Collection of notes for various classes I've taken.
The diode has an anode (positive terminal) and a cathode (negative terminal).
The ideal diode may be considered the most fundamental nonlinear circuit element.
The positive terminal of the diode is called the anode and the negative terminal the cathode.
The ideal diode’s i–v curve is strongly nonlinear, but can be represented by simple piecewise segments. In the idealized picture there are two regions:
Each segment is linear by itself, so when the terminal voltage stays within a single segment the diode can be treated as a linear element for the duration of that operating condition. However, if the signal crosses a segment boundary (a “break point”) the operating region changes and a single linear model no longer applies.
For real diodes the break is not perfectly sharp. Practical points to remember:
A silicon diode typically begins to conduct noticeably near $~0.6–0.8 V$ (the “threshold” or “knee”), and a Schottky diode has a lower threshold.
Around a chosen DC operating point (bias), small AC perturbations can be analysed with a linear small‑signal model. The diode is represented by its dynamic (differential) resistance $r_d = V_T / I_D$ where $V_T ≈ 25–26 mV$ at room temperature and $I_D$ is the DC bias current.
The small‑signal linear approximation is valid only while the AC excursion remains small compared to the distances to the nearest breakpoints (e.g., does not drive the diode from blocking to strongly forward conduction).
The most common implementation of the diode involves a pn junction. Its characteristic curve consists of three regions:
In the forward-bias region, the i-v relationship is closely approximated by
\[i = I_S(e^{v/V_T}-1)\]where the saturation current $I_S$ is a constant for a given diode at a given temperature; we recall that $I_S$ is directly proportional to the cross-sectional area of the diode.
We note that $I_S$ is a strong function of temperature; further, that $I_S$ approximately doubles for each $5\degree C$ rise in temperature.
The thermal voltage $V_T = \frac{kT}{q}$, where $k$ is Boltzmann’s constant ($8.62\times 10^{-5} J/K$), $T$ is the absolute temperature in Kelvins ($273 + \degree C$), and $q$ is the magnitude of electronic charge ($1.60\times 10^{-9} \text{ coulombs}$).
In rapid approximate circuit analysis, we shall assume $V_T \approx 25 mV$ at room temperature.
When $i \gg I_S$, we approximate
\[i \approx I_Se^{v/V_T}\]or alternatively
\[v = V_T \ln \frac{i}{I_S}\]And we observe:
\[\frac{I_2}{I_1} = e^{(V_2-V_1)/V_T}\]or
\[V_2 - V_1 = V_T \ln \frac{I_2}{I_1}\]In the reverse-bias region, the i-v relationship is closely approximated by
\[i \approx -I_S\]and while $I_S$ is observed to roughly double for every $5\degree C$ rise in temperature, $-I_S$ is observed to approximately double for every $10\degree C$ rise in temperature.
The diode enters the breakdown region when $v \lt -V_{BR}$. $-I_S$ increases rapidly, although the increase in voltage drop is very small.
There are a few models for the diode, among which are the:
The exponential model is the most accurate representation of the diode’s behavior and is based on the fundamental physics of the pn junction.
The complete exponential model is given by:
\[I_D = I_Se^{V_D/V_T}\]and
\[I_D = \frac{V_{DD}-V_D}{R}\]Iterative analysis is used to solve the nonlinear equation that results from combining the exponential diode equation with the circuit equation.
Given a circuit with:
The equations are:
Start with an initial guess for $V_D$ and iterate:
Initial guess: $V_{D0} = 0.7V$ (typical forward voltage)
Calculate current: $I_{D1} = I_S e^{V_{D0}/V_T}$
Find new voltage: $V_{D1} = V_{DD} - I_{D1} \cdot R$
Repeat: Use $V_{D1}$ as the new guess and continue until convergence
Define the function: $f(V_D) = I_S e^{V_D/V_T} - \frac{V_{DD} - V_D}{R} = 0$
Calculate derivative: $f’(V_D) = \frac{I_S}{V_T} e^{V_D/V_T} + \frac{1}{R}$
Iterative formula: $V_{D(n+1)} = V_{D(n)} - \frac{f(V_{D(n)})}{f’(V_{D(n)})}$
Given:
Iteration using Method A:
Iteration | $V_D$ (V) | $I_D$ (mA) | $V_{D,new}$ (V) | Error |
---|---|---|---|---|
0 | 0.700 | 4.32 | 0.568 | - |
1 | 0.568 | 0.31 | 0.691 | 0.123 |
2 | 0.691 | 3.45 | 0.576 | 0.115 |
3 | 0.576 | 0.39 | 0.688 | 0.112 |
4 | 0.688 | 3.17 | 0.579 | 0.109 |
… | … | … | … | … |
Stop iterations when:
Voltage criterion: $ | V_{D(n+1)} - V_{D(n)} | < \epsilon_V$ (e.g., $\epsilon_V = 1mV$) |
Current criterion: $ | I_{D(n+1)} - I_{D(n)} | < \epsilon_I$ (e.g., $\epsilon_I = 0.1mA$) |
START
↓
Set initial guess V_D0
↓
Calculate I_D = I_S * exp(V_D/V_T)
↓
Calculate V_D_new = V_DD - I_D * R
↓
Check: |V_D_new - V_D| < tolerance?
↓ NO ↓ YES
V_D = V_D_new SOLUTION FOUND
↓ ↓
Go back to END
calculate I_D
Load line from $i=\frac{V_{DD}}{R}$ to $v=V_{DD}$. Q point at intersection with diode characteristic, at coordinates $(I_D, V_D)$.
The constant-voltage-drop model assumes that a conducting diode maintains a constant voltage drop across its terminals, typically 0.7 V for silicon diodes and 0.3 V for Schottky diodes.
where $V_{on} = 0.7$ V for silicon diodes.
Consider a voltage divider with $V_S = 10$ V, $R_1 = 2$ kΩ, $R_2 = 3$ kΩ, and a diode in parallel with $R_2$.
Step 1: Assume diode is ON ($v_D = 0.7$ V)
Step 2: Current through $R_1$: \(i_{R1} = \frac{V_S - v_D}{R_1} = \frac{10 - 0.7}{2000} = 4.65 \text{ mA}\)
Step 3: Current through $R_2$: \(i_{R2} = \frac{v_D}{R_2} = \frac{0.7}{3000} = 0.233 \text{ mA}\)
Step 4: Diode current: \(i_D = i_{R1} - i_{R2} = 4.65 - 0.233 = 4.42 \text{ mA} > 0\)
Since $i_D > 0$, our assumption is correct.
For circuits with multiple diodes, test all possible combinations of ON/OFF states:
Example: Two diodes with different voltage sources
Choose the solution where all assumed states are consistent with calculated currents.
The ideal diode model treats the diode as a perfect switch: a short circuit when forward-biased and an open circuit when reverse-biased.
Consider a peak detector with an AC source $v_s = 10\sin(\omega t)$ V, diode, capacitor $C$, and load resistor $R_L$.
Analysis:
Result: Output voltage $v_o$ follows the peak of the input signal.
Half-wave rectifier with $v_s = V_m\sin(\omega t)$:
The small-signal model linearizes the diode around a DC operating point, useful for analyzing small AC variations superimposed on a DC bias.
The small-signal resistance (dynamic resistance) is:
\[r_d = \frac{dv}{di}\bigg|_{Q-point} = \frac{V_T}{I_D}\]where $I_D$ is the DC bias current at the Q-point.
At room temperature: $r_d = \frac{25 \text{ mV}}{I_D}$
Circuit: DC source $V_{DC} = 5$ V, AC source $v_{ac} = 0.1\sin(\omega t)$ V, resistor $R = 1$ kΩ, and diode in series.
Step 1: Find DC operating point (using constant-voltage-drop model): \(I_D = \frac{V_{DC} - 0.7}{R} = \frac{5 - 0.7}{1000} = 4.3 \text{ mA}\)
Step 2: Calculate small-signal resistance: \(r_d = \frac{25 \text{ mV}}{4.3 \text{ mA}} = 5.81 \text{ Ω}\)
Step 3: AC analysis - total AC resistance: \(R_{total} = R + r_d = 1000 + 5.81 ≈ 1006 \text{ Ω}\)
Step 4: AC voltage across diode: \(v_{d,ac} = v_{ac} \cdot \frac{r_d}{R + r_d} = 0.1 \cdot \frac{5.81}{1006} ≈ 0.58 \text{ mV}\)
The small-signal model is valid when: