," a "+" or a shaped end.^[5] The 1N4148's band is black,^[10] and on the 1N4001 family "color band denotes cathode end."^[13] An LED's longer leg is the anode.^[12] Bridge rectifiers, which hold four diodes, have their pins marked on the body.^[14]

In U.S. part numbers, "1N" means a one-junction semiconductor, a diode; a suffix such as "A" marks an improved version.^[5]

6. The numbers on a datasheet

A diode's datasheet is the maker's specification document. These are the numbers that matter most:^[5]

• Forward voltage (V_F): the voltage dropped while conducting, always quoted at a stated current.
• Forward current (I_F or I_F(AV)): the continuous or average current. The surge current (I_FSM) is a one-off pulse it survives: 30 A for one 60 Hz half-cycle for the 1N4001.^[13]
• Peak reverse voltage: the most reverse voltage the diode can block without breaking down, one of the most important ratings. NEETS calls it PRV; datasheets list V_RRM, the repetitive peak reverse voltage; it is also called PIV, peak inverse voltage.^[5]​^[13]
• Reverse current (I_R): the leakage when blocking.
• Reverse recovery time (t_rr): how long the diode keeps conducting backward after the voltage reverses.^[5] A conducting PN diode fills with extra minority carriers and cannot turn off until they clear.^[15] The 1N4148 recovers in 4 ns or less.^[10]
• Power dissipation (P_tot) and maximum junction temperature (T_J): the heat limits. All ratings must be reduced at high temperature.^[5]

7. Kinds of diode

Schottky diodes replace the p-type side with a metal. Electrons diffuse from the semiconductor into the metal, leaving a depletion layer and a barrier, as in a PN junction.^[19] Only electrons carry the current, so no minority carriers build up and the diode turns off almost instantly.^[19]​^[15] The price is more leakage:^[15] a 1N5819 may leak 1 mA at its 40 V rating.^[16]

Zener diodes are designed to work in breakdown, where their voltage stays almost constant whatever the current.^[5] Below about 5 V, the heavily doped junction breaks down by tunneling: electrons pass straight through the very thin depletion region. Above about 5 V, breakdown is by avalanche.^[5] Makers set the voltage, from about 4 V to several hundred volts, by controlling the junction's thickness.^[11] Maximum current is the power rating divided by the voltage.^[17]

Light-emitting diodes (LEDs) are forward-biased junctions in which electrons and holes meet, recombine and give off their energy as light.^[22]​^[23] Each photon carries about the band-gap energy, so the band gap sets the wavelength, and wavelength is color.^[24] Red light at 700 nm takes 1.77 eV per photon; violet at 400 nm takes 3.1 eV.^[23] (An electron volt, eV, is the energy one electron gains crossing one volt.) Silicon and germanium junctions give off heat rather than light, so LEDs use compounds such as gallium arsenide phosphide and gallium nitride.^[23] Bigger gaps need bigger forward voltages: about 2.2 V for red, 3.3 V for blue.^[4] Efficient blue LEDs, which won the 2014 Nobel Prize in Physics, made white LED lamps possible: blue light makes a phosphor glow red and green, or red, green and blue LEDs are mixed.^[22]

For example (calculated), a red LED (about 2 V) run from 5 V at 10 mA needs a resistor dropping 3 V: R = 3 ÷ 0.010 = 300 Ω. The next standard value, 330 Ω, is SparkFun's suggestion for most LEDs.^[12]

Photodiodes work the other way round. Light falling on the junction frees carriers, and a reverse-biased photodiode conducts in proportion to the light.^[5] The BPW34 responds from 430 to 1,100 nm, visible through near infrared.^[20] Solar cells are photodiodes,^[4] and the first PN junction was noticed in 1940 because its current jumped under bright light.^[6]

Varactors use the fact that a reverse-biased junction is a capacitor, with the depletion region as its insulating gap. More reverse voltage widens the gap and lowers the capacitance, by up to 10 to 1, so a DC voltage can tune a radio circuit.^[5] (Recap: why a wider gap means less capacitance.)

Transient-voltage-suppression (TVS) diodes are breakdown diodes with very large surge ratings.^[4] A P6KE-series TVS absorbs 600 W pulses from inductive switching and lightning, and bidirectional versions work both ways.^[21]

Laser diodes differ from LEDs: they emit a narrow beam of a single wavelength, for disc players, bar-code readers and laser pointers, and are regulated as lasers.^[25] The same gallium nitride work behind blue LEDs produced the blue laser used in Blu-ray discs.^[22]

8. Diodes in circuits

Rectifiers

A half-wave rectifier is one diode in series with the load. It passes the positive half of each AC cycle and blocks the negative half, giving pulses at the supply frequency, 60 per second on 60 Hz mains: the ripple frequency.^[5]

A full-wave rectifier uses both halves. In one version, a transformer's secondary winding has a connection at its middle, a center tap; each half feeds its own diode, and the diodes take turns sending current the same way through the load. The ripple frequency doubles to 120 Hz, which is easier to smooth.^[5] (Transformers are article 8.) The bridge rectifier does the same job with four diodes and no center tap. With the same transformer it gives nearly twice the output voltage of the center-tapped circuit.^[5] Its current passes through two diodes each half-cycle, losing two forward drops, about 1.4 V. Plug-in power adapters usually contain one.^[4]

Each diode must survive the AC's peak voltage, not its everyday RMS (root mean square) rating, the DC value that would deliver the same power. NEETS works the conversion: 117 V RMS has a peak of 117 × 1.414 = 165 V.^[5] U.S. 120 V mains therefore peaks at about 170 V (calculated). A 1N4001 (50 V) would fail; a 1N4004 (400 V) or higher leaves margin.^[13]

Smoothing with a capacitor

Pulses are not yet useful DC. A large capacitor across the output charges to the peak while the diodes conduct, then alone supplies the load between peaks, discharging slowly.^[5] (Recap: smoothing capacitors.) The result is steady DC with a small sawtooth ripple (Figure 5). Ripple can be estimated from Q = CV.^[26] If the capacitor delivers a load current I for a time t, its voltage falls by It/C. With full-wave rectification, t is at most half a mains cycle, so the peak-to-peak ripple is about V_ripple ≈ I / (2fC), where f is the mains frequency. For 0.5 A from 2,200 µF on 60 Hz, that is 0.5 ÷ (2 × 60 × 0.0022) ≈ 1.9 V (calculated; slightly pessimistic, since recharging starts before the half-cycle ends). Half-wave rectification doubles the gap between pulses, and the ripple.

A Zener regulator, worked through

A Zener diode in reverse, fed through a series resistor, holds a nearly fixed voltage across a load connected in parallel with it.^[5] Suppose we want 5.1 V at 20 mA from a 12 V supply, using a 1N4733A (5.1 V).^[17] All the numbers below are calculated (Figure 6):

1. The resistor drops 12 − 5.1 = 6.9 V.
2. Allow about 10 mA for the Zener, to keep it firmly in breakdown, plus the 20 mA load: 30 mA. R = 6.9 ÷ 0.030 = 230 Ω. The nearest standard value is 220 Ω.
3. With 220 Ω, the current is 6.9 ÷ 220 = 31.4 mA: 20 mA to the load and 11.4 mA through the Zener.
4. The resistor dissipates 6.9 V × 31.4 mA ≈ 0.22 W, so use a 0.5 W part for margin (recap: power ratings).
5. If the load is unplugged, all 31.4 mA flows through the Zener: 5.1 × 0.0314 ≈ 0.16 W. That is far inside the 1N4733A's 178 mA maximum current and 1.3 W rating.^[17]
6. A load above 31.4 mA would leave the Zener nothing, ending regulation. Simple Zener regulators suit only small loads.^[4]

Two refinements. The 5.1 V is specified ±5% at a 49 mA test current, so at 11 mA the output will be a little lower. The Zener's dynamic resistance, typically 7 Ω, sets how much the output moves with current.^[17] If the supply rises to 13 V, the extra 4.5 mA lifts the output only about 0.03 V (calculated).

Reverse-polarity protection

A diode in series with the positive supply lead conducts normally but blocks if the battery or plug is connected backward.^[27]​^[4] The cost is the forward drop: Texas Instruments notes that a 0.6 V drop leaves a 1.8 V two-cell pack only 1.2 V. A Schottky loses less but costs more, and a transistor switch loses less still.^[27] Transistors are article 7.

The flyback diode

When the current through a coil, such as a relay or motor winding, is switched off, the coil produces a large voltage spike that can destroy the switching transistor (article 5 covers inductive kickback). A flyback diode, also called a freewheeling diode, is connected across the coil so that it is reverse biased while the coil is powered. At switch-off, the coil's current carries on around the coil-and-diode loop and dies away harmlessly.^[4]​^[28] TE Connectivity notes that a diode alone "fully eliminates the transient voltage." Because the slowly decaying current also slows the relay's release, TE suggests adding a Zener in series when fast release matters.^[28] If you drive relays or motors from boards like those in the Microcontrollers section, fit one.

Clamping and limiting

A limiter (or clipper) stops part of a waveform from going past a set value. It is used for wave shaping and circuit protection.^[29] A clamper does something subtler. Using a diode and a capacitor, it shifts the whole waveform up or down so that its top or bottom sits at a fixed DC level, keeping its shape and size. NEETS's example turns a square wave swinging ±25 V into one swinging from 0 to +50 V. Clampers, also called DC restorers, appear in test equipment, radar and sonar.^[29]

Diode logic and OR-ing

Two diodes with their cathodes joined make an OR gate: if either input is high, the output is high; if both are low, a resistor pulls it low.^[4] In power OR-ing, two supplies feed one load through Schottky diodes. If one fails, its diode is reverse biased and isolates it, and the other carries on.^[30]

9. Testing a diode

Most digital multimeters have a diode-test mode. It passes a small current through the part and displays the voltage across it. Fluke's procedure:^[9]

1. Turn off all power and make sure no voltage remains at the diode. Discharge any capacitors first (recap: capacitor safety).
2. Select diode test. Put the red lead on the anode and the black lead on the cathode (the banded end). A good silicon diode reads 0.5 to 0.8 V; some germanium diodes read 0.2 to 0.3 V.
3. Reverse the leads. A good diode reads OL (overload, meaning no measurable current).
4. OL in both directions means the diode is open. About the same low reading (around 0.4 V) both ways means it is shorted.

In a circuit, other parts can upset the reading, so you may need to disconnect one end.^[9] Without diode test, a resistance range works: forward should read several hundred times lower than reverse for a signal diode, at least 10 times for a power rectifier. This is not conclusive, since the meter's low voltage cannot reveal a diode that fails only at working voltage.^[5]

LEDs may need more voltage than the meter supplies. A Fluke 117's diode range stops at 2.000 V,^[31] so a good blue or white LED (around 3 V)^[4] reads OL, while a red one may read about 1.8 V and glow faintly. The dependable test is a battery and resistor in series with the LED.^[12] A bridge rectifier is four diodes, so test each: from the − pin to each AC pin, and from each AC pin to the + pin, expect a forward reading, and OL with the leads reversed (see Figure 4). On a Zener, diode test checks only the forward direction; the breakdown voltage is beyond the meter's range.

10. Everyday uses

• Power adapters and chargers. Nearly every device powered from a wall outlet has a rectifier, usually a bridge, at its input.^[4] Battery-powered equipment adds diodes or transistor switches against reverse battery installation and reverse current.^[27]
• LED lighting. LED lamps have reached just over 300 lumens (a measure of visible light) per watt, against 16 for incandescent bulbs and about 70 for fluorescent lamps. LEDs can last 100,000 hours, against about 1,000 for an incandescent bulb.^[22]
• Solar panels. A shaded cell in a series string can be forced into reverse bias and overheat. A bypass diode, usually one per group of cells, lets the current go around, limiting the reverse voltage to about one diode drop.^[32]
• Radio and radar detectors. In 1901 Jagadish Chandra Bose filed a U.S. patent for a galena crystal touched by a metal point, and in 1906 Greenleaf Pickard patented a silicon point-contact detector, sold as the "cat's-whisker" crystal detector. By the mid-1920s, vacuum tubes had replaced crystals in most radios.^[33] Crystals returned in World War II. Microwave radar receivers needed detectors working at frequencies vacuum tubes could not reach, and the point-contact diode's very low internal capacitance suited them.^[5]​^[1] Millions of crystal diodes, each a metal point touching a sliver of silicon or germanium, went into Allied radar receivers, and 99.999% pure silicon was available by war's end.^[1] Russell Ohl found the PN junction in 1940 while working on silicon radar detectors at Bell Labs.^[6] Today, Schottky diodes are the standard microwave mixer and detector diodes, because they store no charge.^[19]

11. Safety

Mains rectifier circuits

A rectifier on 120 V mains sees about 170 V peaks, and its smoothing capacitor can hold that charge after unplugging (recap: stored charge). OSHA notes that 50 to 150 mA flowing hand to foot for one second can stop breathing and may kill; at 1,000 to 4,300 mA the heart stops pumping.^[34] Beginners should stick to low-voltage circuits fed by a ready-made, enclosed adapter. Before touching any power-supply circuit, unplug it, discharge its capacitors and confirm with a meter that no voltage remains.^[9]

Hot diodes

A diode's heat is its forward drop times its current. A 1N4001 at its full 1 A drops up to 1.1 V, about 1.1 W; at a typical 50 °C per watt to the air, it runs about 55 °C above its surroundings (calculated), with its junction rated to 150 °C.^[13] Heat also frees more carriers, which raises the current, which makes more heat: thermal runaway, which ends in destruction.^[5] Schottky leakage climbs especially steeply with temperature, so Schottkys need careful thermal design.^[30] NEETS also advises never fitting or removing a diode with power on, and avoiding excess soldering heat.^[5]

LED and laser eye safety

Bright LEDs, especially blue and white ones, can harm the eye. Blue light between 400 and 500 nm can cause photochemical damage to the retina, the blue light hazard. The international standard IEC 62471 sorts light sources into risk groups from 0 (exempt) to 3 (high risk). A group 3 source can exceed the exposure limit in under a quarter of a second.^[35] Do not stare into high-power LEDs. Laser diodes are more dangerous still. The FDA's classes run from I, considered non-hazardous, to IV. Class IIIa (IEC 3R), typical of laser pointers, can be momentarily hazardous viewed directly; IIIb and IV are immediate eye hazards, and Class IV is hazardous even by reflection and can start fires.^[25] Never look into a laser beam or point one at anyone.

Coming up in this section: 7 Transistors (two PN junctions working together) and 8 Transformers.

The whole story in brief

• A diode conducts from anode to cathode and blocks the other way, thanks to the PN junction's depletion region and barrier.
• Forward current rises steeply past a knee: about 0.6 to 0.7 V for silicon, 0.3 V for germanium and Schottky. Reverse leakage is tiny until breakdown.
• The band marks the cathode; an LED's long leg is the anode. Key ratings: V_F, I_F, V_RRM (PIV), t_rr, power.
• Rectifiers plus a capacitor make DC. Zeners regulate, Schottkys save voltage, TVS diodes absorb surges, flyback diodes tame coils, and LEDs glow in a color set by the band gap.
• Diode test: 0.5 to 0.8 V forward, OL reverse. Respect mains peaks, heat and bright light.

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. HyperPhysics (Georgia State University), "The P-N Junction": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/pnjun.html
3. OpenStax, University Physics Volume 3, 9.7 "Semiconductor Devices": https://openstax.org/books/university-physics-volume-3/pages/9-7-semiconductor-devices
4. SparkFun Learn, "Diodes": https://learn.sparkfun.com/tutorials/diodes/all
5. U.S. Navy, NEETS Module 7: Introduction to Solid-State Devices and Power Supplies (PDF): https://legacy.maritime.org/doc/neets/mod07.pdf
6. Computer History Museum, The Silicon Engine, "1940: Discovery of the p-n Junction": https://www.computerhistory.org/siliconengine/discovery-of-the-p-n-junction/
7. C. Honsberg and S. Bowden, PVEducation, "Formation of a PN-Junction": https://www.pveducation.org/pvcdrom/pn-junctions/formation-of-a-pn-junction
8. HyperPhysics (Georgia State University), "Diodes": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/diod.html
9. Fluke, "How to Test Diodes with a Digital Multimeter": https://www.fluke.com/en-us/learn/blog/digital-multimeters/how-to-test-diodes
10. Vishay, 1N4148 Small Signal Fast Switching Diode data sheet (PDF): https://www.vishay.com/docs/81857/1n4148.pdf
11. HyperPhysics (Georgia State University), "Zener effect and Zener diodes": http://hyperphysics.phy-astr.gsu.edu/hbase/Solids/zener.html
12. SparkFun Learn, "Light-Emitting Diodes (LEDs)": https://learn.sparkfun.com/tutorials/light-emitting-diodes-leds/all
13. Vishay, 1N4001 to 1N4007 General Purpose Plastic Rectifier data sheet (PDF): https://www.vishay.com/docs/88503/1n4001.pdf
14. Vishay, GBU4A to GBU4M Glass Passivated Single-Phase Bridge Rectifier data sheet (PDF): https://www.vishay.com/docs/88614/gbu4a.pdf
15. Toshiba Electronic Devices & Storage, "Do Schottky barrier diodes (SBDs) have reverse recovery time (trr)?": https://toshiba.semicon-storage.com/us/semiconductor/knowledge/faq/diode/do-sbds-have-reverse-recovery-characteristictrr.html
16. Vishay, 1N5817, 1N5818, 1N5819 Schottky Barrier Plastic Rectifier data sheet (PDF): https://www.vishay.com/docs/88525/1n5817.pdf
17. Vishay, 1N4728A to 1N4761A Zener Diodes data sheet (PDF): https://www.vishay.com/docs/85816/1n4728a.pdf
18. Vishay, BZX55 Series Small Signal Zener Diodes data sheet (PDF): https://www.vishay.com/docs/85604/bzx55.pdf
19. Skyworks Solutions, "Mixer and Detector Diodes" (application note 200826A, PDF): https://www.skyworksinc.com/-/media/SkyWorks/Documents/Products/1-100/200826A.pdf
20. Vishay, BPW34 Silicon PIN Photodiode data sheet (PDF): https://www.vishay.com/docs/81521/bpw34.pdf
21. Vishay, P6KE6.8A thru P6KE540A TRANSZORB Transient Voltage Suppressors data sheet (PDF): https://www.vishay.com/docs/88369/p6ke.pdf
22. NobelPrize.org, "The Nobel Prize in Physics 2014: Popular information" (blue LEDs): https://www.nobelprize.org/prizes/physics/2014/popular-information/
23. HyperPhysics (Georgia State University), "Light Emitting Diodes": http://hyperphysics.phy-astr.gsu.edu/hbase/Electronic/led.html
24. P. Hadley, Graz University of Technology, Physics of Semiconductor Devices, "Light Emitting Diode": https://lampz.tugraz.at/~hadley/psd/L15/led.php
25. U.S. Food and Drug Administration, "Laser Products and Instruments": https://www.fda.gov/radiation-emitting-products/home-business-and-entertainment-products/laser-products-and-instruments
26. OpenStax, University Physics Volume 2, 8.1 "Capacitors and Capacitance": https://openstax.org/books/university-physics-volume-2/pages/8-1-capacitors-and-capacitance
27. J. Falin, Texas Instruments, "Reverse Current/Battery Protection Circuits" (SLVA139, PDF): https://www.ti.com/lit/an/slva139/slva139.pdf
28. TE Connectivity, "Relay Coil Suppression with DC Relays" (application note): https://www.te.com/en/products/relays-and-contactors/electromechanical-relays/intersection/relay-coil-suppression-dc-relays.html
29. U.S. Navy, NEETS Module 9: Introduction to Wave-Generation and Wave-Shaping Circuits (PDF): https://legacy.maritime.org/doc/neets/mod09.pdf
30. Texas Instruments, "Basics of Ideal Diodes" (SLVAE57B, PDF): https://www.ti.com/lit/an/slvae57b/slvae57b.pdf
31. Fluke, Fluke 117 Electrician's Multimeter data sheet (PDF): https://media.fluke.com/2dfc6457-2f44-4764-9c44-b10600670b0b_original%20file.pdf
32. C. Honsberg and S. Bowden, PVEducation, "Bypass Diodes": https://www.pveducation.org/pvcdrom/modules-and-arrays/bypass-diodes
33. Computer History Museum, The Silicon Engine, "1901: Semiconductor Rectifiers Patented as ‘Cat's Whisker’ Detectors": https://www.computerhistory.org/siliconengine/semiconductor-rectifiers-patented-as-cats-whisker-detectors/
34. U.S. Occupational Safety and Health Administration, Controlling Electrical Hazards (OSHA 3075, PDF): https://www.osha.gov/sites/default/files/publications/OSHA3075.pdf
35. D. Konjhodzic, "Evaluation of Blue Light Hazard", Proceedings of the 29th CIE Session, 2019 (PDF): https://files.cie.co.at/x046_2019/x046-OP30.pdf