Circuits I
DC and AC circuit analysis: Ohm law, KVL/KCL, node/mesh analysis, Thevenin and Norton equivalents, op-amps, and transient response (RC/RL/RLC).
Introduction; Lumped Element Abstraction
The Lumped Element Abstraction In real circuits, signals propagate at the speed of light. Every wire is an antenna, every component has parasitic capacitance, and Maxwell's equations describe...
The Lumped Element Abstraction
In real circuits, signals propagate at the speed of light. Every wire is an antenna, every component has parasitic capacitance, and Maxwell's equations describe reality precisely - but they are incredibly hard to solve.
The lumped element abstraction lets us ignore most of this. We model a circuit as a network of ideal components (R, L, C, V-source, I-source) connected by ideal wires. The abstraction is valid when the circuit's physical size is much smaller than the wavelength of the signals.
When Does the Abstraction Break Down?
Rule of thumb: the lumped model is valid when the propagation delay through the circuit is much smaller than the signal's rise time:
t_propagation << t_rise
Where:
- t_propagation = L / v (L = length, v ~ 2x10^8 m/s for PCB traces)
- t_rise = rise time of the fastest signal
Example: A 10 cm PCB trace with a 1 ns rise time:
- t_prop = 0.1 / (2x10^8) = 0.5 ns
- 0.5 ns vs 1 ns: the abstraction is marginal
- Rule of thumb: need t_prop < t_rise / 10, so 0.5 ns < 0.1 ns fails
At 100 MHz+ or with fast digital signals, you must think about transmission lines.
Ideal Circuit Elements
Resistor (R): V = I * R (Ohm's law). Unit: ohm.
- Power: P = V*I = I^2*R = V^2/R
- Dissipates energy as heat
Capacitor (C): I = C * dV/dt. Unit: farad.
- Stores energy in electric field: E = 0.5 * C * V^2
- Blocks DC, passes AC
Inductor (L): V = L * dI/dt. Unit: henry.
- Stores energy in magnetic field: E = 0.5 * L * I^2
- Passes DC, blocks AC
Voltage source: provides V regardless of current drawn.
Current source: provides I regardless of voltage across it.
Kirchhoff's Laws
Kirchhoff's Current Law (KCL): The sum of currents entering any node equals the sum leaving.
Sum I_in = Sum I_out
Kirchhoff's Voltage Law (KVL): The sum of voltage drops around any closed loop is zero.
Sum V_loop = 0
Series and Parallel Resistors
Series: R_eq = R1 + R2 + ... + Rn
--[R1]--[R2]--[R3]-- equivalent to --[R_eq]--
Parallel: 1/R_eq = 1/R1 + 1/R2 + ... + 1/Rn
For two: R_eq = (R1 * R2) / (R1 + R2)
+--[R1]--+
----+ +---- equivalent to --[R_eq]--
+--[R2]--+
Voltage Divider
R1
Vin --/\/\/\--+-- Vout
|
R2
|
GND
Vout = Vin * R2 / (R1 + R2)
This is the most common circuit in all of electronics. Use it to:
- Scale a sensor output to a safe ADC range
- Set a reference voltage
- Bias a transistor
Current draw: I = Vin / (R1 + R2). Make R1+R2 large enough to avoid wasting power, but small enough to avoid loading effects.
Example Problem
Question: Design a voltage divider to produce 3.3 V from a 5 V source, with total current draw under 1 mA.
Solution:
- Need R2 / (R1 + R2) = 3.3 / 5 = 0.66
- Total R >= 5 V / 1 mA = 5 kohm
- Pick R1 = 1.7 kohm, R2 = 3.3 kohm
- Verify: 5 * 3.3 / (1.7 + 3.3) = 5 * 3.3 / 5 = 3.3 V correct
- Current: 5 V / 5 kohm = 1 mA correct
Power in Circuits
Every element has a power rating. Exceed it and it smokes.
- Resistor: typical 0.25 W, 0.5 W
- Small LED: 20 mA, V_f ~ 2 V, P = 40 mW
- ESP32 GPIO pin: 20 mA max, 3.3 V, P = 66 mW
Use P = I^2*R or P = V^2/R to check before connecting.
Why This Matters
The lumped model is the foundation of all circuit analysis. Every SPICE simulator uses it. Every schematic capture tool assumes it. Without it, we'd have to solve Maxwell's equations for every LED blinker - which would take hours instead of milliseconds.
The abstraction lets us reason about circuits using algebra, not calculus. It's one of the most successful engineering abstractions ever devised.
Key Takeaways
- Lumped model valid when t_prop << t_rise
- KCL: currents in = currents out at every node
- KVL: voltages around any loop sum to zero
- Series: R adds; Parallel: 1/R adds
- Voltage divider: Vout = Vin * R2/(R1+R2)
- Check power ratings BEFORE connecting
Basic Circuit Analysis Method
We introduce the node method for solving linear circuits. Instead of writing KVL and KCL repeatedly, the node method uses KCL at every node with unknown node voltages as the variables. We derive the...
We introduce the node method for solving linear circuits. Instead of writing KVL and KCL repeatedly, the node method uses KCL at every node with unknown node voltages as the variables. We derive the method formally and apply it to several examples including a voltage divider, a current source with parallel resistances, and a bridge network.
Superposition, Thevenin, and Norton
Superposition: in a linear circuit with multiple independent sources, the total response is the sum of responses to each source acting alone. Thevenin theorem: any linear two-terminal network can be...
Superposition: in a linear circuit with multiple independent sources, the total response is the sum of responses to each source acting alone. Thevenin theorem: any linear two-terminal network can be replaced by a single voltage source V_th in series with a resistance R_th. Norton theorem: dual — a current source I_n in parallel with R_n. We show how to compute these values and when to use each.
The Digital Abstraction
Digital circuits are built on a specific discipline: voltages below V_IL are treated as 0, voltages above V_IH as 1, and we guarantee that outputs stay within valid ranges. This lecture covers noise...
Digital circuits are built on a specific discipline: voltages below V_IL are treated as 0, voltages above V_IH as 1, and we guarantee that outputs stay within valid ranges. This lecture covers noise margins, static discipline, and the relationship between analog reality and digital abstraction.
Inside the Digital Gate
We open the black box: the MOSFET as a switch, the resistor pull-up and pull-down networks, and how CMOS logic gates actually work. This lecture explains why CMOS dominates modern digital design: low...
We open the black box: the MOSFET as a switch, the resistor pull-up and pull-down networks, and how CMOS logic gates actually work. This lecture explains why CMOS dominates modern digital design: low static power, high noise margin, and scalable geometry.
Nonlinear Elements: Diodes and MOSFETs
Real diodes follow the Shockley equation; real MOSFETs have a quadratic I-V characteristic. We introduce nonlinear elements and demonstrate how to solve circuits containing them using graphical and...
Real diodes follow the Shockley equation; real MOSFETs have a quadratic I-V characteristic. We introduce nonlinear elements and demonstrate how to solve circuits containing them using graphical and iterative methods.
Amplifiers and the Small-Signal Model
An amplifier takes a small input signal and produces a larger output. We introduce the linear amplifier abstraction and the small-signal model, then apply it to analyze the MOSFET common-source...
An amplifier takes a small input signal and produces a larger output. We introduce the linear amplifier abstraction and the small-signal model, then apply it to analyze the MOSFET common-source amplifier.
Capacitors and Inductors: Energy Storage
Capacitors store energy in electric fields; inductors store energy in magnetic fields. We define their I-V relationships, show how they behave under DC steady-state, and introduce the concept of...
Capacitors store energy in electric fields; inductors store energy in magnetic fields. We define their I-V relationships, show how they behave under DC steady-state, and introduce the concept of state variables.
📝 Problem Set 1: Lumped Abstraction & KVL/KCL
Q1. For a signal with rise time of 1 ns and a circuit board trace length of 10 cm (signal speed ~2x10^8 m/s), is the lumped-element abstraction valid? Explain.
Show Solution
Propagation delay = 0.1 m / (2x10^8 m/s) = 0.5 ns. Rise time = 1 ns. Since 0.5 ns is comparable to 1 ns (not much smaller), the lumped abstraction is MARGINAL — distributed effects may matter. Typically we require propagation delay < rise time / 10, so 0.5 ns < 0.1 ns fails.
Q2. A circuit has three resistors in parallel: R1=100 ohm, R2=200 ohm, R3=400 ohm. What is the equivalent resistance?
Show Solution
1/R_eq = 1/100 + 1/200 + 1/400 = 4/400 + 2/400 + 1/400 = 7/400. Therefore R_eq = 400/7 ≈ 57.14 ohm.
Q3. In a circuit with a 12V source, R1=1k ohm in series with a parallel combination of R2=2k ohm and R3=3k ohm, find the current through R1, the voltage across R2, and the power dissipated in R3.
Show Solution
R_parallel = (2k x 3k) / (2k + 3k) = 6k/5k = 1.2k ohm. R_total = 1k + 1.2k = 2.2k ohm. I_R1 = 12V / 2.2k ohm ≈ 5.45 mA. V_parallel = 5.45mA x 1.2k ohm ≈ 6.55V. So V_R2 = V_R3 = 6.55V. Power in R3 = V^2/R = (6.55)^2 / 3000 ≈ 14.3 mW.
📝 Problem Set 2: Thevenin and Norton Equivalents
Q1. Find the Thevenin equivalent of a circuit consisting of a 12V source in series with 4k ohm, followed by a parallel branch of 6k ohm.
Show Solution
V_th (open circuit): voltage divider = 12V x 6k/(4k+6k) = 7.2V. R_th: with 12V source shorted, R seen = 4k || 6k = 2.4k ohm. So Thevenin = 7.2V source + 2.4k ohm series resistor.
Q2. Convert the Thevenin circuit from Q1 to its Norton equivalent.
Show Solution
I_n = 7.2V / 2.4k ohm = 3 mA. R_n = 2.4k ohm. Norton = 3 mA current source in parallel with 2.4k ohm.
📝 Problem Set 3: Op-Amps and Transients
Q1. Design a non-inverting op-amp amplifier with a gain of 11 and an input impedance of at least 100 kΩ. Choose standard resistor values and verify.
Show Solution
Gain = 1 + R2/R1 = 11, so R2/R1 = 10. Choose R1 = 10 kΩ, R2 = 100 kΩ. Non-inverting input impedance is set by the op-amp's input stage, typically 10^12 Ω (CMOS) or 10^6 Ω (BJT), so >100 kΩ is easily satisfied. Verify: gain = 1 + 100/10 = 11. ✓
Q2. For an RC circuit with R=2 kΩ, C=100 nF, and V_in stepping from 0V to 10V at t=0, find the time constant, the voltage at t=τ, and the time to reach 99% of final value.
Show Solution
τ = RC = 2000 × 100e-9 = 200 µs. At t=τ, V_C = 10 × (1 - e^-1) = 10 × 0.632 = 6.32 V. 99% is reached at t = 5τ = 1000 µs = 1 ms.
Q3. Calculate the RMS value of a sine wave with amplitude 5V, and compute the average power it delivers to a 50Ω resistor.
Show Solution
V_RMS = 5 / √2 = 3.54 V. P = 3.54² / 50 = 12.5 / 50 = 0.25 W.
🎯 Midterm 1: Node, Mesh, Superposition, Thevenin
Q1. For the circuit shown (10V source, 5 ohm in series with 10 ohm, then 20 ohm parallel), find the voltage across the 20 ohm resistor.
Show Solution
R_parallel = 10 || 20 = 6.67 ohm. R_total = 5 + 6.67 = 11.67 ohm. I_total = 10/11.67 = 0.857A. V_parallel = 0.857 x 6.67 = 5.71V. So V_20 = 5.71V.
Q2. Find the Thevenin equivalent of a balanced bridge with R1=R2=R3=R4=1k ohm and a 12V source across the bridge.
Show Solution
For a balanced bridge, V_th = 0 (the two output nodes are at the same potential). R_th = R1||R3 + R2||R4 = 500 + 500 = 1k ohm.
Q3. Design a voltage divider to produce 3.3V from a 5V source, with total current draw under 1mA.
Show Solution
V_out = V_in x R2/(R1+R2) = 3.3 means R2/(R1+R2) = 0.66. Total R >= 5V/1mA = 5k ohm. Choose R1 = 1.7k ohm, R2 = 3.3k ohm. Verify: 5 x 3.3/(1.7+3.3) = 5 x 3.3/5 = 3.3V. Current = 5V/5k ohm = 1mA.
Q4. Analyze the transient response of an RC circuit with R=1k ohm, C=1 uF, V_source stepping from 0 to 5V at t=0.
Show Solution
tau = RC = 1ms. V_C(t) = 5(1 - e^(-t/tau)). At t=1ms, V_C = 5 x 0.632 = 3.16V. Current I(t) = (V_source/R) x e^(-t/tau).
🎯 Final Exam: All Topics
Q1. Design an op-amp circuit to amplify a 100mV signal to 2V with an input impedance of at least 100k ohm.
Show Solution
Gain = 20. Use non-inverting configuration: 1 + R2/R1 = 20 means R2/R1 = 19. For input impedance > 100k ohm, use JFET-input op-amp (e.g., TL071). Choose R1 = 10k ohm, R2 = 190k ohm.
Q2. For an RLC series circuit with R=10 ohm, L=1mH, C=1 uF, find the resonant frequency and Q factor.
Show Solution
omega_0 = 1/sqrt(LC) = 1/sqrt(10^-3 x 10^-6) = 1/sqrt(10^-9) = 31623 rad/s. f_0 = 5.03 kHz. Q = omega_0 L / R = 31623 x 10^-3 / 10 = 3.16. Bandwidth = f_0/Q = 1.59 kHz.
Q3. A CMOS inverter has V_DD = 3.3V. If V_IL = 1.0V and V_IH = 2.3V, calculate the noise margins NM_H and NM_L.
Show Solution
NM_H = V_DD - V_IH = 3.3 - 2.3 = 1.0V. NM_L = V_IL - 0 = 1.0V. Both are 1.0V — a well-designed inverter.
🎯 Midterm 2: Transients, Op-Amps, Frequency Response
Q1. Design an inverting amplifier with gain -10 using an op-amp with an input resistance of 1 kΩ. What is the bandwidth if the op-amp GBW is 1 MHz?
Show Solution
Gain = -R2/R1 = -10, so R2/R1 = 10. Choose R1 = 1kΩ, R2 = 10kΩ. Bandwidth = GBW / |Gain| = 1 MHz / 10 = 100 kHz.
Q2. A parallel RLC circuit has R=1kΩ, L=1mH, C=1µF. Find the resonant frequency and the Q factor.
Show Solution
f_0 = 1/(2π√(LC)) = 1/(2π√(1e-3 × 1e-6)) = 1/(2π × 3.16e-5) = 5.03 kHz. Q = R√(C/L) = 1000 × √(1e-6/1e-3) = 1000 × 0.0316 = 31.6.
Q3. Given the circuit with Vin=12V, R1=4.7kΩ (to node A), R2=2.2kΩ (node A to ground), R3=3.3kΩ (node A to node B), R4=1kΩ (node B to ground), find V_A and V_B using nodal analysis.
Show Solution
Node A: (V_A-12)/4.7k + V_A/2.2k + (V_A-V_B)/3.3k = 0. Node B: (V_B-V_A)/3.3k + V_B/1k = 0. Solve: V_A(1/4.7 + 1/2.2 + 1/3.3) - V_B/3.3 = 12/4.7. From B: V_A/3.3 = V_B(1/3.3 + 1/1) = V_B × 1.303, V_A = 4.3 V_B. Substitute: 4.3V_B × 0.637 - V_B/3.3 = 2.553. 2.739V_B - 0.303V_B = 2.553. 2.436V_B = 2.553. V_B = 1.048 V. V_A = 4.3 × 1.048 = 4.5 V.