What is the MOSFET Threshold Voltage?
The threshold voltage (Vth) of a MOSFET is the gate-to-source voltage at which a conducting channel begins to form between the source and drain. Below this voltage the transistor is essentially off; above it, the device turns on. Vth is one of the most important parameters in transistor and analog/digital circuit design, setting bias points, logic levels, and leakage behavior.
The Formula
For an n-channel (NMOS) device built on a p-type substrate, the long-channel threshold voltage is:
$$V_{th} = V_{fb} + 2\phi_F + \frac{\sqrt{2\,\varepsilon_s\, q\, N_a \cdot 2\phi_F}}{C_{ox}}$$
where \(V_{fb}\) is the flat-band voltage, \(\phi_F\) is the Fermi potential, \(\varepsilon_s = \varepsilon_r \cdot \varepsilon_0\) is the silicon permittivity (\(\varepsilon_0 \approx 8.854\times10^{-14}\) F/cm), \(q = 1.602\times10^{-19}\) C is the electron charge, \(N_a\) is the substrate acceptor doping concentration (cm⁻³), and \(C_{ox}\) is the gate-oxide capacitance per unit area (F/cm²). The \(2\phi_F\) term sets the strong-inversion surface potential, and the square-root term is the voltage needed to support the depletion-region (body) charge.
How to Use It
Enter the flat-band voltage, Fermi potential, substrate doping, oxide capacitance, and the relative permittivity of silicon (default 11.7). Keep all quantities in CGS-compatible units: doping in cm⁻³ and capacitance/permittivity per cm². The calculator returns \(V_{th}\) along with the surface-inversion term and the depletion-charge term so you can see how each contributes.
Worked Example
Take \(V_{fb} = -0.9\) V, \(\phi_F = 0.3\) V, \(N_a = 1\times10^{16}\) cm⁻³, \(C_{ox} = 3.45\times10^{-7}\) F/cm², \(\varepsilon_r = 11.7\). Then \(\varepsilon_s = 1.036\times10^{-12}\) F/cm. The inside of the root is $$2 \cdot \varepsilon_s \cdot q \cdot N_a \cdot 0.6 = 1.992\times10^{-15},$$ whose square root is \(\approx 4.463\times10^{-8}\) C/cm². Dividing by \(C_{ox}\) gives \(\approx 0.1294\) V. So $$V_{th} = -0.9 + 0.6 + 0.1294 \approx -0.171 \text{ V}.$$
Typical Threshold Voltage, Doping and Oxide Values
The values below are typical for silicon MOS devices at about 300 K. Use them as a sanity check for your inputs.
| Parameter | Typical range | Notes |
|---|---|---|
| Substrate doping \(N_a\) | about 1e15 to 1e18 cm-3 | Enter in cm-3, for example 1e16. |
| Gate oxide thickness \(t_{ox}\) | a few nm to about 100 nm | Thin oxides are used in modern CMOS. Thick oxides are used in power and high-voltage devices. |
| Oxide capacitance \(C_{ox}\) | about 3.45e-8 F/cm² (100 nm) to 1.7e-6 F/cm² (2 nm) | \(C_{ox} = 3.45\times10^{-13}\,\text{F/cm} / t_{ox}\), with \(t_{ox}\) in cm. This uses SiO2 with \(\varepsilon_r = 3.9\). Example: 10 nm = 1e-6 cm gives 3.45e-7 F/cm². |
| Fermi potential \(\phi_F\) | about 0.3 to 0.45 V | \(\phi_F = (kT/q)\ln(N_a/n_i)\) with \(n_i \approx 1.0\times10^{10}\,\text{cm}^{-3}\). This gives about 0.30 V at 1e15 cm-3, about 0.41 V at 1e17 and about 0.48 V at 1e18. |
| Flat-band voltage \(V_{FB}\) | about -0.5 to -1.2 V | Depends on the gate material (work function) and on oxide charge. |
| Enhancement-mode NMOS \(V_{th}\) | about 0.3 to 1 V (modern CMOS); about 2 to 4 V (power MOSFETs) | The exact value depends on the process and the device type. |
Interpreting Your Threshold Voltage Result
Sign. A positive \(V_{th}\) means an enhancement-mode NMOS device. It is off at \(V_{GS}=0\) and needs a positive gate voltage to form an inversion channel. A negative \(V_{th}\) means a depletion-mode device, which conducts at zero gate bias (normally on).
Worked example. Take \(V_{FB} = -1.2\) V, \(\phi_F = 0.30\) V, \(N_a = 10^{15}\) cm-3, \(t_{ox}=20\) nm so \(C_{ox} = 1.725\times10^{-7}\) F/cm², and \(\varepsilon_r = 11.7\), so \(\varepsilon_S = 1.036\times10^{-12}\) F/cm. The body-charge term is
$$\frac{\sqrt{2(1.036\times10^{-12})(1.602\times10^{-19})(10^{15})(0.60)}}{1.725\times10^{-7}} = \frac{1.411\times10^{-8}}{1.725\times10^{-7}} \approx 0.0818\ \text{V}$$
Then \(V_{th} = -1.2 + 0.60 + 0.0818 =\) -0.5182 V. This is negative, so the device is depletion-mode and normally on.
Doping and oxide. The last term is the depletion-charge (body) term. It grows with \(\sqrt{N_a}\), so higher doping raises \(V_{th}\). It is also divided by \(C_{ox}\), so a thinner oxide (larger \(C_{ox}\)) shrinks that term and moves \(V_{th}\) toward \(V_{FB} + 2\phi_F\). Higher doping also increases \(\phi_F\), which adds a second upward shift.
Leakage and subthreshold behavior. A lower \(V_{th}\) gives a stronger drive current at a given supply voltage. However, subthreshold leakage rises roughly exponentially as \(V_{th}\) falls, typically by a decade for every 60 to 100 mV. Designers trade speed against standby power when choosing \(V_{th}\).
Model limits. This is the long-channel, uniform-doping formula with zero source-to-body bias. It ignores short-channel effects, drain-induced barrier lowering (DIBL), the body effect, non-uniform doping, quantum effects in very thin oxides and poly-gate depletion. Treat the result as a first-order estimate. Real devices need measured data or process-specific models.
Constants and Material Parameters Used
| Quantity | Symbol | Value | Units |
|---|---|---|---|
| Elementary charge | \(q\) | 1.602e-19 | C |
| Vacuum permittivity | \(\varepsilon_0\) | 8.854e-14 | F/cm |
| Relative permittivity of silicon | \(\varepsilon_r\) (Si) | about 11.7 | dimensionless |
| Relative permittivity of SiO2 | \(\varepsilon_r\) (SiO2) | about 3.9 | dimensionless |
| Thermal voltage at 300 K | \(kT/q\) | about 0.0259 | V |
| Intrinsic carrier concentration of Si at 300 K | \(n_i\) | about 1.0e10 | cm-3 |
Unit note: the formula works in CGS-style units (cm, F/cm, cm-3). \(\varepsilon_0\) is therefore 8.854e-14 F/cm, which equals 8.854e-12 F/m in SI. Doping must be in cm-3 (1 cm-3 = 106 m-3), and \(C_{ox}\) must be in F/cm² (1 F/m² = 1e-4 F/cm²). Mixing SI and CGS values gives a body term that is wrong by orders of magnitude.
The silicon permittivity is \(\varepsilon_S = 11.7 \times 8.854\times10^{-14} \approx 1.036\times10^{-12}\) F/cm. Values such as \(n_i\) vary with temperature and with the source cited, so treat them as approximate.
FAQ
Does this work for PMOS? The same structure applies, but signs of \(V_{fb}\), \(\phi_F\), and doping type flip; this tool models the NMOS case.
What units should I use? Use cm-based units (cm⁻³ for doping, F/cm² for capacitance) so that the embedded constants \(q\) and \(\varepsilon_0\) are consistent.
Why is Vth sometimes negative? A negative \(V_{th}\) (depletion-mode device) can result when the flat-band voltage is strongly negative relative to the inversion and body terms.