MOSFET basics
MOSFET basics: structure, operation regions, I–V equations, threshold and body effect, small-signal parameters, and typical exam problems with calculations.
Drafted with Aria, reviewed by the AiCanCode.org team · Updated . How we write and check our notes. Spotted an error? Use Give Feedback at the bottom of the page.
Why it matters
Metal–Oxide–Semiconductor Field-Effect Transistors (MOSFETs) are the basic building blocks of almost all modern Integrated Circuits (ICs), from microprocessors to memory and sensors. Engineers use MOSFETs as fast switches in digital logic (CMOS), as amplifying devices in analog circuits (op-amps, RF), and as power switches in motor drives and power supplies.
Key ideas
- The four terminals are: Gate (G), Drain (D), Source (S), and Body/Bulk (B). In many circuits B is tied to S.
- Types by polarity: n-channel MOSFET (nMOS) conducts with electrons (majority carriers in channel); p-channel MOSFET (pMOS) conducts with holes.
- Types by mode: Enhancement-mode MOSFET is normally OFF at V_GS = 0 and requires a threshold voltage to turn ON; Depletion-mode MOSFET is normally ON at V_GS = 0 and requires opposite gate bias to turn it OFF. In basic ICs, enhancement-mode devices are standard.
- Definition: Threshold voltage (V_T) is the gate-to-source voltage at which a strong inversion layer forms at the semiconductor–oxide interface, causing a conducting channel to appear between source and drain. For nMOS, conduction begins when V_GS ≥ V_T; for pMOS when V_SG ≥ |V_T|.
- Definition: Overdrive voltage (V_OV) is V_OV = V_GS − V_T (for nMOS). It represents how much the gate voltage exceeds threshold, controlling channel charge.
- Definition: Regions of operation of an enhancement-mode nMOS (assuming body tied to source so no body effect):
- Cutoff region: V_GS < V_T ⇒ channel absent, I_D ≈ 0.
- Triode/linear region: V_GS ≥ V_T and V_DS < V_OV ⇒ channel forms; current increases approximately linearly with V_DS.
- Saturation region: V_GS ≥ V_T and V_DS ≥ V_OV ⇒ channel pinches off near drain; current becomes mostly independent of V_DS (ideal long-channel), with slight increase if channel-length modulation is present.
- Physical picture (for nMOS on p-type body): With V_G positive relative to S, electrons are attracted to the Si–SiO2 interface under the gate, forming an inversion layer (the channel) that connects n+ source and drain so current can flow from drain to source when V_DS is applied.
Assumptions for textbook long-channel I–V model:
- Uniform channel doping, long channel (short-channel effects neglected), constant mobility, constant temperature, body tied to source (no body effect), and no velocity saturation.
Derivation (textbook level) of I–V in triode and saturation (nMOS):
- The incremental channel charge density per unit area under the gate is Q_c(x) = −C_ox[V_GS − V_T − V(x)], where C_ox is the oxide capacitance per unit area and V(x) is the local channel potential referenced to source (0 at source, V_DS at drain). Negative sign indicates electron charge.
- The drift current I_D is I_D = W·Q_c(x)·v_d(x) = W·μ_n·C_ox [V_GS − V_T − V(x)] · dV/dx, where W is channel width, μ_n is electron mobility, and v_d = −μ_n dV/dx.
- Since I_D is constant along x in steady state, separate variables and integrate from x = 0 (V = 0) to x = L (V = V_DS): ∫_0^{L} I_D dx = W μ_n C_ox ∫_0^{V_DS} [V_GS − V_T − V] dV ⇒ I_D·L = W μ_n C_ox [ (V_GS − V_T)V_DS − (V_DS^2)/2 ] Divide by L and define k_n = μ_n C_ox (W/L) to get triode equation: I_D = k_n [ (V_GS − V_T) V_DS − (V_DS^2)/2 ] for V_DS < V_OV.
- At the boundary V_DS = V_OV, the channel just pinches off at drain. For V_DS ≥ V_OV, the current saturates (ideal long-channel): I_D,sat = (1/2) k_n (V_OV)^2 = (1/2) k_n (V_GS − V_T)^2.
- Channel-length modulation accounts for slight I_D increase with V_DS in saturation, modeled by factor (1 + λ V_DS): I_D ≈ (1/2) k_n (V_GS − V_T)^2 (1 + λ V_DS) in saturation, where λ is the channel-length modulation parameter (units V⁻¹).
pMOS adaptation (magnitude form):
- Use V_SG, V_SD and |V_Tp| with the same equations in terms of magnitudes, or replace n by p, μ_p for mobility, and reverse voltage polarities and current direction convention. Exam answers often state: “Replace n by p and invert voltage polarities.”
Definition: Body effect (substrate-bias effect) is the increase in threshold voltage when the source–bulk junction is reverse-biased. The empirical formula (for nMOS) is:
- V_T = V_T0 + γ ( √(2φ_F + V_SB) − √(2φ_F) ), where V_T0 is threshold at V_SB = 0, γ is the body-effect coefficient (V^0.5), φ_F is the Fermi potential (positive magnitude used here), and V_SB is source-to-bulk voltage (≥ 0 for reverse bias in nMOS). Assumes uniform doping and moderate reverse bias.
Definition: Transconductance parameter (k_n) is k_n = μ_n C_ox (W/L), with units A/V². It sets how strongly gate voltage controls drain current.
Definition: Small-signal transconductance (g_m) is g_m = ∂I_D/∂V_GS at the bias point. For long-channel saturation (λ ≈ 0): g_m = k_n (V_GS − V_T) = 2 I_D / (V_GS − V_T). Units: siemens (S).
Symbols and recommended diagram descriptions:
- MOSFET symbols: Draw nMOS with a three-terminal FET symbol (source arrow absent in many MOS symbols). Show body tied to source with a line if required. Mark G, D, S, and body diode from body (p) to drain/source (n+) for nMOS (anode at body, cathode at source/drain). For pMOS, reverse diode.
- Cross-section (nMOS): Draw p-type substrate, two n+ regions (source, drain), gate oxide (SiO2) above channel, polysilicon gate on top, metal contacts to S, D, and G. Label W (width) into the page and L (length) along the surface channel. Indicate V_G, V_S, V_D, and the inversion channel under positive V_G.
Operating as switch and amplifier:
- As a switch: In digital CMOS, an nMOS passes strong ‘0’ (low) well and weak ‘1’; a pMOS passes strong ‘1’ well and weak ‘0’. A complementary pair (CMOS transmission gate or inverter) gives rail-to-rail switching with very low static power.
- As an amplifier: Bias an nMOS in saturation with a drain resistor and a DC gate bias. Small AC variations on V_GS produce proportional variations in I_D, generating output voltage swings across the load resistor. Gain ≈ g_m·R_D (magnitude), assuming output resistance r_o = 1/(λ I_D) is large.
Formulas
I_D = 0 — I_D: drain current (A). Cutoff region when V_GS < V_T (nMOS) or V_SG < |V_T| (pMOS).
I_D = k_n [ (V_GS − V_T)·V_DS − (V_DS²)/2 ] — nMOS triode/linear region; k_n = μ_n·C_ox·(W/L) (A/V²), V_GS, V_T, V_DS in volts. Valid when V_GS ≥ V_T and 0 ≤ V_DS < V_GS − V_T.
I_D = (1/2)·k_n·(V_GS − V_T)² — nMOS saturation (ideal long-channel, no channel-length modulation). Valid when V_DS ≥ V_GS − V_T.
I_D ≈ (1/2)·k_n·(V_GS − V_T)²·(1 + λ·V_DS) — nMOS saturation including channel-length modulation. λ: channel-length modulation parameter (V⁻¹). Valid in saturation region.
k_n = μ_n·C_ox·(W/L) — Transconductance parameter. μ_n: electron mobility (m²/(V·s)); C_ox: oxide capacitance per unit area (F/m²); W, L in meters.
V_OV = V_GS − V_T — Overdrive voltage (V). Applies to both nMOS and pMOS with sign conventions.
g_m = ∂I_D/∂V_GS — Small-signal transconductance (S). General definition.
g_m = k_n·(V_GS − V_T) = 2·I_D / (V_GS − V_T) — For long-channel saturation (λ ≈ 0). V in volts, I_D in amperes.
r_o = 1 / (λ·I_D) — Output resistance in saturation due to channel-length modulation (Ω). Valid when λ modeling holds.
V_T = V_T0 + γ·( √(2·φ_F + V_SB) − √(2·φ_F) ) — Body effect (nMOS). V_T0 (V), γ (V^0.5), φ_F (V), V_SB (V). Valid for moderate reverse bias and uniform doping.
I_Dp = k_p [ (V_SG − |V_Tp|)·V_SD − (V_SD²)/2 ] — pMOS triode in magnitude form. k_p = μ_p·C_ox·(W/L) (A/V²); use appropriate sign in circuit equations.
I_Dp = (1/2)·k_p·(V_SG − |V_Tp|)² — pMOS saturation in magnitude form.
Worked examples
Example 1: Compute I_D in linear and saturation Given: Enhancement nMOS, μ_n·C_ox = 200 μA/V², W/L = 10, so k_n = μ_n·C_ox·(W/L). Threshold V_T = 1.0 V. Case A (linear): V_GS = 3.0 V, V_DS = 0.20 V. Case B (saturation): V_GS = 3.0 V, V_DS = 5.0 V, ignore channel-length modulation (λ ≈ 0).
Steps:
- Compute k_n. Formula: k_n = μ_n·C_ox·(W/L) Numbers: k_n = (200×10⁻⁶ A/V²)·10 = 2.0×10⁻³ A/V² = 2.0 mA/V².
- Check region for Case A: V_OV = V_GS − V_T = 3.0 − 1.0 = 2.0 V. Since V_DS = 0.20 V < 2.0 V ⇒ triode.
- Case A current. Formula: I_D = k_n [ (V_GS − V_T)·V_DS − (V_DS²)/2 ] Numbers: I_D = 2.0×10⁻³ [ (2.0)(0.20) − (0.20)²/2 ] A = 2.0×10⁻³ [ 0.40 − 0.040/2 ] A = 2.0×10⁻³ [ 0.40 − 0.020 ] A = 2.0×10⁻³ × 0.380 A = 0.000760 A = 0.760 mA.
- Case B region: V_DS = 5.0 V ≥ V_OV = 2.0 V ⇒ saturation (ideal).
- Case B current. Formula: I_D = (1/2)·k_n·(V_GS − V_T)² Numbers: I_D = 0.5 × 2.0×10⁻³ × (2.0)² A = 1.0×10⁻³ × 4.0 A = 4.00 mA.
Example 2: Gate voltage for desired current in saturation; include channel-length modulation effect estimate Given: Same device as Example 1 (k_n = 2.0 mA/V², V_T = 1.0 V). Desired I_D = 1.00 mA in saturation. Assume initially λ = 0; then refine with λ = 0.020 V⁻¹ at V_DS = 4.0 V.
Steps:
- Ignore λ and solve for V_GS. Formula: I_D = (1/2)·k_n·(V_GS − V_T)² Rearranged: V_GS − V_T = √(2·I_D/k_n) Numbers: V_OV = √(2·1.00×10⁻³ / 2.0×10⁻³) V = √(1.0) V = 1.00 V ⇒ V_GS = 1.0 + 1.0 = 2.00 V.
- Include λ to estimate actual I_D at V_DS = 4.0 V if gate kept at 2.00 V. Formula: I_D ≈ (1/2)·k_n·V_OV²·(1 + λ·V_DS) Numbers: I_D ≈ 0.5·2.0×10⁻³·(1.00)²·(1 + 0.020·4.0) A = 1.0×10⁻³·(1 + 0.080) A = 1.08 mA.
- To achieve exactly 1.00 mA at V_DS = 4.0 V with λ ≠ 0, reduce V_OV slightly. Solve: 1.00×10⁻³ = 0.5·2.0×10⁻³·V_OV²·(1 + 0.080) ⇒ V_OV² = (1.00×10⁻³)/(1.0×10⁻³·1.080) = 0.9259 ⇒ V_OV = 0.9623 V (principal root) ⇒ V_GS = V_T + V_OV = 1.962 V (approx.).
Example 3: Threshold increase due to body effect Given: nMOS with V_T0 = 0.70 V, γ = 0.40 V^0.5, 2φ_F = 0.60 V. Source-to-bulk voltage V_SB = 0.50 V (source at higher potential than bulk, reverse-bias for nMOS).
Steps:
- Compute square roots. √(2φ_F) = √(0.60) = 0.7746 V^0.5 (approx.) √(2φ_F + V_SB) = √(0.60 + 0.50) = √(1.10) = 1.0488 V^0.5 (approx.)
- Apply body-effect formula. Formula: V_T = V_T0 + γ ( √(2φ_F + V_SB) − √(2φ_F) ) Numbers: V_T = 0.70 + 0.40 (1.0488 − 0.7746) V = 0.70 + 0.40 × 0.2742 V = 0.70 + 0.1097 V = 0.8097 V ≈ 0.810 V.
Model theory question (short answer sketch): Explain CMOS inverter operation
- A CMOS inverter uses a pMOS at the top (source to V_DD) and an nMOS at the bottom (source to GND), gates tied together as input, drains tied as output.
- When input is low (0 V), pMOS gate–source is high (V_SG ≈ V_DD), it turns ON; nMOS gate–source is 0 V, it is OFF; output charges to V_DD (logic 1) through pMOS.
- When input is high (≈ V_DD), pMOS gate–source is ≈ 0 V (OFF), nMOS gate–source is high (ON); output discharges to GND (logic 0) through nMOS.
- In steady states, one device is OFF, so static current is nearly zero; switching power dominates. The switching threshold depends on device sizing and V_T.
Common mistakes
- Forgetting the (V_DS²)/2 term in the triode equation, leading to overestimated current.
- Ignoring body effect when source is not at the same potential as bulk, giving wrong V_T and I_D.
- Wrong sign conventions for pMOS: using V_GS instead of V_SG, or forgetting to take |V_Tp|.
- Assuming truly zero input current at the gate in power devices; while DC gate current is ~0 (insulator), dynamic gate charge and leakage exist.
- Neglecting channel-length modulation in analog bias calculations, causing gain or current errors.
- Unit mistakes: not converting μA/V² to A/V²; dropping factors of 10.
- Drawing incomplete symbols: missing body diode, unlabeled terminals, or wrong diode orientation.
For your exam
- Short notes: Enhancement vs depletion MOSFET; CMOS inverter basics; channel-length modulation. Full marks: concise explanation, one labeled diagram (symbol or cross-section), key formula or condition.
- Derivations: Long-channel I–V equation from charge-sheet model; saturation current; body-effect threshold equation. Full marks: assumptions listed, step-by-step integration, clear region conditions, final expression boxed.
- Numericals: Compute I_D for given V_GS, V_DS, W/L, μ_n·C_ox, V_T; find V_GS for target I_D; include λ or body effect. Full marks: choose correct region, write formula before substituting, carry SI units, and state final answer clearly.
- Diagrams: nMOS cross-section and symbol with body diode; characteristic curves I_D–V_DS showing regions. Full marks: all terminals labeled (G, D, S, B), indicate p/n regions, oxide, channel, and mark region boundaries on plots (V_DS = V_OV line).
Quick check
- State the condition for saturation in an enhancement nMOS and give the ideal saturation current expression.
- An nMOS has k_n = 1.5 mA/V², V_T = 0.8 V, V_GS = 2.3 V, V_DS = 0.5 V. Which region is it in, and what is I_D (ideal long-channel)?
- What is the effect of a positive V_SB on V_T in an nMOS, and which formula relates them?
Answers: 1) Saturation when V_DS ≥ V_GS − V_T; I_D = (1/2)·k_n·(V_GS − V_T)². 2) V_OV = 1.5 V, V_DS = 0.5 V < 1.5 V ⇒ triode; I_D = 1.5×10⁻³ [ (1.5)(0.5) − (0.5)²/2 ] = 1.5×10⁻³ [0.75 − 0.125] = 1.5×10⁻³·0.625 = 0.0009375 A = 0.938 mA. 3) Positive V_SB increases V_T (body effect); V_T = V_T0 + γ(√(2φ_F + V_SB) − √(2φ_F)).
Finished this topic? Mark it so your progress, study plan and readiness keep up.
Stuck on something here?