IAE Academy

FOR CORE AND IT JOBS · CHAPTER 1 OF 7

The MOSFET as a switch and the CMOS inverter

This chapter treats the MOSFET as a voltage-controlled switch and builds the CMOS inverter from it. Its regions, switching threshold, noise margins and near-zero resting current are what you need to size and check a static CMOS gate.

Updated 30 Oct 2026

What you will be able to do

You will tell whether an nMOS or pMOS transistor is on from its terminal voltages. You will find each inverter transistor's region for any input. You will compute the switching threshold and see how W/L moves it. You will compute a matched inverter's noise margins and explain its near-zero steady current.

Before you start

  1. Kirchhoff's current law at a node

  2. Solving a linear equation and taking a square root

  3. Derivatives: the slope dy/dx, the product rule and the chain rule

  4. A MOSFET's gate, drain, source and body (body tied to source here), and its channel

  5. Carrier mobility and capacitance per unit area

The MOSFET as a voltage-controlled switch

A MOSFET conducts between drain and source only when a channel forms under its gate, so comparing the gate voltage with a threshold gives a switch. The nMOS source is its lower-voltage terminal: it conducts when the gate is high enough above the source. The pMOS source is its higher-voltage terminal: it conducts when the gate is low enough below it. V_DD is the supply, V_Tn the nMOS threshold and |V_Tp| the size of the negative pMOS threshold V_Tp.

SWITCH RULE

Otherwise the transistor is off and its drain current is taken as zero (leakage neglected). V_GS is the nMOS gate-to-source voltage, V_SG the pMOS source-to-gate voltage. Volts.

An on transistor has two regions. The channel's drain end vanishes when V_GS - V_DS falls to V_T, at V_DS = V_GS - V_T. Below this V_DS is the linear region (V_DS = 0 included); at or above it, saturation. The formulas are for the nMOS, with and overdrive ; for the pMOS use V_SG, V_SD and |V_Tp|. (carrier mobility times gate-oxide capacitance per unit area) is the process transconductance, or ; W/L is channel width over length. Channel-length modulation is neglected.

DRAIN CURRENT IN THE LINEAR REGION

Holds for V_GS > V_T and 0 ≤ V_DS < V_ov. The bracket's slope, V_ov - V_DS, is positive there, so I_D rises from zero at V_DS = 0 (zero only there) toward the saturation value. I_D in A, k' in A/V², voltages in V.

DRAIN CURRENT IN SATURATION

Holds for V_GS > V_T and V_DS ≥ V_ov; units as above. At V_DS = V_ov the linear bracket is V_ov²/2, so the two formulas meet at the boundary.

WORKED EXAMPLE

An nMOS has k'_n = 100 µA/V², W/L = 10, V_Tn = 1 V and V_GS = 3 V. Find I_D at V_DS = 3 V and 1 V.

Solution
  1. The transistor is on because V_GS = 3 V > V_Tn = 1 V. Its overdrive is 3 - 1 = 2 V, which sets the region boundary.

  2. At V_DS = 3 V we have V_DS ≥ 2 V, so the device is in saturation. I_D = (100 µA/V² / 2)(10)(2 V)² = 50 µA/V² × 10 × 4 V² = 2 mA.

  3. At V_DS = 1 V we have V_DS < 2 V, so the device is in the linear region. Here k'(W/L) = 1000 µA/V² = 1 mA/V², so I_D = 1 mA/V² × [2 V × 1 V - (1 V)²/2] = 1 mA/V² × 1.5 V² = 1.5 mA.

  4. Check: 1.5 mA is below the 2 mA saturation value, as it must be, since the linear current rises toward the saturation value as V_DS nears 2 V.

Answer: I_D = 2 mA at V_DS = 3 V (saturation) and 1.5 mA at V_DS = 1 V (linear).

The CMOS inverter and its transfer table

The inverter ties the gates of a pMOS and an nMOS together as the input. The pMOS pull-up, when on, connects the output to V_DD; the nMOS pull-down, when on, to ground. Each body is tied to its source. The output drives only the next gate, which takes no steady current, so by KCL the current the pMOS supplies equals the current the nMOS sinks. In the netlist, the nMOS has V_GS = V_in and V_DS = V_out; the pMOS has V_SG = V_DD - V_in and V_SD = V_DD - V_out.

Transistor

Type

Drain

Gate

Source

M1

pMOS

M2

nMOS

ground

This chapter assumes V_DD > V_Tn + |V_Tp|, so both are on for V_Tn < V_in < V_DD - |V_Tp|. By the boundary, an on nMOS is saturated while V_out ≥ V_in - V_Tn (V_DS ≥ V_GS - V_Tn), an on pMOS while V_out ≤ V_in + |V_Tp| (V_SD ≥ V_SG - |V_Tp|).

The currents decide the regions. As V_in rises, the nMOS saturation current grows and the pMOS one shrinks; V_M is the input where they are equal. Below V_M the nMOS one is smaller, so both cannot be saturated. Nor can the nMOS be linear: that needs V_out < V_in - V_Tn, which saturates the pMOS, and its current would exceed the most the nMOS can carry. So below V_M the nMOS is saturated and the pMOS linear; above V_M the roles swap.

Nearing V_M from below, the pMOS current nears its saturation value, so V_out nears the pMOS edge V_M + |V_Tp|; from above, V_M - V_Tn. At V_M any V_out between these meets both conditions, V_M included. For V_in ≤ V_Tn the nMOS is off, so the on pMOS carries zero current, which in the linear region means V_SD = 0: V_out = V_DD. The last row mirrors this.

V_in

nMOS, pMOS

V_out

off, linear

saturated, linear

saturated, saturated

linear, saturated

linear, off

0

An arrow shows the output falling across a row. The middle row is a vertical drop: with channel-length modulation neglected, both devices are saturated only at V_M. In the first and last rows one transistor is off, so no path joins V_DD to ground: a steady input below V_Tn or above V_DD - |V_Tp| draws only leakage. A static CMOS gate works the same way: steady inputs at 0 or V_DD turn off its pull-up or its pull-down network.

Switching threshold and noise margins

At V_M the output passes through V_out = V_in, so V_M is called the switching threshold. Each saturation current is zero at one end of the range V_Tn to V_DD - |V_Tp| and positive at the other, so they cross inside it: at V_M both devices are on, saturated and with positive overdrives. Equal currents fix V_M.

CURRENT BALANCE AT THE SWITCHING THRESHOLD

Both saturated, same current. Left: nMOS with V_GS = V_M. Right: pMOS with V_SG = V_DD - V_M. k' in A/V², voltages in V.

Take the positive square root of each side, since both overdrives are positive. With this gives , so .

SWITCHING THRESHOLD

Holds for V_DD > V_Tn + |V_Tp|; r dimensionless, voltages in V. V_M is a weighted average of V_Tn (weight r) and V_DD - |V_Tp| (weight 1). For r = 1 and V_Tn = |V_Tp|, V_M = V_DD/2.

A larger r, a stronger nMOS, weights V_Tn more: V_M moves down toward V_Tn. A smaller r moves it toward V_DD - |V_Tp|. For V_M = V_DD/2 with V_Tn = |V_Tp| you need r = 1, that is (W/L)_p/(W/L)_n = k'_n/k'_p. Since k'_n is usually larger than k'_p, the pMOS is then made wider.

WORKED EXAMPLE

V_DD = 3 V, V_Tn = |V_Tp| = 0.5 V, k'_n = 100 µA/V², k'_p = 40 µA/V², (W/L)_p = 5. Find V_M for (W/L)_n = 2 and 8.

Solution
  1. For (W/L)_n = 2: r² = (100 × 2)/(40 × 5) = 200/200 = 1, so r = 1. The two devices are equally strong.

  2. V_M = (3 - 0.5 + 1 × 0.5)/(1 + 1) = 3/2 = 1.5 V, which is V_DD/2 as expected for r = 1 with equal thresholds.

  3. For (W/L)_n = 8: r² = 800/200 = 4, so r = 2.

  4. V_M = (3 - 0.5 + 2 × 0.5)/(1 + 2) = 3.5/3 ≈ 1.17 V. This lies between V_Tn = 0.5 V and V_DD - |V_Tp| = 2.5 V, as it must.

  5. A stronger nMOS moved V_M down from 1.5 V to 1.17 V, as the argument above predicts.

Answer: V_M = 1.5 V for (W/L)_n = 2 and about 1.17 V for (W/L)_n = 8.

A noise margin is how far a logic level can be disturbed before the next stage misreads it. The gain is the slope of the transfer characteristic. V_IL and V_IH are the inputs where the gain is -1, the unity-gain points; between them its size exceeds 1, so noise is amplified. With output levels V_OH = V_DD and V_OL = 0 from the table, a high output may droop by and a low one rise by and still be read correctly.

Take a matched inverter: V_Tn = |V_Tp| = V_T and r = 1, so V_M = V_DD/2. Below V_M the nMOS is saturated and the pMOS linear. Divide their current balance by k'_n(W/L)_n = k'_p(W/L)_p.

CURRENT BALANCE BELOW THE SWITCHING THRESHOLD

Matched inverter, V_T < V_in < V_DD/2. Left: nMOS saturated. Right: pMOS linear, overdrive V_DD - V_in - V_T, V_SD = V_DD - V_out. V².

Differentiate both sides with respect to V_in (product and chain rules) and set . This gives , so . Put this back: the right side becomes . Multiply by 8 and expand; the terms cancel, leaving . Divide by .

UNITY-GAIN INPUTS AND NOISE MARGINS OF A MATCHED INVERTER

Holds for V_Tn = |V_Tp| = V_T, r = 1, V_DD > 2V_T; V_IL then lies between V_T and V_DD/2, as assumed. V_IH = V_DD - V_IL because swapping the devices' roles maps V_in to V_DD - V_in and V_out to V_DD - V_out. Volts.

Check yourself

CHECK YOURSELF 1

A CMOS inverter has V and V. The input is held steady at V. Which pair describes the nMOS and the pMOS, in that order?

  1. (A)

    nMOS off, pMOS linear

  2. (B)

    nMOS saturated, pMOS linear

  3. (C)

    nMOS linear, pMOS off

  4. (D)

    nMOS saturated, pMOS saturated

Answer and solution

Answer: (C)

  1. Compare the input with V.

  2. Since V, the input is in the last row of the transfer table.

  3. The pMOS has V, below V, so it is off.

  4. The nMOS has V V, so it is on. The off pMOS lets no current through, so the nMOS current is zero. In saturation it would carry a positive current, so it must be linear with : .

  5. So the nMOS is linear and the pMOS is off: option C.

CHECK YOURSELF 2

A CMOS inverter has V and V. The process gives and . The transistors have and . Find the switching threshold in volts.

Answer and solution

Answer: (accepted range )

  1. Compute , so .

  2. Use .

  3. V.

  4. Check: 0.6 V lies between V and V, and is below because the nMOS is the stronger device.

CHECK YOURSELF 3

A matched CMOS inverter has V, and V. Find its noise margin in volts. (Here .)

Answer and solution

Answer: (accepted range )

  1. The condition holds: .

  2. Use .

  3. V.

  4. Check: V lies between V and V, as assumed.

CHECK YOURSELF 4One or more options are correct

In a CMOS inverter you widen the nMOS, so that increases while , , and the pMOS stay unchanged. Which statements are correct? Assume .

  1. (A)

    moves down toward .

  2. (B)

    moves up toward .

  3. (C)

    stays between and .

  4. (D)

    becomes exactly whatever the thresholds.

Answer and solution

Answer: (A), (C)

  1. is a weighted average of (weight 1) and (weight ).

  2. The assumption means . A larger puts more weight on the smaller value , so moves down toward . A is correct and B is wrong.

  3. A weighted average with positive weights lies between its two values, so C is correct.

  4. Even at , , which equals only when . So D is wrong.

Used in these GATE topics